Path: gmdzi!unido!fauern!math.fu-berlin.de!msi.umn.edu!news.cs.indiana.edu!purdue!spaf
From: s...@cs.purdue.EDU (Gene Spafford)
Newsgroups: news.announce.newusers,news.admin
Subject: What is Usenet?
Message-ID: <14692@ector.cs.purdue.edu>
Date: 21 May 91 04:49:10 GMT
Expires: 19 Aug 91 04:49:10 GMT
Followup-To: news.announce.newusers
Organization: Dept. of Computer Sciences, Purdue Univ.
Lines: 261
Approved: s...@cs.purdue.EDU
Original from: c...@count.tct.com (Chip Salzenberg)
[Most recent change: 19 May 1991 by s...@cs.purdue.edu (Gene Spafford)]
The first thing to understand about Usenet is that it is widely
misunderstood. Every day on Usenet, the "blind men and the elephant"
phenomenon is evident, in spades. In the opinion of the author, more
flame wars arise because of a lack of understanding of the nature of
Usenet than from any other source. And consider that such flame wars
arise, of necessity, among people who are on Usenet. Imagine, then,
how poorly understood Usenet must be by those outside!
Any essay on the nature of Usenet cannot ignore the erroneous
impressions held by many Usenet users. Therefore, this article will
treat falsehoods first. Keep reading for truth. (Beauty, alas, is
not relevant to Usenet.)
WHAT USENET IS NOT
------------------
1. Usenet is not an organization.
Usenet has no central authority. In fact, it has no central
anything. There is a vague notion of "upstream" and "downstream"
related to the direction of high-volume news flow. It follows
that, to the extent that "upstream" sites decide what traffic
they will carry for their "downstream" neighbors, that "upstream"
sites have some influence on their neighbors. But such influence
is usually easy to circumvent, and heavy-handed manipulation
typically results in a backlash of resentment.
2. Usenet is not a democracy.
A democracy can be loosely defined as "government of the people,
by the people, for the people." However, as explained above,
Usenet is not an organization, and only an organization can be run
as a democracy. Even a democracy must be organized, for if it
lacks a means of enforcing the peoples' wishes, then it may as
well not exist.
Some people wish that Usenet were a democracy. Many people
pretend that it is. Both groups are sadly deluded.
3. Usenet is not fair.
After all, who shall decide what's fair? For that matter, if
someone is behaving unfairly, who's going to stop him? Neither
you nor I, that's certain.
4. Usenet is not a right.
Some people misunderstand their local right of "freedom of speech"
to mean that they have a legal right to use others' computers to
say what they wish in whatever way they wish, and the owners of
said computers have no right to stop them.
Those people are wrong. Freedom of speech also means freedom not
to speak; if I choose not to use my computer to aid your speech,
that is my right. Freedom of the press belongs to those who own
one.
5. Usenet is not a public utility.
Some Usenet sites are publically funded or subsidized. Most of
them, by plain count, are not. There is no government monopoly
on Usenet, and little or no control.
6. Usenet is not a commercial network.
Many Usenet sites are academic or government organizations; in
fact, Usenet originated in academia. Therefore, there is a Usenet
custom of keeping commercial traffic to a minimum. If such
commercial traffic is generally considered worth carrying, then it
may be grudgingly tolerated. Even so, it is usually separated
somehow from non-commercial traffic; see "comp.newprod."
7. Usenet is not the Internet.
The Internet is a wide-ranging network, parts of which are
subsidized by various governments. The Internet carries many
kinds of traffic; Usenet is only one of them. And the Internet is
only one of the various networks carrying Usenet traffic.
8. Usenet is not a UUCP network.
UUCP is a protocol (some might say "protocol suite," but that's a
technical point) for sending data over point-to-point connections,
typically using dialup modems. Usenet is only one of the various
kinds of traffic carried via UUCP, and UUCP is only one of the
various transports carrying Usenet traffic.
9. Usenet is not a UNIX network, nor even an ASCII network.
Don't assume that everyone is using "rn" on a UNIX machine. There
are Vaxen running VMS, IBM mainframes, Amigas, and MS-DOS PCs
reading and posting to Usenet. And, yes, some of them use
(shudder) EBCDIC. Ignore them if you like, but they're out there.
10. Usenet is not software.
There are dozens of software packages used at various sites to
transport and read Usenet articles. So no one program or package
can be called "the Usenet software."
Software designed to support Usenet traffic can be (and is) used
for other kinds of communication, usually without risk of mixing
the two. Such private communication networks are typically kept
distinct from Usenet by the invention of newsgroup names different
from the universally-recognized ones.
Well, enough negativity.
WHAT USENET IS
--------------
Usenet is the set of machines that exchange articles tagged with one
or more universally-recognized labels, called "newsgroups" (or
"groups" for short).
(Note that the term "newsgroup" is correct, while "area," "base,"
"board," "bboard," "conference," "round table," "SIG," etc. are
incorrect. If you want to be understood, be accurate.)
DIVERSITY
---------
If the above definition of Usenet sounds vague, that's because it is.
It is almost impossible to generalize over all Usenet sites in any
non-trivial way. Usenet encompasses government agencies, large
universities, high schools, businesses of all sizes, home computers of
all descriptions, etc, etc.
CONTROL
-------
Every administrator controls his own site. No one has any real
control over any site but his own.
The administrator gets his power from the owner of the system he
administers. As long as the owner is happy with the job the
administrator is doing, he can do whatever he pleases, up to and
including cutting off Usenet entirely. Them's the breaks.
PROPAGATION
-----------
In the old days, when UUCP over long-distance dialup lines was the
dominant means of article transmission, a few well-connected sites had
real influence in determining which newsgroups would be carried where.
Those sites called themselves "the backbone."
But things have changed. Nowadays, even the smallest Internet site
has connectivity the likes of which the backbone admin of yesteryear
could only dream. In addition, in the U.S., the advent of cheaper
long-distance calls and high-speed modems has made long-distance
Usenet feeds thinkable for smaller companies. There is only one
pre-eminent UUCP transport site today in the U.S., namely UUNET. But
UUNET isn't a player in the propagation wars, because it never refuses
any traffic -- it gets paid by the minute, after all; and besides, to
refuse based on content would jeopardize its legal status as an
enhanced service provider.
All of the above applies to the U.S. In Europe, different cost
structures favored the creation of strictly controlled hierarchical
organizations with central registries. This is all very unlike the
traditional mode of U.S. sites (pick a name, get the software, get a
feed, you're on). Europe's "benign monopolies," long uncontested, now
face competition from looser organizations patterned after the U.S.
model.
NEWSGROUP CREATION
------------------
As discussed above, Usenet is not a democracy. Nevertheless, the
current most popular way to create a new newsgroup involves a "vote"
to determine popular support for (and opposition to) a proposed
newsgroup. The document that describes this procedure is entitled
"How To Create A New Newsgroups." Its common name, however, is "the
guidelines."
If you follow the guidelines, it is probable that your group will be
created and will be widely propagated.
HOWEVER: Because of the nature of Usenet, there is no way for any user
to enforce the results of a newsgroup vote (or any other decision, for
that matter). Therefore, for your new newsgroup to be propagated
widely, you must not only follow the letter of the guidelines; you
must also follow its spirit. And you must not allow even a whiff of
shady dealings or dirty tricks to mar the vote.
So, you may ask: How is a new user supposed to know anything about the
"spirit" of the guidelines? Obviously, he can't. This fact leads
inexorably to the following recommendation:
>> If you are a new user, don't try to create a new newsgroup. <<
If you have a good newsgroup idea, then read the "news.groups"
newsgroup for a while (six months, at least) to find out how things
work. If you're too impatient to wait six months, then you really
need to learn; read "news.groups" for a year instead. If you just
can't wait, find a Usenet old hand to run the vote for you.
Readers may think this advice unnecessarily strict. Ignore it at your
peril. It is embarassing to speak before learning. It is foolish to
jump into a society you don't understand with your mouth open. And it
is futile to try to force your will on people who can tune you out
with the press of a key.
IF YOU ARE UNHAPPY...
---------------------
Property rights being what they are, there is no higher authority on
Usenet than the people who own the machines on which Usenet traffic is
carried. If the owner of the machine you use says, "We will not carry
alt.sex on this machine," and you are not happy with that order, you
have no Usenet recourse. What can we outsiders do, after all?
That doesn't mean you are without options. Depending on the nature of
your site, you may have some internal political recourse. Or you
might find external pressure helpful. Or, with a minimal investment,
you can get a feed of your own from somewhere else. Computers capable
of taking Usenet feeds are down in the $500 range now, and
UNIX-capable boxes are going for under $2000, and there are at least
two UNIX lookalikes in the $100 price range.
No matter what, though, appealing to "Usenet" won't help. Even if
those you read such an appeal are sympathetic to your cause, they will
almost certainly have even less influence at your site than you do.
By the same token, if you don't like what some user at another site is
doing, only the administrator and/or owner of that site have any
authority to do anything about it. Persuade them that the user in
question is a problem for them, and they might do something (if they
feel like it).
If the user in question is the administrator or owner of the site from
which he or she posts, forget it; you can't win. Arrange for your
newsreading software to ignore articles from him or her if you can,
and chalk one up to experience.
WORDS TO LIVE BY #1:
USENET AS SOCIETY
--------------------
Those who have never tried electronic communication may not be aware
of what a "social skill" really is. One social skill that must be
learned, is that other people have points of view that are not only
different, but *threatening*, to your own. In turn, your opinions may
be threatening to others. There is nothing wrong with this. Your
beliefs need not be hidden behind a facade, as happens with
face-to-face conversation. Not everybody in the world is a bosom
buddy, but you can still have a meaningful conversation with them.
The person who cannot do this lacks in social skills.
-- Nick Szabo
WORDS TO LIVE BY #2:
USENET AS ANARCHY
--------------------
Anarchy means having to put up with things that really piss you off.
-- Unknown
--
Gene Spafford
NSF/Purdue/U of Florida Software Engineering Research Center,
Dept. of Computer Sciences, Purdue University, W. Lafayette IN 47907-1398
Internet: s...@cs.purdue.edu phone: (317) 494-7825
Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <1991May24.200549.16629@newshost.anu.edu.au>
Date: 24 May 91 20:05:49 GMT
References: <14692@ector.cs.purdue.edu>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 332
In article <14...@ector.cs.purdue.edu> s...@cs.purdue.EDU
(Gene Spafford) writes:
>Original from: c...@count.tct.com (Chip Salzenberg)
>[Most recent change: 19 May 1991 by s...@cs.purdue.edu (Gene Spafford)]
I find that this essay includes a good deal of wisdom but also has
what seem to me some quite glaring omissions in the light of recent
events, despite having been updated so recently. I would appreciate
any light that others with more experience of Usenet can shed on the
issues raised below.
>WHAT USENET IS NOT
>------------------
> 1. Usenet is not an organization.
>
> Usenet has no central authority. In fact, it has no central
> anything.
Not having a central authority seems to me to imply that Usenet is
decentralized, not that it isn't an organization. Most organizations
have central authorities because of the great benefits of doing so,
but the two concepts are by no means synonymous. Quite plainly there
is a GREAT DEAL of organization involved in keeping the news flowing.
It would be surprising if this was achieved without any central
authority, though not impossible.
I always get suspicious when somebody says, authoritatively, that
"there are no authorities here". My suspicion is that there is
indeed an authority but it does not welcome scrutiny.
> 2. Usenet is not a democracy.
>
> A democracy can be loosely defined as "government of the people,
> by the people, for the people." However, as explained above,
> Usenet is not an organization, and only an organization can be run
> as a democracy. Even a democracy must be organized, for if it
> lacks a means of enforcing the peoples' wishes, then it may as
> well not exist.
>
> Some people wish that Usenet were a democracy. Many people
> pretend that it is. Both groups are sadly deluded.
I have heard MANY arguments about why people who "wish" that some
organization were more democratically organized are "sadly deluded",
but that is about the weakest.
Even Governments generally have quite limited central authority
and no matter how democratically organized they may be, they are
unable to enforce their wishes on matters considered outside the
sphere of Government authority.
Private organizations are called democratic when they have a membership
which has a right to determine the leadership and policies of the
organization by some kind of majority voting system. Many other forms
of organization exist, such as the Catholic Church and Usenet,
but democratic forms of organization do not imply a means of enforcing
the wishes of the membership over anything other than the organization
itself.
For example if Usenet was a fully democratic organization
that would NOT imply any kind of authority
to enforce the decisions of the majority of members as to what
members should do with computers they own, whether they should
subscribe to certain newsgroups and so forth, because the organization
itself has no such authority any more than a democratically elected
Government can decide what you should eat for breakfast.
It can be argued that Usenet should be run by site sysadmins
or by a small group of "backbone" sysadmins or by a
smaller group of exceptionally wise and experienced "netgods"
rather than by all users. The Catholic Church has got on quite
well without any concessions to democracy. But simply saying
that advocates of democracy are "sadly deluded" with arguments
that would not convince a schoolchild in any other context, only
highlights the inadequcy of such views.
>WHAT USENET IS
>--------------
>Usenet is the set of machines that exchange articles tagged with one
>or more universally-recognized labels, called "newsgroups" (or
>"groups" for short).
Well, if it is a set of machines then of course it cannot be
an organization, and certainly cannot become a democratic
organization, since such things are comprised of sets of people,
not machines. I think this statement is clearly inaccurate -
when I see messages that say they are asking "the net" a
question or telling "the net" something I am sure they are
addressing a group of people. Describing Usenet as a set of
machines obscures the fact that there IS a definable (though
unregistered) group of people who can be said to be "members"
of "the net" (and who are a much larger group than the site
administrators).
That leaves something like this:
"Usenet is an organization which facilitates the exchange of
articles tagged with one or more universally-recognized labels,
called "newsgroups"..."
But HOW does it organize this? Not by providing any telecommunications
equipment, or computer equipment or software or even software standards,
let alone article content etc.
The GLARING omission from the essay that I see, is that it takes the
"universally-recognized labels" for granted instead of explaining how
they come to be "universally-recognized".
I would be more inclined to describe Usenet as the organization which
defines a certain set of labels used to tag articles.
After all, the essay points out that there are other organizations
which do this, distinguishing their separate exchange of articles
from those that are involved in Usenet. As well as the alt hierarchy
(which has even less of a central authority and no more democracy)
there are also name spaces organized by clarinet as a business or
emanating from FidoNet echos and so forth.
The regular lists of currently active newsgroups sharply distinguish
between the lists of comp soc rec misc news and sci which are "official"
and those provided for some other namespaces, which are "unofficial".
It is hard to square the concept of an "official" list of "Usenet"
newsgroups and moderators, distinguished from "non-Usenet"
newsgroups, with the claim that there is no Usenet organization
or that whatever Usenet organization there is has no central
authority. Who else makes the lists "official"?
>NEWSGROUP CREATION
>------------------
>As discussed above, Usenet is not a democracy. Nevertheless, the
>current most popular way to create a new newsgroup involves a "vote"
>to determine popular support for (and opposition to) a proposed
>newsgroup. The document that describes this procedure is entitled
>"How To Create A New Newsgroups." Its common name, however, is "the
>guidelines."
>
>If you follow the guidelines, it is probable that your group will be
>created and will be widely propagated.
>
>HOWEVER: Because of the nature of Usenet, there is no way for any user
>to enforce the results of a newsgroup vote (or any other decision, for
>that matter). Therefore, for your new newsgroup to be propagated
>widely, you must not only follow the letter of the guidelines; you
>must also follow its spirit. And you must not allow even a whiff of
>shady dealings or dirty tricks to mar the vote.
Here I see another GLARING omission.
There is simply no mention at all of the "official" lists of active
newsgroups and their moderators or the corresponding checkgroups
messages.
Yet unless I have misunderstood the setup COMPLETELY (which no doubt
some kind soul will point out with enthusiasm :-) those do play
a sufficiently important part in the organization of Usenet that
it would be difficult to avoid mentioning them by accident.
Specifically, my understanding is that there are large numbers
of system administrators who treat the "official" lists as in
some way authoritative.
Many sites simply take a "full feed" of the official list while far
fewer take a "fuller feed" that includes all the available alt
groups etc. The "full feed" sites include most of the more important ones
that ensure wide propagation. Many other sites use the "official"
lists as a basis for selection, excluding entire hierarchies or
individual groups on the basis of the descriptions provided, and
adding other groups independently if they happen to know of them
and are interested.
None of this implies that any site participating in Usenet is
obliged to take a full feed or to pay any attention to the "official"
lists at all. But I would suggest that providing such "official"
lists is very central to what Usenet as an organization DOES.
Ignoring that and describing the propagation of newsgroups without
mentioning it is quite misleading.
So what?
Well, in the recent m.a.g. flamewars there were a lot of announcements
about how this or that site wasn't going to carry the group and a
lot of lectures about how nobody could impose their will on sysadmins
through a vote (together with some gratuitous advocacy of feudalism
as a model for Usenet).
A lot of that could have been avoided if it was understood that the
issue was whether m.a.g. would appear in the "official" lists,
not whether any particular site would therefore take it.
The advice that new users should wait at least six months before
attempting to start a news group strikes me as very sound in the
light of m.a.g. events.
But it would also have been helpful to provide a better description
of what Usenet is and the role of the "official" lists.
Of course it can be argued that the lists are not really official,
since Usenet is not an organization and has no central authority,
just as the essay I am responding to, and the guidelines mentioned
in it are just the opinion of their authors.
However I think that just evades reality. If an alternative list
was published, claiming to be "official" and listing different
groups or different moderators and topics for the same labels,
that would be widely seen as a disruption of Usenet whereas
adding another alt group would not be regarded the same way.
Participants in Usenet would not just say "oh well, that one
is spaf's list and shows who he thinks should be moderators
of certain groups and the other one is so-and-so's list".
Moderators on spaf's list would be quite pissed off about
mail for them being diverted to some usurper.
In short, Usenet is an organization, and it has a central
authority, the main function of which is to publish the "official" lists.
So, is that organization democratic?
Well, as the essay notes, there is a voting procedure, which
decides not merely who leads the organization and takes decisions,
but what the decisions are to be on each issue as it arises.
That is more than you get in most democratic organizations. On the
other hand the people who implement the results of the votes and
give them an "official" seal on behalf of the organization are
not democratically elected and neither are the people who write
the guidelines according to which those votes are conducted
(and neither are the returning officers under those guidelines,
but that is a separate issue).
I would say the Usenet organization meets most of the criteria
for being democratic, but has an unelected leadership of
"authoritative" figures who volunteer or are "universally
recognized" as a result of their wisdom and experience
rather than being elected. That is not uncommon, especially
with informal organizations.
It is quite an achievement, and required a long period of
evolution, to establish guidelines and "official" lists
that are so widely respected by sysadmins. It would be
pointless implementing procedures that resulted in lists
which were NOT widely respected by sysadmins. That can be
the basis of arguments for and against broader democracy
and even for or against particular controversial newsgroups.
But ignoring the whole issue under a facade of "there are
no authorities here" merely denigrates what has actually
been achieved by those authorities.
Certainly ANYBODY can establish an alt.group, which is
VERY democratic, but not nearly as useful as establishing
a Usenet group, which gets far wider propagation as a result
of the efforts of the central authorities that run Usenet.
Some people say that Usenet has "netgods" but I don't think
it is a glaring omission for the essay not to mention them. The
term is generally used tongue in cheek to refer to people
whose wisdom and experience makes their views authoritative.
Although I have seen messages which actually call for a decision
by netgods in the most grotesquely obsequious way, that is
a reflection on the individual authors of those messages and
does not require refutation in the regular posting on "What is Usenet".
The far more widespread misunderstanding of the propagation of
newsgroups and the role of the "official" lists in that does
make THAT omission a "glaring" one.
I think the essay should explain that the "official" lists and
guidelines for voting are published by individual volunteers
because there has not yet been any need for Usenet to establish
a formal constitution that specifies such matters in the way
that less informal democratic organizations do.
If that explanation is not correct, then some other explanation
should be provided. But I suspect the complete omission of any
mention of the "official lists" or of how the guidelines are
decided on reflects some kind of quite
unnecessary embarassment over the lack of "democracy", which
also could explain the ridiculous arguments about why
democracy would be inappropriate.
So, my conclusion is that Usenet IS a partial democracy with an
unelected leadership and an uncodified constitution.
Britain is a democracy with an uncodified constitution and
it works quite well. So does Usenet. The British head of state
is unelected and the head of government is only indirectly
elected. That does not make it any less of a democracy. Since
the day-to-day decision making in Usenet is based on votes
open to all members, the lack of democracy in electing
leaders is unimportant and would only become important in
some kind of "constitutional crisis".
People like Gene Spafford and David Lawrence are not Gods or even
absolute monarchs but more like equivalents of Bagehot and Hansard.
(An authoritative commentator on the British Constitution,
and an authoritative Parliamentary reporter.)
That is VERY different from the "feudalism" which some
people advocate, let alone from the absolute monarchy which
would be implied if the volunteers providing a public service
by issuing "official" lists and essays on the constitution
actually had the authority to overide votes of the members.
Denying that Usenet is an organization and has a central authority
is simply denying reality. That does not make reality go away but
contributes to the "what is an elephant" flamewars. The Usenet central
authority has no powers over individual sites and does little
but issue official lists of Usenet groups and moderators. It
would be a dangerous illusion to imagine it has authority over
more than that and an equally dangerous illusion to imagine that
just because it consists of self-appointed (or universally
recognized) volunteers that therefore Usenet is not a democratic
organization and it's central authorities could ignore the
wishes of the Usenet membership as expressed through votes.
It IS a dangerous illusion to pretend that there is no central
authority issuing official lists of newsgroups and their moderators
and future issues of "What is Usenet" should correct that illusion.
Even if there WERE "netgods" they would be "constitutional netgods"
that existed only while believed in. If people did not regard Elizabeth
Windsor as Queen of Australia then she would not be. That does not change
the fact that she is, nor does it change the fact that if she forgot that
in Australia the people and not the Queen have sovereignty she would
very rapidly cease to be Queen. (Some time ago a
British monarch lost his head for such an oversight and there have
been few problems from "royals" since then. :-)
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!fauern!ira.uka.de!sol.ctr.columbia.edu!lll-winken!elroy.jpl.nasa.gov!swrinde!zaphod.mps.ohio-state.edu!rpi!clarkson!grape.ecs.clarkson.edu!nelson
From: nel...@sun.soe.clarkson.edu (Russ Nelson)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <NELSON.91May25231909@sun.clarkson.edu>
Date: 26 May 91 04:19:09 GMT
References: <14692@ector.cs.purdue.edu>
<1991May24.200549.16629@newshost.anu.edu.au>
Sender: use...@grape.ecs.clarkson.edu
Reply-To: nel...@clutx.clarkson.edu (aka NEL...@CLUTX.BITNET)
Organization: Clarkson University, Potsdam NY
Lines: 14
In-Reply-To: cmf851@anu.oz.au's message of 24 May 91 20:05:49 GMT
In article <1991May24.200549.16...@newshost.anu.edu.au> cmf...@anu.oz.au (Albert Langer) writes:
So, my conclusion is that Usenet IS a partial democracy with an
unelected leadership and an uncodified constitution.
A dog is a partial rabbit with a carnivore diet and shorter ears.
In other words, Usenet is an anarchy, and I don't want no chocolate dog
for Easter.
--
--russ <nel...@clutx.clarkson.edu> I'm proud to be a humble Quaker.
Clear cutting is criminal, spiking trees is criminal, and using hyperbole of
this magnitude in a serious discussion is criminal. -- Irv Chidsey
Path: gmdzi!unido!math.fu-berlin.de!ox.com!emory!samsung!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <1991May26.134920.4757@newshost.anu.edu.au>
Date: 26 May 91 13:49:20 GMT
References: <14692@ector.cs.purdue.edu> <1991May24.200549.16629@newshost.anu.edu.au> <NELSON.91May25231909@sun.clarkson.edu>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 19
In article <NELSON.91May25231...@sun.clarkson.edu> nel...@clutx.clarkson.edu
(aka NEL...@CLUTX.BITNET) writes:
>A dog is a partial rabbit with a carnivore diet and shorter ears.
>
>In other words, Usenet is an anarchy, and I don't want no chocolate dog
>for Easter.
Uh... you mean Usenet is a partial anarchy... like an anarchy with a
central authority that issues "official" lists of Usenet newsgroups
based on democratic voting?
Whaddya mean p-partial wabbit? What's up doc?
I don't want no chocolate in my ears.
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!math.fu-berlin.de!ox.com!emory!swrinde!zaphod.mps.ohio-state.edu!rpi!clarkson!grape.ecs.clarkson.edu!nelson
From: nel...@sun.soe.clarkson.edu (Russ Nelson)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <NELSON.91May26121435@sun.clarkson.edu>
Date: 26 May 91 17:14:35 GMT
References: <14692@ector.cs.purdue.edu>
<1991May24.200549.16629@newshost.anu.edu.au>
<NELSON.91May25231909@sun.clarkson.edu>
<1991May26.134920.4757@newshost.anu.edu.au>
Sender: use...@grape.ecs.clarkson.edu
Reply-To: nel...@clutx.clarkson.edu (aka NEL...@CLUTX.BITNET)
Organization: Clarkson University, Potsdam NY
Lines: 36
In-Reply-To: cmf851@anu.oz.au's message of 26 May 91 13:49:20 GMT
In article <1991May26.134920.4...@newshost.anu.edu.au> cmf...@anu.oz.au (Albert Langer) writes:
In article <NELSON.91May25231...@sun.clarkson.edu> nel...@clutx.clarkson.edu
(aka NEL...@CLUTX.BITNET) writes:
>A dog is a partial rabbit with a carnivore diet and shorter ears.
>
>In other words, Usenet is an anarchy, and I don't want no chocolate dog
>for Easter.
Uh... you mean Usenet is a partial anarchy... like an anarchy with a
central authority that issues "official" lists of Usenet newsgroups
based on democratic voting?
The central authority only has the authority we grant it.
The official lists are only as official as we believe them to be.
The democratic votes are merely advisory.
Usenet is an anarchy. That doesn't mean that there aren't authorities
(centralized or otherwise), nor does it mean that there aren't rules,
nor does it mean that there is no organization.
Power *always* flows from obedience. Hitler's evil came from his power to
convince people to follow him. Otherwise he would have been another
nut. The IRS (US govt's taxing authority) does not have the power to
compel us to pay taxes. The tax system only works because nearly
everyone pays taxes voluntarily.
Unlike other systems of governance, obedience in an anarchy is acknowledged
to be voluntary. Other systems of governance pretend that obedience can
be compelled.
--
--russ <nel...@clutx.clarkson.edu> I'm proud to be a humble Quaker.
Clear cutting is criminal, spiking trees is criminal, and using hyperbole of
this magnitude in a serious discussion is criminal. -- Irv Chidsey
Path: gmdzi!unido!math.fu-berlin.de!ox.com!yale!think.com!zaphod.mps.ohio-state.edu!sample.eng.ohio-state.edu!purdue!spaf
From: s...@cs.purdue.EDU (Gene Spafford)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <14819@ector.cs.purdue.edu>
Date: 27 May 91 02:57:31 GMT
References: <14692@ector.cs.purdue.edu>
<1991May24.200549.16629@newshost.anu.edu.au>
Sender: n...@cs.purdue.EDU
Organization: Department of Computer Science, Purdue University
Lines: 102
Well, I've read most of the posts on this topic over the past few
days.
A few comments:
1) I've been posting the list of lists and informational postings for
almost 10 years now. It was only around the time of the grand
renaming of groups and the short-lived "backbone cabal" that I had any
slight delusion that the postings were in any sense "official."
Every six months or so, I post a reminder to people that the postings
are advisory. I list things that appear appropriate to me, as derived
from things other people send me and based on what I observe. Thus,
if someone sends me a description of a new hierarchy of groups, I list
it. If I see a new group created that looks like the majority of
sites might be likely to carry it (e.g., not patently silly or
illegal), I include it in my list.
Let me stress again that I consider these things as advisory. I treat
the "votes" that occur for new groups as advisory as well, as it seems
to be a good clue about whether a group will be widely carried. I
never complain about any site that carries more or less than what I
have in my lists. I do not threaten, coerce, plead, or lecture others
about the contents of the postings. I list mailing lists and
newsgroups no matter what their content, as long as they are clearly
intended seriously and are not being established primarily for illegal
activities. That is, perhaps, one reason why the lists have endured:
I am largely successful in keeping my personal agendas well-separate
of the postings.
Some people seem to find the postings helpful. I'm glad if that is
the case. It takes a fair amount of effort to keep the postings as
up-to-date as they are, and I don't always have the time.
2) I don't pretend that I will do these lists forever, or that they
are any kind of basis for future growth. Someday, when the load gets
unbearable, or the percentage of crap on the net gets still higher
than it is now, I will simply stop posting the articles. Some people
think I should have stopped long ago, perhaps because my view of the
world doesn't match theirs. Others are in some sense jealous because
fewer people heed their views of how the net should be organized than
who adhere to the view embodied in the postings. Either way, I
understand that not everyone appreciates the work or the contents of
the postings.
A combination of extra workload and pneumonia kept me from posting
updates between January and earlier this month. It didn't seem a lot
of people noticed. I'm not sure what that means, exactly, but maybe
the postings aren't serving the same purpose now as they were a few
years ago. A few years ago, when I made noises about quitting, I got
lots and lots of main asking me not to. Would it happen now?
I dunno....
In general, I suspect that as the net evolves, and especially as the
lack of simple consideration becomes more prevalent on net postings,
reminders on etiquette and proper places to post will become
anachronisms (if they haven't already).
3) I really don't like the term "net.god" especially when
someone refers to me that way. If people respect the tenacity,
effort, moderation, or dedication -- great. I don't maintain the
postings for any sense of "power" or to get "fans" -- I do it because
I believe it helps others get more out of the Usenet, and makes it a
better place for me to exchange information. Besides, I haven't yet
seen any Usenet groupies!
4) Usenet is as much an organization as is the group of people who
visit the park on a Sunday afternoon. The composition changes from
day to day, no one takes formal attendance, and everyone is doing
pretty much what they fancy; luckily, most of the people are staying
on the paths and not pissing on the flowers.
5) If you don't agree with me or my postings, congratulations. If you
do agree, great. If you don't care, join the rest. If you have
suggestions or corrections, mail them to me and I'll consider them for
the next batch. I'll also gladly accept any new articles to include
in the group if they seem informative and non-redundant.
In case you haven't noticed, the following people wrote the original
articles I maintain and dozens more have made small contributions:
Randall Atkinson <rand...@uvaarpa.virginia.edu>
a...@uunet.uu.net (Andrew Partan)
b...@looking.on.ca (Brad Templeton)
c...@apple.COM (Chuq Von Rospach)
hoptoad!gnu (John Gilmore)
je...@eagle.UUCP (Jerry Schwarz)
m...@stargate.com (Mark Horton)
o...@gatech.edu (A. Jeff Offutt VI)
r...@hazel.circ.upenn.edu (Rich Kulawiec)
tay...@intuitive.com (Dave Taylor)
wo...@ncar.ucar.edu (Greg Woods)
I credit each and every one of them every time I post their works. I
have never pretended to be the font of wisdom about the net. A source
of experience, perhaps; wisdom is something made clear only by the
passage of time, and not by one's own judgement.
--
Gene Spafford
NSF/Purdue/U of Florida Software Engineering Research Center,
Dept. of Computer Sciences, Purdue University, W. Lafayette IN 47907-1398
Internet: s...@cs.purdue.edu phone: (317) 494-7825
Path: gmdzi!unido!unidui!math.fu-berlin.de!ox.com!caen!sdd.hp.com!think.com!rpi!batcomputer!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <1991May30.020747.19060@newshost.anu.edu.au>
Date: 30 May 91 02:07:47 GMT
References: <NELSON.91May25231909@sun.clarkson.edu> <1991May26.134920.4757@newshost.anu.edu.au> <NELSON.91May26121435@sun.clarkson.edu>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 113
In article <NELSON.91May26121...@sun.clarkson.edu> nel...@clutx.clarkson.edu
(aka NEL...@CLUTX.BITNET) writes:
>The central authority only has the authority we grant it.
>The official lists are only as official as we believe them to be.
I won't argue with those two statements.
>The democratic votes are merely advisory.
This one however is ambiguous and it is precisely that ambiguity I am
trying to clear up. Advisory to who?
My impression is that the "official" or "standard" USENET lists
are "merely advisory" to the administrators of each independent site.
If that is what you and others mean by "anarchy" then I think that is
an inaccurate use of the term. All kinds of things are "merely advisory"
without anybody being under the illusion that the organizations
producing them are anarchistic. For example most "standards" are just
as "advisory" as the "standard" USENET list, without anybody thinking
that ANSI, IEEE standards board etc are not organizations or have
no central authorities.
However I'm not really interested in debating the correct use of the
term "anarchy". My concern is that there is another possible
interpretation of "merely advisory" and that the "What is Usenet"
essay obscures rather than clarifies the issue.
It could be thought that the votes regularly organized in
news.groups.announce are "merely advisory" to anyone compiling a
list of "official" ("traditional" or "standard" newsgroups) or are
"just an opinion poll" (as stated recently by Chip Salzenberg though
not expressed so openly in his essay on "What is Usenet").
If that was so, it would not establish that USENET is an anarchy or
has no organization, but it would certainly establish that
whatever organization it has, is not democratic.
I believe that understanding of the nature of USENET would be as
misleading as to claim that the decisions of Parliament are
"merely advisory" to a constitutional monarch. As I pointed out,
one British monarch lost his head for making that mistake, and the
decision to remove it from his shoulders was carried out, despite
that particular Act of Parliament never having received the Royal
Assent.
>Usenet is an anarchy. That doesn't mean that there aren't authorities
>(centralized or otherwise), nor does it mean that there aren't rules,
>nor does it mean that there is no organization.
I agree that an anarchy can have an organization, and even "rules"
(though in a more limited sense than is often understood by that
term). But your contention that an anarchy can have centralized
authorities indicates you are using the term in a quite different
sense from any that I am familiar with. Again, I see no real
point in debating the semantics. Call it what you like - if we are
agreed that USENET has a centralized authority which is bound by the
votes of members when issuing lists of "standard" Usenet newsgroups,
but that has no authority over members or anybody else,
then you may call it an anarchy or even a cheesecake for all I
care. The point under discussion is that others would disagree
with that description of the characteristics of this particular
cheesecake, not whether an entity with those characteristics
should or should not be called a cheesecake.
>Power *always* flows from obedience. Hitler's evil came from his power to
>convince people to follow him. Otherwise he would have been another
>nut. The IRS (US govt's taxing authority) does not have the power to
>compel us to pay taxes. The tax system only works because nearly
>everyone pays taxes voluntarily.
The IRS does have power to compel individuals to pay taxes. Since many
people do NOT pay taxes "voluntarily" but only because of the legal
compulsion to do so (whereas they may for example donate to charities
"voluntarily"), the IRS is an excellent example of how power does NOT
always flow from voluntary conviction.
Like most (though not all) private organizations, and unlike
Government agencies such as the IRS, USENET has no means to compel
obedience. That does not mean it is an anarchy any more than any
other voluntary association is an anarchy.
>Unlike other systems of governance, obedience in an anarchy is acknowledged
>to be voluntary. Other systems of governance pretend that obedience can
>be compelled.
Unlike other systems of governance, anarchy is imaginary. But we are
not discussing systems of governance. A more conventional definition
of anarchy asserts that social affairs can be administered through
voluntary associations without any need for the compulsion of State
Power. My point is that USENET is an example of such an (informal)
voluntary association with no power to coerce. If USENET Governed
some country then that country would indeed be an anarchy.
There are many thousands of voluntary associations
with no power to coerce organized for one purpose or another, but none
of them are organized for the purpose of governing countries so
they are not examples of "anarchy". I have never heard of voluntary
associations being confused with "anarchies" before and I suspect you
must be suffering from the fantasy that "netland" is a country.
BTW Abraham Lincoln took the position that when the people grow
weary of their Government they have TWO rights - their constitutional
right to reform it and their revolutionary right to dismember and
overthrow it. That is consistent with your view that Government compulsion
of individuals only works when the people voluntarily accept the power
of Government. That view is generally described as (revolutionary)
democracy, not anarchy.
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!mcsun!uunet!pdn!tscs!tct!chip
From: c...@tct.com (Chip Salzenberg)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <28451CB8.1085@tct.com>
Date: 30 May 91 15:39:36 GMT
References: <1991May26.134920.4757@newshost.anu.edu.au> <NELSON.91May26121435@sun.clarkson.edu> <1991May30.020747.19060@newshost.anu.edu.au>
Organization: Teltronics/TCT, Sarasota, FL
Lines: 85
According to cmf...@anu.oz.au (Albert Langer):
>My impression is that the "official" or "standard" USENET lists
>are "merely advisory" to the administrators of each independent site.
Agreed. (With the reservation that I consider the words "official" and
"standard" inaccurate.)
>All kinds of things are "merely advisory" without anybody being under
>the illusion that the organizations producing them are anarchistic.
Usenet is an anarchy because *everything* is only advisory. Newsgroup
votes are only the most visible example of this fundamental truth.
>My concern is that there is another possible interpretation of "merely
>advisory" and that the "What is Usenet" essay obscures rather than
>clarifies the issue.
In my opinion, the "What Is Usenet?" article is excruciatingly clear
on the subject of authority (i.e. that there is none). I would like
Mr. Langer to stop claiming that the article is "obscuring" the issues
just because it disagrees with his version of the facts.
>It could be thought that the votes regularly organized in
>news.groups.announce ...
Votes are not organized in news.announce.newgroups. In the usual case,
they are announced there.
>"just an opinion poll" (as stated recently by Chip Salzenberg though
>not expressed so openly in his essay on "What is Usenet")
Granted, those exact words do not appear. Consider, however, these
excerpts:
As discussed above, Usenet is not a democracy. Nevertheless, the
currently most popular way to create a new newsgroup involves a vote
to determine popular support for (and opposition to) a proposed
newsgroup. [...]
If you follow the guidelines, it is PROBABLE that your group will be
created and will be widely propagated. [emphasis added]
HOWEVER: Due to the nature of Usenet, there is no way for any user
to enforce the results of a newsgroup vote (or any other decision,
for that matter).
In my opinion, these quotations make clear that a newsgroup vote has
no authority, that it is only advisory. Hmm... that sounds a lot like
an opinion poll, doesn't it?
Still, this point could be made more clear. I'll probably put in a
paragraph that explicitly explains that "vote" is a misnomer and that
it is only an opinion poll.
Incidentally, if the phrase "not expressed so openly" is meant to
imply that I have held back facts intentionally, I must protest in the
strongest possible terms. I have *not* been dishonest with anyone,
nor have I held back anything I know. I have put forth the facts as I
see them. Others may disagree with my statements, but there is no
reason to impugn my honesty.
>If we are agreed that USENET has a centralized authority which is
>bound by the votes of members when issuing lists of "standard" Usenet
>newsgroups ...
Hahahahaha(h). That's a good one.
"We" are not agreed on the statement quoted above, except perhaps in
your own mind, because it is just plain false:
(1) If Usenet has a centralized authority of some kind, where is it?
Gene Spafford has already denied that he's it. If he isn't it,
then who is?
(2) If Usenet is bound by the votes of "members" (whoever they are),
then why is comp.protocols.tcp-ip.eniac not on anyone's list?
These two questions must be answered by anyone who claims that Usenet
is an organization with a central authority.
--
Chip Salzenberg at Teltronics/TCT <c...@tct.com>, <uunet!pdn!tct!chip>
perl -e 'sub do { print "extinct!\n"; } do do()'
--
Chip Salzenberg at Teltronics/TCT <c...@tct.com>, <uunet!pdn!tct!chip>
perl -e 'sub do { print "extinct!\n"; } do do()'
Path: gmdzi!unido!mcsun!uunet!spool.mu.edu!caen!ox.com!msen.com!emv
From: e...@msen.com (Ed Vielmetti)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <EMV.91May31001113@bronte.aa.ox.com>
Date: 31 May 91 04:11:20 GMT
References: <1991May26.134920.4757@newshost.anu.edu.au>
<NELSON.91May26121435@sun.clarkson.edu>
<1991May30.020747.19060@newshost.anu.edu.au> <28451CB8.1085@tct.com>
Sender: use...@ox.com (Usenet News Administrator)
Organization: MSEN, Inc. Ann Arbor MI
Lines: 46
In-Reply-To: chip@tct.com's message of 30 May 91 15:39:36 GMT
In article <28451CB8.1...@tct.com> c...@tct.com (Chip Salzenberg) writes:
(1) If Usenet has a centralized authority of some kind, where is it?
Gene Spafford has already denied that he's it. If he isn't it,
then who is?
Who makes the lists, makes the rules. Gene Spafford makes the list of
newsgroups, many people follow it, ergo he has some measure of power.
Dave Lawrence publishes the list of newly created newsgroups, he has
some measure of editorial control over its contents and the course of
newsgroup voting, he has some measure of power. Brian Reid collects
the arbitron statistics, he disseminates information about what people
are really reading, he has some measure of power.
Of these three, s...@cs.purdue.edu is probably the easiest to unseat.
He hardly reads any news these days, posting a list that doesn't
change that much is not that hard work, and t...@rpi.edu does most of
the hard work for him. There's the huge flow of data going into the
arbitrons and the Usenet Top 1000 sites polls that would be hard to
duplicate, so I guess r...@decwrl.dec.com is safe; besides, there's
very little editorial comment in the arbitron lists so it'd be hard to
rewrite.
If you want to have power, post to news.lists twice monthly. I
presume that anyone who knows how to write a perl script and run a
cron job and create a moderated message has the authority to do that.
(2) If Usenet is bound by the votes of "members" (whoever they are),
then why is comp.protocols.tcp-ip.eniac not on anyone's list?
The eniac group is not on Gene Spafford's list because it is not on
Brian Reid's list. It's not on Brian's list because no one's talking
about it. You'll need to ask a better question.
These two questions must be answered by anyone who claims that Usenet
is an organization with a central authority.
Chip, you state that with such authority that I'm afraid some people
will believe you.
--
Edward Vielmetti, MSEN Inc. moderator, comp.archives e...@msen.com
"often those with the power to appoint will be on one side of a
controversial issue and find it convenient to use their opponent's
momentary stridency as a pretext to squelch them"
Path: gmdzi!unido!mcsun!uunet!spool.mu.edu!news.cs.indiana.edu!purdue!spaf
From: s...@cs.purdue.EDU (Gene Spafford)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <14865@ector.cs.purdue.edu>
Date: 31 May 91 23:00:20 GMT
References: <1991May26.134920.4757@newshost.anu.edu.au>
<NELSON.91May26121435@sun.clarkson.edu>
<1991May30.020747.19060@newshost.anu.edu.au> <28451CB8.1085@tct.com>
<EMV.91May31001113@bronte.aa.ox.com>
Sender: n...@cs.purdue.EDU
Organization: Department of Computer Science, Purdue University
Lines: 36
In article <EMV.91May31001...@bronte.aa.ox.com> e...@msen.com (Ed Vielmetti) writes:
Of these three, s...@cs.purdue.edu is probably the easiest to unseat.
He hardly reads any news these days, posting a list that doesn't
change that much is not that hard work, and t...@rpi.edu does most of
the hard work for him.
Even easier than that. Just get enough people to tell me they would
like me to stop the postings, and I will.
I "hardly read the news" in the sense that I only read news in about
20 groups, with low volume. I have a "real life" and Usenet is not
high priority there. My wife, my committments as a university
professor and researcher, and my activities in professional societies
rank WAY up above reading people's quibbles about naming new groups,
posting chain letters, and general flaming. If that's grounds for
anything one way or another related to posting the lists, I'd be
interested in hearing the rationale.
As for the lists not changing, well, I average between 125 and 150
messages a week (as of last 6 weeks) request updates to the lists,
mailed copies, inquiries, flames, and other assorted stuff. You may
not see massive changes at any particular point, but a fair amount
happens on an on-going basis.
Dave Lawrence (t...@cs.rpi.edu) does a fair amount of work moderating
the newgroups group, but that is only a fraction of what is involved
in maintaining ALL the lists.
But you're right -- "unseating" me is very simple, and would take very
little effort. Just ask politely.
--
Gene Spafford
NSF/Purdue/U of Florida Software Engineering Research Center,
Dept. of Computer Sciences, Purdue University, W. Lafayette IN 47907-1398
Internet: s...@cs.purdue.edu phone: (317) 494-7825
Path: gmdzi!unido!mcsun!uunet!looking!brad
From: b...@looking.on.ca (Brad Templeton)
Newsgroups: news.admin
Subject: Re: What is Usenet? LONG!
Message-ID: <1991Jul05.045509.22233@looking.on.ca>
Date: 5 Jul 91 04:55:09 GMT
References: <1991Jun23.050938.29045@wolves.uucp> <1991Jun30.100732.11768@newshost.anu.edu.au> <91Jul2.215538edt.1086@smoke.cs.toronto.edu> <1991Jul4.155833.23710@newshost.anu.edu.au>
Organization: Looking Glass Software Ltd.
Lines: 14
Usenet is not a commune at all. It is a co-operative, but that is quite
different from a commune. Usenet exists within societies that are all
pretty property oriented. You don't lose your copyright rights by posting
to USENET any more than George Lucas lost the copyright on "Star Wars" when
it was broadcast to the general airwaves. Each site admin has complete
and absolute control over her local machine, with only a few exceptions
(such as not having the right to snoop at E-mail in the USA)
To get technical, USENET is a minarchist co-operative. It would be
an anarchist co-operative (not an oxymoron, if you think it is you don't
know what anarchism is) but it exists within societies that do have laws
about copyright, libel etc.
--
Brad Templeton, ClariNet Communications Corp. -- Waterloo, Ontario 519/884-7473
Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet? LONG!
Message-ID: <1991Jul6.084236.11568@newshost.anu.edu.au>
Date: 6 Jul 91 08:42:36 GMT
References: <91Jul2.215538edt.1086@smoke.cs.toronto.edu> <1991Jul4.155833.23710@newshost.anu.edu.au> <1991Jul05.045509.22233@looking.on.ca>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 149
In article <1991Jul05.045509.22...@looking.on.ca> b...@looking.on.ca
(Brad Templeton) writes:
>Usenet is not a commune at all. It is a co-operative, but that is quite
>different from a commune. Usenet exists within societies that are all
>pretty property oriented. You don't lose your copyright rights by posting
>to USENET any more than George Lucas lost the copyright on "Star Wars" when
>it was broadcast to the general airwaves. Each site admin has complete
>and absolute control over her local machine, with only a few exceptions
>(such as not having the right to snoop at E-mail in the USA)
Hmmm, I'm inclined to agree that "co-operative" is a more accurate
description of the kind of unincorporated association involved in
Usenet than "commune" (or "conspiracy" :-) though not for exactly
the same reasons.
When "Star Wars" is broadcast, the retention of copyright does not
establish that the broadcast stations either are or are not cooperatives
or communes. One could imagine communal broadcasters, cooperative
broadcasters or, as is actually the case, private and Government
corporations.
The EXISTENCE of a copyright in "Star Wars" depends on the fact that
this collectively produced work was created as a "work for hire" by
employees and/or by contractors of the copyright owner who assigned
their rights. It was not produced by the broadcasters and in no way
does it define what kinds of organization the broadcasters are. Again,
"Star Wars" could have been produced by a coop or a commune instead of by a
corporation, but it wasn't.
Does a Jazz musician have copyright in the music he or she
contributes during a "jam" session? Or is the only resulting musical "work",
one that was produced collectively by all participating? Or do both the
individual works and the collective work exist side by side?
Can one claim copyrignt in the "public performance" of one's own
part in a conversation? Or in a Usenet "thread"?
Some individual articles or collections of articles broadcast through
Usenet may be the intellectual property of specific copyright owners
who licence such broadcast.
But Usenet is not just a broadcaster or cooperative of broadcasters.
It is also (and more importantly) the people who exchange the articles
broadcast and who produce collective works while doing so, but are
not employees of, and have not assigned rights to, any copyright
owner.
While a moderator might plausibly claim a "compilation copyright"
for a moderated news group, no such claim has ever been tested legally.
For unmoderated Usenet newsgroups it seems clear that the resulting
work is a collective product and I suspect the same might be true
for most moderated news groups.
Don't forget that as well as explicitly broadcasting any article
distributed to any news group, one is also inviting others to reply to it,
to quote from it (even in full) and to further develop it as
part of the collective work being produced in the thread, news group
or group of news groups concerned. (When one is NOT joining in the
collective work this way, but merely making use of Usenet for distribution,
that tends to be resented as an abuse of Usenet by other participants -
as with "commercial" material).
Conventional concepts of intellectual property ownership do not fit
in well with what is essentially a collective or communal creative
process. The usual way collective work is organized in this soceity
based on private property is as work for hire, with the employer
owning the collectively produced work. That model clearly does NOT
apply to Usenet, since those contributing work for different employees.
One should not overlook the distinctively "communal" rather than
merely "cooperative" aspects of this. There is no concept of an
"exchange" between people contributing their endeavours to Usenet
in return for which they receive benefits from Usenet, whether
in proportion to their contributions or otherwise. Instead we have
the communistic principle "from each according to their ability",
for contributing to the net, and "to each according to their
need" for distributing the results of those contributions.
That is not normally the basis for any mere "cooperative" (where distribution
according to work, shares in the co-op or various other private
property based principles prevails).
This is especially striking when one looks at the original and still
major focus of Usenet, for exchange of technical information among
computer specialists. While we are continually assured that the
computer industry and especially software development is a classic
illustration of the superiority of capitalist entrepreneurial
organization, a glance behind the scenes shows a very different
picture.
The whole industry is quite dependent on a free exchange of
technical assistance among employees of ferociously "competing"
corporations, organized by those employees themselves, often
behind the backs of their managers and without the assistance
of any traditional "corporate" structure.
Some of the best software is developed without pay, as an
entirely voluntary contribution "from each according to their
ability" and is then distributed globally and instantaneously,
far more efficiently than has ever been achieved by conventional
wholesalers and retailers, "to each according to their need".
Software needs that cannot be met by spare time endeavours, are produced
using grants from military and Government sources,
in military, Government and academic organizations, again
as "freely available" public goods, not as commodities for sale.
This has a lot of problems and difficulties, and it can't extend
to the rest of the economy and especially those areas that don't
produce "public goods", without major upheavals. But isn't it
interesting that the most high-tech, dynamic sector in modern
capitalist society is also so communistic?
By way of contrast, all the "suits" have been able to achieve,
with their copyright and patent laws and marketing schemes, is
to hinder the collective production of software with trade
secrecy etc for source code, and hinder collective distribution
with shrink wrap.
>To get technical, USENET is a minarchist co-operative. It would be
>an anarchist co-operative (not an oxymoron, if you think it is you don't
>know what anarchism is) but it exists within societies that do have laws
>about copyright, libel etc.
I don't understand the term "minarchist" - does it mean something
like "minimally anarchist"? Anyway, I don't see that "anarchist" is
any more accurate than "feudalist" as an addition to "cooperative"
or even just to "association".
I think Usenet is only a co-operative, rather than a commune, and only a low
level cooperative at that, because it lacks any collective ownership and
control of common resources that can deploy those resources according
to a common plan. More like a simple "mutual-aid team" where peasants
exchange cattle and farm implements and stocks of grain and labor for
mutual benefit while cultivating their OWN land.
Now imagine what Usenet could achieve if it DID have some collectively
owned resources and an organization that could work to a common plan.
It's managed a lot better than the PTTs at organizing an email
and news service WITHOUT that (though with assistance from both open and
"back door" diversion or military, government, academic and corporate
resources). Perhaps it could do a lot more with more organization
(or perhaps it could get stuffed up by bureaucracy - both are possible).
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!math.fu-berlin.de!ox.com!yale.edu!qt.cs.utexas.edu!zaphod.mps.ohio-state.edu!samsung!uunet!alembic!csu
From: c...@alembic.acs.com (Dave Mack)
Newsgroups: news.admin
Subject: Re: What is Usenet? LONG!
Message-ID: <1991Jul6.214904.18944@alembic.acs.com>
Date: 6 Jul 91 21:49:04 GMT
References: <1991Jul4.155833.23710@newshost.anu.edu.au> <1991Jul05.045509.22233@looking.on.ca> <1991Jul6.084236.11568@newshost.anu.edu.au>
Organization: Alembic Computer Services, McLean VA
Lines: 21
In article <1991Jul6.084236.11...@newshost.anu.edu.au> cmf...@anu.oz.au (Albert Langer) writes:
>Now imagine what Usenet could achieve if it DID have some collectively
>owned resources and an organization that could work to a common plan.
Imagine what would happen (in the United States, at least) as soon
as an actual property-owning target for litigation existed.
Imagine all the marvelous political fluff that would begin as
soon as something that thought of itself as a governing body
existed.
Imagine the flame wars over the "common plan". (Votes *will* be
taken in accordance with the "Common Plan Amendment Guidelines"
as soon as we figure out what those guidelines are.)
Imagine attempting to collect dues from all the people who
think that Usenet is free.
--
Dave Mack
"I've fallen and I can't get up. Call my lawyer."
Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet? LONG!
Message-ID: <1991Jul7.182322.26915@newshost.anu.edu.au>
Date: 7 Jul 91 18:23:22 GMT
References: <1991Jul05.045509.22233@looking.on.ca> <1991Jul6.084236.11568@newshost.anu.edu.au> <1991Jul6.214904.18944@alembic.acs.com>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 47
In article <1991Jul6.214904.18...@alembic.acs.com> c...@alembic.acs.com
(Dave Mack) writes:
>>Now imagine what Usenet could achieve if it DID have some collectively
>>owned resources and an organization that could work to a common plan.
>
>Imagine what would happen (in the United States, at least) as soon
>as an actual property-owning target for litigation existed.
>
>Imagine all the marvelous political fluff [...]
>
>Imagine the flame wars over the "common plan" [...]
>
>Imagine attempting to collect dues [...]
Gee, I think I preferred John Lennon's version of "Imagine".
The paragraph you (and Peter da Silva) started quoting from ended
with:
"Perhaps it could do a lot more with a more organization (or perhaps
it could get stuffed up by bureaucracy - both are possible)."
I really don't think either of you are showing great insight by
focussing on one possibility when attention has already been drawn to two.
It's not as though you are deepening or widening the discussion by
drawing attention to some new aspect not previously mentioned (though
the selective quoting could make it look that way).
It's more like you are proudly proclaiming a DESIRE to only look for
problems rather than also look for solutions.
If I am wrong about that, how about you actually DO some "imagining"
and suggest some possible ways to overcome the (very real) obstacles
you mention to Usenet getting more organized.
I have some suggestions, but I'd be interested in hearing yours first.
Or are you SERIOUSLY suggesting that the points you mention are
self-evidently INSUPERABLE barriers to Usenet ever achieving any
higher level of organization than it already has?
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!mcsun!uunet!zaphod.mps.ohio-state.edu!menudo.uh.edu!sugar!taronga!peter
From: pe...@taronga.hackercorp.com (Peter da Silva)
Newsgroups: news.admin
Subject: Re: What is Usenet? LONG!
Message-ID: <I1V5B79@taronga.hackercorp.com>
Date: 8 Jul 91 01:05:12 GMT
References: <1991Jul7.182322.26915@newshost.anu.edu.au>
Organization: A corner of our bedroom
Lines: 44
cmf...@anu.oz.au (Albert Langer) writes:
> I really don't think either of you are showing great insight by
> focussing on one possibility when attention has already been drawn to two.
The experience with other networks with central control... or for that
matter the experience of the old backbone cabal... is what leads us to
view any such center as an undesirable thing.
> If I am wrong about that, how about you actually DO some "imagining"
> and suggest some possible ways to overcome the (very real) obstacles
> you mention to Usenet getting more organized.
This has been rehashed any number of times. You're welcome to do some
brainstorming and I'll be happy to stand aside. But I like Usenet the way
it is. If I want a different net, I can join one. Fidonet, for example,
or Peacenet. Or Clarinet.
> I have some suggestions, but I'd be interested in hearing yours first.
Be my guest.
> Or are you SERIOUSLY suggesting that the points you mention are
> self-evidently INSUPERABLE barriers to Usenet ever achieving any
> higher level of organization than it already has?
No, I'm seriously suggesting that a network like Usenet with its lack
of organisation is a good thing. There is no reason you can't set up
a parallel network with tighter central control. It's been done already,
over an over again: look at the gnu.* hierarchy, or vmsnet, clarinet,
bionet. There are also weaker networks, such as alt.*. Regional ones,
like tx.*. Over in Europe they have EUNET with its own hierarchy.
The best way to convince people that a better organised Usenet would
be a better Usenet is to demonstrate it. Set up your own network,
starting by establishing a few co-operating nodes and build it up
from there by connecting nodes that are willing to abide by your
rules. If you thus produce a better net, you *will* supplant Usenet.
I've suggested the same thing to other people who have wanted to set
up something like this. I'll even contribute ideas if you like... but
please run the test first before you try changing Usenet itself.
--
Peter da Silva. `-_-' Taronga Park BBS +1 713 568 0480 2400/n/8/1
Taronga Park. 'U` "Have you hugged your wolf, today?"
Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet? LONG!
Message-ID: <1991Jul8.191839.8793@newshost.anu.edu.au>
Date: 8 Jul 91 19:18:39 GMT
References: <1991Jul7.182322.26915@newshost.anu.edu.au> <I1V5B79@taronga.hackercorp.com>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 151
In article <I1V5...@taronga.hackercorp.com> pe...@taronga.hackercorp.com
(Peter da Silva) quotes me and writes:
>The experience with other networks with central control... or for that
>matter the experience of the old backbone cabal... is what leads us to
>view any such center as an undesirable thing.
"Central control" and "getting more organized" are not synonymous.
Experience is something to learn from, not become frightened by.
>> If I am wrong about that, how about you actually DO some "imagining"
>> and suggest some possible ways to overcome the (very real) obstacles
>> you mention to Usenet getting more organized.
>
>This has been rehashed any number of times. You're welcome to do some
>brainstorming and I'll be happy to stand aside. But I like Usenet the way
>it is. If I want a different net, I can join one. Fidonet, for example,
>or Peacenet. Or Clarinet.
Well, there is plenty of brain-storming going on - e.g. an interesting
thread in comp.archives.admin, not to mention various things that
software developers are thinking about which will ultimately have a
major impact on the way the net functions. The problem I see is that
there is no machinery for reaching decisions after brain-storming,
so the changes that WILL occur will happen haphazardly and without
adequate preparation.
The Usenet you like "the way it is" has never been that way before
and never will be again. It is one of the fastest growing and most
rapidly changing entities around. Nothing that grows as rapidly as
Usenet can avoid periodically changing it's structures quite
fundamentally.
>No, I'm seriously suggesting that a network like Usenet with its lack
>of organisation is a good thing. There is no reason you can't set up
>a parallel network with tighter central control. It's been done already,
>over an over again: look at the gnu.* hierarchy, or vmsnet, clarinet,
>bionet. There are also weaker networks, such as alt.*. Regional ones,
>like tx.*. Over in Europe they have EUNET with its own hierarchy.
Again, "tighter central control" isn't the issue. Nor are newsgroup
hierarchies. Indeed, what makes you so certain the future Usenet will
even HAVE newsgroup hierarchies? Your thinking is constricted by
taking those for granted. I've tried to bring out the fact that
Usenet DOES have more organization than it is usually given credit
for and had hoped that this might be recognized as not being such
a bad thing.
>The best way to convince people that a better organised Usenet would
>be a better Usenet is to demonstrate it. Set up your own network,
>starting by establishing a few co-operating nodes and build it up
>from there by connecting nodes that are willing to abide by your
>rules. If you thus produce a better net, you *will* supplant Usenet.
Not necessarily. The problems that arise in a new network of a few
cooperating nodes, and solutions found for those problems, do not
necessarily scale up to a network as big as Usenet and may therefore
demonstrate nothing. Indeed that is the problem, Usenet structures
lag behind it's growth.
The reason why I am probing and challenging the way Usenet is
organized is NOT because I think I have some simple alternative set
of "rules" that could be superior to what has evolved from the
efforts of many wise and experienced people over some considerable
time. On the contrary it is my respect for what has been achieved
that leads me to want to understand it better and I happen to
believe that the best way to understand anything is by probing
and challenging it. If I didn't think there was something to learn
from here, I wouldn't bother trying to change it.
>I've suggested the same thing to other people who have wanted to set
>up something like this. I'll even contribute ideas if you like... but
>please run the test first before you try changing Usenet itself.
Fine, but I cannot possibly "try changing Usenet itself" by merely
writing some articles here and in the meantime the "test" is reaction
to ideas and proposals, which may well result in significant changes
to those ideas and proposals before trying anything in practice.
Let me spell out my (unhidden) agenda for reference in case you want
to contribute some ideas.
I want to see email and news technology widely used by the "general public"
including small community organizations in both developed and developing
countries. I want to see it becoming a new form of interactive mass
media at least as significant as the print media and accessible to anyone
that has access to a simple PC wordprocessor (which I expect to include
most office workers and most homes in developed countries within a few
years). Accessible not just as a "consumer" but also as a "producer"
so that instead of people just soaking up ideas from the mass media,
they can also discuss what to do about the world and other topics of
interest.
That involves a MASSIVELY larger scale operation than Usenet and
requires solution to many problems such as:
1. Fully automatic store and forward nodes that don't need skilled
sysadmins any more than telephones need telephone operators.
2. Broadcast and transport technologies that can cope with the
higher volume.
3. Filtering, hypertext and related technologies that can select
from the higher volume and make it accessible to people interested
in relatively low volumes.
4. Intuitive user interfaces for people that can only just handle
a wordprocessor or text editor on a Mac.
All those are EASY (well, relatively easy :-)
But the social and political problems are MUCH more interesting
and complex.
Who builds it, who pays for it, how is it organized?
As far as the "mass market" is concerned we have the experience
of IBM's "Prodigy", the French Minitel, Prestel (or Australia's
"Viatel" and now "Discovery"), CI$ and so forth, and
the Baby Bells are clearly waiting in the wings.
Is that the future of this technology? Seems to me the corporations
that COULD introduce it just don't think in terms of truly
INTERACTIVE networking. Their whole logic points to "customers"
for "services".
I don't know whether Usenet itself could actually evolve to
BECOME something that really serves the general public. There
are obvious arguments against. But I'm quite sure that it at
least has important lessons that should be studied on how to
get something going without relying on corporate structures
to do it.
P.S. Your clarification that you did not mean that Usenet consists of
Marxists and Anarchists is understood, and also that you are not one
or other of these yourself (though it was not necessary to go so far as
to suggest that Trotsky was an Anarchist in order to establish that
you are not even mildly interested in these subjects).
Other digression dropped as requested... except to remark that
neither Marx nor Trotsky were Anarchists, that neither of them were known
for advocating structures similar to Usenet and that the widespread
belief the net has some kind of anarchist structure is not unrelated
to widespread ignorance of both the structure of the net and the (various)
ideas of Anarchism (not to mention Marxism and Trotskyism).
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!math.fu-berlin.de!ox.com!caen!zaphod.mps.ohio-state.edu!wuarchive!uunet!stanford.edu!mcnc!duke!wolves!ggw
From: g...@wolves.uucp (Gregory G. Woodbury)
Newsgroups: news.admin,comp.org.eff.talk,alt.folklore.computers
Subject: Commercial Usenet? (was: What is Usenet? LONG!)
Keywords: Usenet Prodigy history Bell Laboratories anecdotes
Message-ID: <1991Jul10.030757.6007@wolves.uucp>
Date: 10 Jul 91 03:07:57 GMT
References: <1991Jul7.182322.26915@newshost.anu.edu.au> <I1V5B79@taronga.hackercorp.com> <1991Jul8.191839.8793@newshost.anu.edu.au>
Followup-To: news.admin
Organization: Wolves Den UNIX
Lines: 299
X-Checksum-Snefru: 94c4622b a010393d cedcb919 a784b34c
{Is there still a line eater around? I don't think so . . . :-}
[Later in this article, I relate some opinions on Prodigy and comment on
the development of its predecessors. This is part of the ongoing
discussion of where is "Usenet" going. - ggw]
In <1991Jul8.191839.8...@newshost.anu.edu.au> cmf...@anu.oz.au
(Albert Langer) writes:
>
>In article <I1V5...@taronga.hackercorp.com> pe...@taronga.hackercorp.com
>(Peter da Silva) quotes me and writes:
>
>>The experience with other networks with central control... or for that
>>matter the experience of the old backbone cabal... is what leads us to
>>view any such center as an undesirable thing.
>
>"Central control" and "getting more organized" are not synonymous.
>Experience is something to learn from, not become frightened by.
>
Albert, once again we are falling into the old argument about
the structure of the "current Usenet" rather than focusing on the
future. It seems more and more that you are failing to grasp certain
basic concepts that underlie the USA portions of Usenet.
I DO NOT mean to be parochial or provincal about this, but the
main disagreements I hear from you seem firmly rooted in the distinction
between the USA infrastructure and the way the rest of the world seems
to view Usenet.
I understand your point about getting more organized in terms of
preparing for the future, and the distinction you make from "central
control". It is just that in the USA, for the sites that participate in
Usenet to be able to make the changes and transitions you indicate, they
must be free from any official control.
:
>
>Well, there is plenty of brain-storming going on - e.g. an interesting
>thread in comp.archives.admin, not to mention various things that
>software developers are thinking about which will ultimately have a
>major impact on the way the net functions. The problem I see is that
>there is no machinery for reaching decisions after brain-storming,
>so the changes that WILL occur will happen haphazardly and without
>adequate preparation.
I do wish, however, that you would make up your mind! Either
there is an organization to Usenet (which you claim makes the decision
about newsgroups) which can organize the transition, or there is NOT an
organization to provide the machinery for reaching a decision! Please
decide whether you think there is or is not an organization for Usenet.
Or at least settle on a clear vocabulary.
>The Usenet you like "the way it is" has never been that way before
>and never will be again. It is one of the fastest growing and most
>rapidly changing entities around. Nothing that grows as rapidly as
>Usenet can avoid periodically changing it's structures quite
>fundamentally.
Usenet has (several times) changed its basic structures and each
time, while there has been a seemingly chaotic transition, the resulting
ad hoc relationships between sites have been able to deal with the
change and continue operating. If there were some authority that
dictated how the net operated, transitions might occur more rapidly, but
they would occur MUCH LESS FREQUENTLY. They would also occur only after
it was obvious that the change was REQUIRED (as opposed to desireable.)
The minarchist community that is Usenet (Brad's term is really
quite descriptive!) is able to adapt rapidly BECAUSE it is not
restrained by any formal requirements to maintain any particular service
or protocol.
>>No, I'm seriously suggesting that a network like Usenet with its lack
>>of organisation is a good thing. There is no reason you can't set up
>>a parallel network with tighter central control. It's been done already,
>>over an over again: look at the gnu.* hierarchy, or vmsnet, clarinet,
>>bionet. There are also weaker networks, such as alt.*. Regional ones,
>>like tx.*. Over in Europe they have EUNET with its own hierarchy.
>
>Again, "tighter central control" isn't the issue. Nor are newsgroup
>hierarchies. Indeed, what makes you so certain the future Usenet will
>even HAVE newsgroup hierarchies? Your thinking is constricted by
>taking those for granted. I've tried to bring out the fact that
>Usenet DOES have more organization than it is usually given credit
>for and had hoped that this might be recognized as not being such
>a bad thing.
This point keeps surfacing again and again because there is a
fundamental misunderstanding. (Again) I believe that this lies in the
cultural differences between your point of view and a generic USAan
point of view. As I mentioned above, please be consistent about this
"organization" that Usenet (allegedly) has. There is certainly a great
deal of information exchanged over many links that maintain the nebulous
virtual network called Usenet, and that there is a high degree of
"organization" in that virtual network. This is easily recognized, and
most will say that maintaining this virtual organization is "a good
thing".
The rub comes when you (apparently) insist that the world as a
whole (and especially the USAan contingent) take this virtual
organization and give it some reality. In OZ, europe, and other parts
of the world, the formal networks have no problems with Usenet traffic,
and there is no appeal if those networks suppress portions of the
traffic. Try to find a site in the fj net that will publicly and
offically admit to carrying alt.sex.* (this is theoretical - my point is
that certain formal networks can AND DO suppress certain topics which
the virtual network of Usenet still carries.) In the USA, while some
specific sites don't supply certain groups, the "free market" principles
still generally apply, and the traffic can be obtained elsewhere if
the desire can support the cost.
[Usenet II commentary elided]
>
>The reason why I am probing and challenging the way Usenet is
>organized is NOT because I think I have some simple alternative set
>of "rules" that could be superior to what has evolved from the
>efforts of many wise and experienced people over some considerable
>time. On the contrary it is my respect for what has been achieved
>that leads me to want to understand it better and I happen to
>believe that the best way to understand anything is by probing
>and challenging it. If I didn't think there was something to learn
>from here, I wouldn't bother trying to change it.
Several folks have pointed out that a major portion of the
character of the virtual network called Usenet lies in the peculiar
USAan situations that require USAans to maintain the position that
Usenet (as a formal entity) does not exist. We can discuss and
speculate all we want to about a "virtual network", but it is an
insurmountable political situation that makes US maintain that there is
no "objective" network called Usenet.
Your probing and challenging has certainly helped to clarify and
deepen my own understanding of Usenet, and gives me a better handle on
planning for the transitions that are occurring even as I write these
words.
>Fine, but I cannot possibly "try changing Usenet itself" by merely
>writing some articles here and in the meantime the "test" is reaction
>to ideas and proposals, which may well result in significant changes
>to those ideas and proposals before trying anything in practice.
Have you ever heard the proverb "The Pen is mightier than the
sword"? Some of the most significant changes in the evolution of Usenet
began as simple questions or comments in an unrelated discussion.
>Let me spell out my (unhidden) agenda for reference in case you want
>to contribute some ideas.
>
>I want to see email and news technology widely used by the "general public"
>including small community organizations in both developed and developing
>countries. I want to see it becoming a new form of interactive mass
>media at least as significant as the print media and accessible to anyone
>that has access to a simple PC wordprocessor (which I expect to include
>most office workers and most homes in developed countries within a few
>years). Accessible not just as a "consumer" but also as a "producer"
>so that instead of people just soaking up ideas from the mass media,
>they can also discuss what to do about the world and other topics of
>interest.
Not a bad vision, however, it is NOT Usenet as currently
constituted. At best, I suspect that "Usenet" will be a special
interest subset of this WorldNet you forsee.
>That involves a MASSIVELY larger scale operation than Usenet and
>requires solution to many problems such as:
>
>1. Fully automatic store and forward nodes that don't need skilled
>sysadmins any more than telephones need telephone operators.
You neglect the fact that there is quite an infrastructure that
operates the network over which telephone conversations take place. And
the evolution of the public telephone network is not terribly different
from the evolution of Usenet.
The early telephone users had to deal with the operators. They
had to crank the magneto themselves to even signal that they wanted to
use the telephone.
Have you ever needed a telephone number in a distant location?
Try doing without directory assistance.
Have you ever tried calling Antarctica, or Arctic research
stations? Operators are still used for some places - Universal Direct
Dial is still only an idea, ugly reality shows the myth.
And even if most of the use of telephones is done without direct
human assistance, some of the more demanding uses of the phone network
require operator assistance. The special circuits used to link the
All-Star Baseball game broadcasters in Toronto to the rest of the world
are set-up and taken down manually (even if they go over private
satellite uplinks!)
>2. Broadcast and transport technologies that can cope with the
>higher volume.
Already well under development. ISDN and Consumer FDDI are
coming down the pike already. The house I rent here in Durham NC is
conveniently located near one of the local telco switching offices, and
major cable trunks are under the manhole covers in my front yard :-) I
could (just luckily) get ISDN delivered to my computer at home if I was
willing to pay the price.
Note, more and more, broadcast is becoming less effective as a
distribution method. Wide area distribution is much more than
"broadcasting" ever was.
>3. Filtering, hypertext and related technologies that can select
>from the higher volume and make it accessible to people interested
>in relatively low volumes.
This is exactly what is being discussed elsewhere.
>4. Intuitive user interfaces for people that can only just handle
>a wordprocessor or text editor on a Mac.
This is part of the first point, telephones aren't really that
easy, it just seems that way on the basis of cultural familiarity and a
lot of effort behind the facade.
>All those are EASY (well, relatively easy :-)
>
>But the social and political problems are MUCH more interesting
>and complex.
>
>Who builds it, who pays for it, how is it organized?
>
>As far as the "mass market" is concerned we have the experience
>of IBM's "Prodigy", the French Minitel, Prestel (or Australia's
>"Viatel" and now "Discovery"), CI$ and so forth, and
>the Baby Bells are clearly waiting in the wings.
Funny you should mention Prodigy and the Baby Bells in the same
breath. The Prodigy technology is from early 1980s and was developed at
Bell Laboratories (Piscataway) for . . . . . . . guess who. . . . .
CBS, the television people! [I was there for a while.] Bellcore still
owns the technology that Prodigy uses. Unless IBM really changed and
rewrote massive portions of the software, its underpinnings are Unix.
Every time I see a Prodigy ad or demo, I get flashbacks. :-)
>Is that the future of this technology? Seems to me the corporations
>that COULD introduce it just don't think in terms of truly
>INTERACTIVE networking. Their whole logic points to "customers"
>for "services".
Aye, there's the rub!
When you get into the situation where you are dealing with "mass
market" you run into a certain mentality. Why did NAPLPS select such a
low resolution? Why is there only a 45 character screen width? Why the
limited color palette? Basically, all these have the same answer. It
fit into using low-cost, consumer grade products. The original
Videotext Field Trial (In Fairfield NJ) had special terminal boxes and
screens, but you could do without the screens and plug it into your
television if you had the necessary connections available.
Remember please, that this was about the same time that
Commodore shocked the consumer computer market with the 22 character
wide VIC-20 at less than $1000 for a 5K memory machine!
Mass marketing offers a promise of "you can interact!" but
really says "we don't want you to cause its expensive". The marketing
of Prodigy (and its predecessor) are full of this premise.
>I don't know whether Usenet itself could actually evolve to
>BECOME something that really serves the general public. There
>are obvious arguments against. But I'm quite sure that it at
>least has important lessons that should be studied on how to
>get something going without relying on corporate structures
>to do it.
Once again, in developing VFT, the early Usenet (prior to Bnews)
had an effect on its development. The project picked up a full feed
from the Neptune Lab (because we weren't supposed to connect to other
Piscataway sites for "security" reasons - mail was routed via a py site
though :-) and early experiments with visual environments for
newsreading took place there and never went anywhere else (it was
something like a cross between vnews and trn!). One programming team
did provide a way to dump Usenet articles into the Videotext database
and provide a way to read Usenet from the Videotext side of the lab.
They were working on the reverse gateway when upper levels learned of
the experiment and squashed the concept - Usenet was "too uncontrolled"
to ever allow the consumer to access it.
I guess this was because the primary designers seemed to be
former IBM marketing types who were refugees in the Business Systems
Division at the Piscataway Lab. PY was nowhere near as much fun to work
at as Holmdel or Murray Hill, the only worse lab I saw was Whippany, and
even so, one or two sites there (the old marx brothers) were important
to early Usenet connectivity.
>P.S. Your clarification that you did not mean that Usenet consists of
>Marxists and Anarchists is understood, and also that you are not one
>or other of these yourself (though it was not necessary to go so far as
>to suggest that Trotsky was an Anarchist in order to establish that
>you are not even mildly interested in these subjects).
Gee, I guess I sort of knew this reference to Marx was here. How
droll that I just mentioned the Marx' Brothers!
The upshot of this whole comment is to point out that while
certain elements from Usenet are going to be important to the massive
public nets that will evolve, the testbed functions of Usenet are too
esoteric to be totally subsumed by the public nets.
One of the more interesting references to the future shapes of
networks is a side note in Orson Scott Card's "Ender's Game". Quite
different from the Cyberpunk vision.
--
Gregory G. Woodbury @ The Wolves Den UNIX, Durham NC
UUCP: ...dukcds!wolves!ggw ...duke!wolves!ggw [use the maps!]
Domain: g...@cds.duke.edu ggw%wol...@duke.cs.duke.edu
[The line eater is a boojum snark! ] <standard disclaimers apply>
Path: gmdzi!unido!mcsun!uunet!zaphod.mps.ohio-state.edu!samsung!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin,comp.archives.admin
Subject: Re: What is Usenet? LONG!
Message-ID: <1991Jul11.184850.10889@newshost.anu.edu.au>
Date: 11 Jul 91 18:48:50 GMT
References: <1991Jul6.214904.18944@alembic.acs.com> <1991Jul7.182322.26915@newshost.anu.edu.au> <1991Jul9.092710.5075@iccgcc.decnet.ab.com>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 107
In article <1991Jul9.092710.5...@iccgcc.decnet.ab.com>
herri...@iccgcc.decnet.ab.com writes:
>Here's one of the points where you are having trouble communicating
>with your audience, Albert. Many of us are not convinced that
>another organization would be a higher level.
Perhaps, but I suspect it is a deeper and more visceral reaction than
that. If there were some specific organizational proposals on the
table, then legitimate objections could be raised to show that they
would not improve things. But the earlier discussion showed there was
a reluctance to even admit that WE DO HAVE some kind of democratic
and centralized organization for deciding on new newsgroups and their
names and moderators. The repeated and completely irrelevant
affirmations of the independence of each site suggest to me there is
some more deep rooted opposition to the whole idea of any higher
level organization, rather than concern over whether any particular
form of organization would be better or worse.
That's unfortunate. For example it means there is no straight forward
way to determine user requirements for software development, and
instead a great deal of smugness about the 1970s technology still
in use.
There's an interesting discussion in comp.archives.admin at the
moment (to which I am cross-posting this) concerning the need for
improved internet user services and the difficulty obtaining funding
for necessary facilities like comp.archives and archie and improvements
to those. I've been following it carefully and recently printed out
a "selection" of messages which came to over 130 pages.
Most of the participants clearly know what they are talking about and
are experienced people, some of them responsible for administration
of large sites and regional networks, and presumably familiar with
the way funding organizations operate.
Yet, as with numerous discussions in news.admin and elsewhere, there
is no indication of the discussion actually leading towards any
kind of decision, assignment of responsibility and action. It is
simply another exchange of views which may indeed be very useful
to the individual participants in guiding their own separate actions
in their own organizations.
Part of the problem may be that news is such a wonderful adjunct to
organization that one almost feels as though it is a complete
substitute, even though it isn't.
When people face a common problem, in needing or wanting some
facility that would be useful to them, they tend to get together
in either formal or informal groups to hold meetings, take
decisions, assign responsibilities and then take action. Sometimes
it is just left informal, sometimes a club might be setup with
a defined membership for whom services will be provided and dues
and other support collected to provide those services, or a
lobby group may be setup to exert pressure for funding to be
provided. There are inumerable ways to get organized.
Even academics do it, with associations for the advancement of this
and that and their corresponding newsletters, journals and so forth.
Computer users certainly do it with user's groups that likewise
have publications and both help their members directly and speak
on behalf of their members to computer and software suppliers etc.
In Usenet it is so easy to call a meeting and have a discussion,
and even issue a regular publication, and it is also so easy
to take an "unofficial" initiative without needing to reach agreement
first, that the usefulness of being able to take decisions, assign
responsibilities and then take action in the name of an organized
group seems to be largely forgotton. (As though the only things
an organized group could possibly do would be tell sysadmins how to
run their own sites or impose censorship on users!)
I suspect this is compounded by the fact that conventional corporate
structures are providing the backbone transport that Usenet needs
(Internet and regionals and uunet), so being "organized" can be
left to them.
But it strikes me that the Internet and funding bodies like NSF are
following a sensible policy in not getting involved in "value added"
services like those under discussion in comp.archives.admin and
indeed Usenet itself. If users and admins want improvements in
the way services operate, and especially if they want funding for
such improvements, then they should get their act together enough
to at least form a User's Group that can speak for its members,
however few, in defining requirements. Separation of communications
carriers from content service providers seems a desirable policy.
It IS possible to take decisions, assign responsibilities and carry
out those decisions as is proved by the regular, continuous examples
of doing so in news.groups. Even though an insignificant fraction of
net posters, let alone net readers takes part, and despite the noise
level, the mere separation of formal proposals to be voted on
(in news.announce.newgroups) from general discussion, and the
subsequent announcement of voting results, is sufficient to turn
new newsgroups on, with the same moderators or unmoderated status,
more or less simultaneously at tens of thousands of sites, after
relatively brief and focussed discussions.
It should be possible to apply that principle to other kinds of
initiatives. The alternative is ONLY being able to do things
that are organized for us either by conventional corporate structures
like those that run the Internet, or as a result of individual
uncoordinated initiatives writing new software etc.
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!mcsun!uunet!zaphod.mps.ohio-state.edu!samsung!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Keywords: Usenet Prodigy history Bell Laboratories anecdotes
Message-ID: <1991Jul11.194153.11310@newshost.anu.edu.au>
Date: 11 Jul 91 19:41:53 GMT
References: <I1V5B79@taronga.hackercorp.com> <1991Jul8.191839.8793@newshost.anu.edu.au> <1991Jul10.030757.6007@wolves.uucp>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 128
In article <1991Jul10.030757.6...@wolves.uucp> g...@wolves.uucp
(Gregory G. Woodbury) quotes me and writes:
>> [...] The problem I see is that
>>there is no machinery for reaching decisions after brain-storming,
>>so the changes that WILL occur will happen haphazardly and without
>>adequate preparation.
>
> I do wish, however, that you would make up your mind! Either
>there is an organization to Usenet (which you claim makes the decision
>about newsgroups) which can organize the transition, or there is NOT an
>organization to provide the machinery for reaching a decision! Please
>decide whether you think there is or is not an organization for Usenet.
>Or at least settle on a clear vocabulary.
I am saying there is OBVIOUSLY some highly centralized, and democratic
organization for taking decisions about new newsgroups and EQUALLY
OBVIOUSLY no similar organization for taking decisions about other
matters. If further organization was established it might well use
the same mechanisms as "news.groups" etc but a glance shows that
is NOT happening at the moment. There is no contradiction in what
I am saying and no amount of demands that I make up my mind, with
or without exclamation marks, will create one.
> Usenet has (several times) changed its basic structures and each
>time, while there has been a seemingly chaotic transition, the resulting
>ad hoc relationships between sites have been able to deal with the
>change and continue operating. If there were some authority that
>dictated how the net operated, transitions might occur more rapidly, but
>they would occur MUCH LESS FREQUENTLY. They would also occur only after
>it was obvious that the change was REQUIRED (as opposed to desireable.)
>
> The minarchist community that is Usenet (Brad's term is really
>quite descriptive!) is able to adapt rapidly BECAUSE it is not
>restrained by any formal requirements to maintain any particular service
>or protocol.
On the contrary, my impression is that major upheavals, like the
"great renaming", have occurred infrequently and LONG AFTER they were
required. Adaptations like the increased reliance on the Internet
have been a result of policy decisions "dictated" from outside by
conventional organizational structures (which planned the Internet etc).
Usenet has been very SLOW to adapt, and is still essentially using a
1970s technology, BECAUSE the lack of any organization that could
take decisions about upgrading technology RESTRAINS developers to
maintaining the old services and protocols (as distinct from ensuring
gateways and conversion for backward compatability and transition,
which any competent decision making body would specify as part of
an upgrade). All sorts of good and interesting things are happening,
but it's no longer at the leading edge of computer communications
(and it's still at the bleeding edge as regards needing continuous
maintenance by admin staff).
Of course it would be POSSIBLE to setup an organization for USENET
which wasted its energy attempting to "dictate" (unsuccessfully)
and was so bureaucratic that it resulted in adaptation occurring
even more slowly. But it would require fairly careful design for
such an organization to be able to keep the main centers of
technical innovation in the computer industry around the world,
linked with a technology that is fundamentally twenty years out
of date. That is easier to achieve with a feeling that there is
no way to take decisions than with a hidebound bureaucracy
that could at least be overthrown after say 5 years or at most
a decade of technical stagnation.
BTW, I would still like to be told the meaning of "minarchist".
> [...] (this is theoretical - my point is
>that certain formal networks can AND DO suppress certain topics which
>the virtual network of Usenet still carries.) In the USA, while some
>specific sites don't supply certain groups, the "free market" principles
>still generally apply, and the traffic can be obtained elsewhere if
>the desire can support the cost.
Fine. If anybody proposes organization for the purpose of suppressing
certain topics, let them be lashed ten thousand times with wet noodles,
AS WELL AS failing ignominuously in so patently foolish an endeavour,
given the world-wide interconnectivity of the PSTN. But why are the wet
noodles (which are superfluous anyway, given the devastating effect of
PSTN reality), brought out during discussions of ORGANIZATION rather than
just during discussions of SUPPRESSION?
> Funny you should mention Prodigy and the Baby Bells in the same
>breath. The Prodigy technology is from early 1980s and was developed at
>Bell Laboratories (Piscataway) for . . . . . . . guess who. . . . .
>CBS, the television people! [I was there for a while.] Bellcore still
>owns the technology that Prodigy uses. Unless IBM really changed and
>rewrote massive portions of the software, its underpinnings are Unix.
>
>Every time I see a Prodigy ad or demo, I get flashbacks. :-)
>
>>Is that the future of this technology? Seems to me the corporations
>>that COULD introduce it just don't think in terms of truly
>>INTERACTIVE networking. Their whole logic points to "customers"
>>for "services".
>
> Aye, there's the rub!
> When you get into the situation where you are dealing with "mass
>market" you run into a certain mentality. [...]
> [...] One programming team
>did provide a way to dump Usenet articles into the Videotext database
>and provide a way to read Usenet from the Videotext side of the lab.
>They were working on the reverse gateway when upper levels learned of
>the experiment and squashed the concept - Usenet was "too uncontrolled"
>to ever allow the consumer to access it.
> I guess this was because the primary designers seemed to be
>former IBM marketing types [...]
Seems we are agreed that IBM and company aren't going to provide
the sort of INTERACTIVE and "UNCONTROLLED" news and mail that the
general public should have.
So, do we forget about it, or hope it will somehow just happen,
or get organized to DO SOMETHING ABOUT IT?
> One of the more interesting references to the future shapes of
>networks is a side note in Orson Scott Card's "Ender's Game". Quite
>different from the Cyberpunk vision.
I assume that is an SF book? I'll look out for it. (And the other
one cited in another reply to you).
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!unidui!math.fu-berlin.de!ox.com!caen!zaphod.mps.ohio-state.edu!wuarchive!uunet!icd.ab.com!iccgcc.decnet.ab.com!herrickd
From: herri...@iccgcc.decnet.ab.com
Newsgroups: news.admin,comp.archives.admin
Subject: Re: What is Usenet? LONG!
Message-ID: <1991Jul12.120227.5108@iccgcc.decnet.ab.com>
Date: 12 Jul 91 17:02:27 GMT
References: <1991Jul6.214904.18944@alembic.acs.com> <1991Jul7.182322.26915@newshost.anu.edu.au> <1991Jul9.092710.5075@iccgcc.decnet.ab.com> <1991Jul11.184850.10889@newshost.anu.edu.au>
Lines: 138
In article <1991Jul11.184850.10...@newshost.anu.edu.au>, cmf...@anu.oz.au (Albert Langer) writes:
> In article <1991Jul9.092710.5...@iccgcc.decnet.ab.com>
> herri...@iccgcc.decnet.ab.com writes:
>
>>Here's one of the points where you are having trouble communicating
>>with your audience, Albert. Many of us are not convinced that
>>another organization would be a higher level.
>
> Perhaps, but I suspect it is a deeper and more visceral reaction than
> that. If there were some specific organizational proposals on the
> [...]
> That's unfortunate. For example it means there is no straight forward
> way to determine user requirements for software development, and
> instead a great deal of smugness about the 1970s technology still
> in use.
>
This discussion you have precipitated and are perpetuating is
(or can be if you so direct it) a requirements negotiation.
> There's an interesting discussion in comp.archives.admin at the
> moment (to which I am cross-posting this) concerning the need for
> improved internet user services and the difficulty obtaining funding
> for necessary facilities like comp.archives and archie and improvements
> [...]
> When people face a common problem, in needing or wanting some
> facility that would be useful to them, they tend to get together
> in either formal or informal groups to hold meetings, take
> decisions, assign responsibilities and then take action. Sometimes
> [...]
> In Usenet it is so easy to call a meeting and have a discussion,
> and even issue a regular publication, and it is also so easy
> to take an "unofficial" initiative without needing to reach agreement
> first, that the usefulness of being able to take decisions, assign
> responsibilities and then take action in the name of an organized
> group seems to be largely forgotton. (As though the only things
> an organized group could possibly do would be tell sysadmins how to
> run their own sites or impose censorship on users!)
>
> I suspect this is compounded by the fact that conventional corporate
> structures are providing the backbone transport that Usenet needs
> (Internet and regionals and uunet), so being "organized" can be
> left to them.
>
> But it strikes me that the Internet and funding bodies like NSF are
> following a sensible policy in not getting involved in "value added"
> services like those under discussion in comp.archives.admin and
> indeed Usenet itself. If users and admins want improvements in
> the way services operate, and especially if they want funding for
> such improvements, then they should get their act together enough
> to at least form a User's Group that can speak for its members,
> however few, in defining requirements. Separation of communications
> carriers from content service providers seems a desirable policy.
>
> It IS possible to take decisions, assign responsibilities and carry
> out those decisions as is proved by the regular, continuous examples
> of doing so in news.groups. Even though an insignificant fraction of
> [...]
> It should be possible to apply that principle to other kinds of
> initiatives. The alternative is ONLY being able to do things
> that are organized for us either by conventional corporate structures
> like those that run the Internet, or as a result of individual
> uncoordinated initiatives writing new software etc.
>
I'm beginning to get a perspective on what is going on here.
Usenet, here in the USofA, is the purest marketplace I have ever seen.
There is continuing discussion of various ideas by the people who are
interested in those ideas. It is aimless and meandering. Participants
come and go while the discussion continues. Every once in a while,
someone gets excited enough to take action.
The action can be to try to topple the RSA patent by freely publishing
an unfettered alternative.
It can be to gather some corporations into an expensive co-operative
effort to implement Privacy Enhanced Mail.
It can be to write a tool like Cnews.
And the market responds. RSA shouts, "You're infringing!" and the
author backs off. PEM grows in the corner, maybe to proliferate,
maybe to sit in a niche, maybe to fade away. Cnews seems about to
take over the world, though Bnews lingers on, and at least one
non-customer is very unhappy with Cnews.
You look at this and say, "It's not neat. Nobody knows what is
happening. It is out of control. Nobody is planning the future.
I know what the future should look like, but there is no one to
whom I can go and show him my vision and get him to decide that
this is the future we will build."
I look at it and say, "Hey! That's neat! There's no out of reach
authority deciding what my network is going to look like. If I want
to install a fribblewhiz, all I have to do is persuade the sysadmins
on the other end of my phone calls to install one, too. And, if
the sysadmins whose systems I currently call don't want to co-operate,
I can forget them and call some who do."
A vision of the future that is currently being applied in an island
of the net is a transport that allows a reader to subscribe to a
group that does not currently come to the system on which they
read news. A few minutes or a few hours later, the traffic shows
up so they can read it. This will spread, or fade away, or spring
up in other unconnected islands, or just stay where it is. There
isn't anyone who can tell us to do it that way.
There are other nets, connected in in various ways, where there is
such an authority. NSFNET (have fun explaining your vision to the
US Congress), PRODIGY (they may be beginning to realize they need
a better vision of what a net is), COMPUSERVE (don't they seriously
try to provide what their customers want - mail connection to other
networks is following a classic marketplace action path), FIDO (I
may be misunderstanding, but some things posted here suggest they
have some people who make decisions), EUNET (has recently displayed
some responsiveness to the voice of the market), PSINET (maybe they
would be interested in your ideas as a tool to gain an advantage
over UUNET).
In many parts of the US, the local telephone company will soon have
ISDN on the pole in front of people's houses. They want things on
ISDN that will make individuals subscribe. This could be the most
fruitful path for you to sell your vision to someone who can decide
to realize it.
However, here in the US, your vision will have to compete with a
bunch of other visions, and individuals will decide to buy your
vision, or buy a competing vision, or continue their lives without
computer networks.
I like it that way.
dan herrick d...@ncoast.org
PS. When you posted some indications of what your vision is, I saw
some things that look good. I think some others did, too. You could
just redirect this discussion into the development of a specification.
Someone might throw together a prototype before you thought the
specification was done.
Path: gmdzi!unido!mcsun!uunet!sugar!taronga!peter
From: pe...@taronga.hackercorp.com (Peter da Silva)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <I516JQK@taronga.hackercorp.com>
Date: 13 Jul 91 13:05:00 GMT
References: <1991Jul11.194153.11310@newshost.anu.edu.au>
Organization: A corner of our bedroom
Lines: 19
cmf...@anu.oz.au (Albert Langer) writes:
> but it's no longer at the leading edge of computer communications
What is? IRC? BIX? Fax?
> (and it's still at the bleeding edge as regards needing continuous
> maintenance by admin staff).
I have spent a total of half an hour of maintainance on Sugarland UNIX
over the past six months. Add a couple of hours over the same period
for Taronga, mainly because of disk space problems, and you hardly
get "continuous maintainance".
I think there must be something wrong with the Usenet software at ANU.
I really do. ANU is where that VAX/VMS news software came out of, isn't
it?
--
Peter da Silva. `-_-' Taronga Park BBS +1 713 568 0480 2400/n/8/1
Taronga Park. 'U` "Have you hugged your wolf, today?"
Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin,comp.archives.admin
Subject: Usenet User's Association? (Was: Re: What is Usenet? LONG!
Message-ID: <1991Jul14.173626.3010@newshost.anu.edu.au>
Date: 14 Jul 91 17:36:26 GMT
References: <1991Jul9.092710.5075@iccgcc.decnet.ab.com> <1991Jul11.184850.10889@newshost.anu.edu.au> <1991Jul12.120227.5108@iccgcc.decnet.ab.com>
Sender: n...@newshost.anu.edu.au
Followup-To: news.admin
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 71
In article <1991Jul12.120227.5...@iccgcc.decnet.ab.com>
herri...@iccgcc.decnet.ab.com writes:
>This discussion you have precipitated and are perpetuating is
>(or can be if you so direct it) a requirements negotiation.
Peter da Silva and I already seem to be in a dialog on requirements
negotiation in news.admin and you are welcome to join. My responses will have
a new subject line "Requirements for General Public Access". There is also
a partially overlapping discussion of mail addressing in comp.mail.misc
(some of it cross-posted to comp.org.eff.talk and even comp.protocols.tcp-ip).
>I'm beginning to get a perspective on what is going on here.
Thanks for your comments, which at least take the discussion beyond the
old paranoia about any organization being for censorship or to tell
sysadmins how to run their own sites. I will respond shortly in a
separate article, only in news.admin, as the points I raised relevant
to comp.archives.admin have not been followed up (though I hope some people
following the comp.archives discussion might be interested in joining
the wider discussion of organization in news.admin.)
As we have moved beyond "What is Usenet", I will respond under the
subject "Usenet User's Association?"
>PS. When you posted some indications of what your vision is, I saw
>some things that look good. I think some others did, too. You could
>just redirect this discussion into the development of a specification.
>Someone might throw together a prototype before you thought the
>specification was done.
Please see (and join) the discussion with Peter under "Requirements for
General Public Access".
BTW, part of the problem, which relates back to organization and also
perhaps to comp.archives is that software development for Usenet and
related services is organized ONLY at a level where people CAN throw
together a prototype before a proper specification is done. Not all
software CAN usefully be produced that way. Nor can all useful and
even essential services be operated without suitable organization,
once the prototype has been successfully demonstrated. (See
comp.archives.admin discussion).
Leaving aside whether or not I am right about needing X.400 and X.500
and a major software engineering effort in order to meet requirements
for General Public Access, I am sure you would agree that SOME projects
are simply too large and complex for individuals or small teams to
organize informally and spread spontaneously without some apparatus
for decision making. For example one could not launch a satellite
broadcast of Usenet without more organization than we currently have
even if it was proved beyond doubt that doing so could result in
substantial cost savings for EVERY Usenet site (if necessary by
re-distributing the cost benefits to ensure this).
If it DOES happen to be true that such benefits COULD be achieved by
a satellite broadcast then we are ALL suffering significant costs
for not having an organization capable of doing it. Likewise with
any major software development that may be necessary or desirable.
Should any such projects be left to the conventional corporate world?
Or is it possible that the people who have demonstrated such creativity
and imagination in building Usenet and the freely available software
that goes with it, and in setting up archives etc, all by voluntary
effort (or diversion of resources from other jobs), could also
build larger scale software projects that require greater coordination?
They do it at work. Can it be done for fun too?
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Requirements for General Public Access (Was: Re: What is Usenet?
Message-ID: <1991Jul14.185420.3553@newshost.anu.edu.au>
Date: 14 Jul 91 18:54:20 GMT
References: <1991Jul11.194153.11310@newshost.anu.edu.au> <I516JQK@taronga.hackercorp.com>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 81
In article <I516...@taronga.hackercorp.com> pe...@taronga.hackercorp.com
(Peter da Silva) quotes me and writes:
>> but it's no longer at the leading edge of computer communications
>
>What is? IRC? BIX? Fax?
No, none of those, though Fax, despite usually needing a separate
phone line and using phone numbers for addressing, HAS become a
MUCH more widely used general public service than email DESPITE
the clear superiority of email (which can include Fax). That illustrates
how enormous are the opportunities that have been lost as a result of
the technical stagnation in Usenet. Email could EASILY have been
developed for direct exchange over the PSTN with suitable standards
(which could also embrace Fax picture files). That was only taken up
by hobbyists like FidoNet rather than by bodies that could establish
standards, so email and news lag behind for several years while a
much more primitive technology takes off.
For details of the leading edge take a look at some of the research
projects going on - e.g. as summarized in the book of papers on
"Computer Supported Collective Work" by Irene Greibach(sp?). Or even
just look at the May issue of Byte magazine (WAIS and Information Lens).
When the current Usenet technology was introduced - in the 1970s - it
WAS as part of a leading edge "research" network. It's quite amusing
to read books like Vallee's "Network Revolution" complaining about
the unfriendliness of that 1970s technology, which we STILL use.
But what is particularly galling is that commercial and corporate
email networks are leap-frogging Usenet with X.400 and X.500
implementation (not just reseach), RIGHT NOW, while Usenet gurus
are still saying it's "too big" for them to even read the damn
specs (which were first published 6 years ago and revised nearly
3 years ago)! Even Fax will be absorbed.
>> (and it's still at the bleeding edge as regards needing continuous
>> maintenance by admin staff).
>
>I have spent a total of half an hour of maintainance on Sugarland UNIX
>over the past six months. Add a couple of hours over the same period
>for Taronga, mainly because of disk space problems, and you hardly
>get "continuous maintainance".
Well, there you have me. I must admit that my belief in the necessity
for X.400 and X.500 to achieve an unattended node (apart from any
other benefits), was based on an assumption that an average of 10
seconds maintenance time per day would be very difficult to achieve.
Even 40 seconds per day for Taronga is pretty impressive.
If you have achieved it already, even though at only two small sites and
with a skilled C programmer as sysadmin, the possibility of doing it
more widely with just current RFC based software must be taken seriously.
I will refrain from responding to some of your (many) other remarks on this
point as it is more useful to examine the empirical evidence.
Is there any chance of you packing up a complete copy of your configuration,
including ALL the relevant files as well any software you have customized,
so I can have a look at how you did this and try it myself? (I realize
that installation and customization for particular neighbours etc are a
separate issue, but if such low maintenance time CAN be achieved with an
RFC based system it is worth investigating how to design simplified
installation and customization scripts - and perhaps even worth just putting
up with initial time-consuming manual installation).
>I think there must be something wrong with the Usenet software at ANU.
>I really do. ANU is where that VAX/VMS news software came out of, isn't
>it?
This is the sixth newsgroup and third message in which you have referred
to my invalid header line, escalating from a suggestion to near certainty
about the nature of the problem within a few days and before I got around
to replying on this very uninteresting subject. I will try to eventually
explain the situation in comp.mail.misc. Meanwhile please note that I am
only a visitor at ANU and my estimates of the workload for system
administration have nothing to do with the setup here.
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!mcsun!uunet!zaphod.mps.ohio-state.edu!menudo.uh.edu!sugar!taronga!peter
From: pe...@taronga.hackercorp.com (Peter da Silva)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access (Was: Re: What is Usenet?
Message-ID: <TV26ZI9@taronga.hackercorp.com>
Date: 15 Jul 91 02:28:25 GMT
References: <1991Jul14.185420.3553@newshost.anu.edu.au>
Organization: A corner of our bedroom
Lines: 102
cmf...@anu.oz.au (Albert Langer) writes:
> No, none of those, though Fax, despite usually needing a separate
> phone line and using phone numbers for addressing, HAS become a
> MUCH more widely used general public service than email DESPITE
> the clear superiority of email (which can include Fax). That illustrates
> how enormous are the opportunities that have been lost as a result of
> the technical stagnation in Usenet.
Nothing to *do* with Usenet. The reason FAX is so popular is simply that
the machine that could send FAX, with a price tag of $X000, was developed
in Japan. And it was developed because Japan has massive problems with
email. Their whole written language is pictographic, and until *very*
recently machines for entering it were (a) very expensive, and (b) very
hard to use.
FAX is simply a product of the dominance of the Japanese in consumer
electronics.
> Even Fax will be absorbed [by X.400 and X.500].
I don't believe so. The technology for X.400 addressing aided by X.500
directory services is just too expensive for consumers. You can walk into
a store and get a FAX machine for $500.
Plus, FAX has a major advantage in ease of addressing. In fact, I believe
that is its greatest advantage. People *need* short, unambiguous names.
> Well, there you have me. I must admit that my belief in the necessity
> for X.400 and X.500 to achieve an unattended node (apart from any
> other benefits), was based on an assumption that an average of 10
> seconds maintenance time per day would be very difficult to achieve.
> Even 40 seconds per day for Taronga is pretty impressive.
It doesn't *need* that much. It just *gets* that much because it's
sitting right there in my bedroom.
And, you know what, my telephone has cost the phone company more time
than that, what with changes to my class of service.
> If you have achieved it already, even though at only two small sites and
> with a skilled C programmer as sysadmin, the possibility of doing it
> more widely with just current RFC based software must be taken seriously.
Thanks for the kudos, but I'm just running straight C-news under UNIX.
I haven't written any special code, just used what's out there on the net.
The point is that RFC822, phone numbers, Zip codes, and X.400 addressing
protocols are just that. Protocols. They can be implemented well or ill.
They can have research implementations, or commercial ones. And that's
what's needed: commercial quality implementations. C news is better than
B news or Notes, but it's got a ways to go yet.
If I may be permitted a digression, that's where the GNU effort falls
apart. It's not capable of producing those implemantations. Sure, you
could spend six months polishing gfoo (for foo = cc, emacs, etc) and
try to sell it for enough to pay for the work, but you wouldn't be
able to because your first customer would take it and sell it out from
under you... or you'd have to price it out of reach of that first
customer. The only way to make money off GNU software is consulting
fees, or use it to help sell hardware if you're a computer vendor...
and that's *not* the way to create a standard, portable system.
> Is there any chance of you packing up a complete copy of your configuration,
> including ALL the relevant files as well any software you have customized,
> so I can have a look at how you did this and try it myself?
Just get a copy of:
cnews
elm
smail 2.5
pathalias
To build these, you'll also need
unshar
patch
Installation is a bear, but you only have to do that once. That's where
the commercial grade software part comes in.
For map unpacking you'll have to wait. I'll see what Karl has set up at
sugar... I didn't do that part. Here I just depend on sugar and uhnix1 for
routing. Ferranti is a different matter... it's running under a 286 version
of Xenix and is held together by scotch tape and baling wire. Politics. I'm
trying to get the news moved over to a 386 box but right now it's all I can
do to keep Ferranti on the net.
So add that to your list of requirements... an 80386 or equivalent
processor, at least with the current news software. Both cnews and
pathalias blow out at FICC.
> This is the sixth newsgroup and third message in which you have referred
> to my invalid header line,
Well, it's getting on my nerves. Why can't you just put a "reply-to" line
in and let the automatic tools handle it properly? I know you can edit
your headers, because you've been moving messages around from one group to
another.
--
Peter da Silva. `-_-' Taronga Park BBS +1 713 568 0480 2400/n/8/1
Taronga Park. 'U` "Have you hugged your wolf, today?"
Path: gmdzi!unido!mcsun!uunet!zaphod.mps.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!ncoast!dlh
From: d...@NCoast.ORG (daniel lance herrick)
Newsgroups: news.admin
Subject: Re: Usenet User's Association? (Was: Re: What is Usenet? LONG!
Message-ID: <1991Jul15.100530.15096@NCoast.ORG>
Date: 15 Jul 91 10:05:30 GMT
References: <1991Jul11.184850.10889@newshost.anu.edu.au>
<1991Jul12.120227.5108@iccgcc.decnet.ab.com>
<1991Jul14.173626.3010@newshost.anu.edu.au>
Organization: North Coast Public Access Un*x (ncoast)
Lines: 106
In article <1991Jul14.173626.3...@newshost.anu.edu.au> cmf...@anu.oz.au (Albert Langer) writes:
>In article <1991Jul12.120227.5...@iccgcc.decnet.ab.com>
>herri...@iccgcc.decnet.ab.com writes:
>
>>This discussion you have precipitated and are perpetuating is
>>(or can be if you so direct it) a requirements negotiation.
>
>Peter da Silva and I already seem to be in a dialog on requirements
>negotiation in news.admin and you are welcome to join. My responses will have
Yes, I saw the first pieces of that discussion a few minutes after
writing the piece you reference here.
>perhaps to comp.archives is that software development for Usenet and
>related services is organized ONLY at a level where people CAN throw
>together a prototype before a proper specification is done. Not all
>software CAN usefully be produced that way. Nor can all useful and
>even essential services be operated without suitable organization,
>once the prototype has been successfully demonstrated. (See
>comp.archives.admin discussion).
>
"Not all software..." one of my consulting clients
maintains a stable of 50-60 programmers to keep a particular
software product in the market. As I understand it, there is
an organization of four or five guys in a garage who are knocking
them silly in the marketplace. The market share of my client
is said to come from the same kind of name loyalty as sells small
computers with the three magic blue letters.
I would hate to think how big and complicated the Usenet software
suite would be if it came out of the software making systems of
this client.
>Leaving aside whether or not I am right about needing X.400 and X.500
>and a major software engineering effort in order to meet requirements
>for General Public Access, I am sure you would agree that SOME projects
>are simply too large and complex for individuals or small teams to
>organize informally and spread spontaneously without some apparatus
>for decision making. For example one could not launch a satellite
>broadcast of Usenet without more organization than we currently have
>even if it was proved beyond doubt that doing so could result in
>substantial cost savings for EVERY Usenet site (if necessary by
>re-distributing the cost benefits to ensure this).
>
There is a long tradition in the United States of foundations
providing seed money to start a worthy project in the expectation
that it will become self sustaining in a year or two. I don't
share your pessimism about the viability of satellite broadcast
of Usenet.
One audio channel with enough bandwidth for an fm broadcast signal
for a year would not require the budget of a small country. It is
well within the resources of a small group to start the broadcasts
on speculation. They would require one or two year's funding and
the imagination to work out a payment scheme that the Usenet
community would buy into. Probably the biggest impediment is the
militant conviction in some portions of the Usenet that somebody
else should pay. However, the people holding that conviction are
not the ones who would subscribe to such a service.
There is a large constituency that already has the technology
necessary to read such a broadcast. It is not a development
project. You will find some of them gathered in rec.ham.radio,
or whatever they call that group now.
I believe the OSCAR satellites provide an explicit counterexample
for your pessimism. The AR stands for Amateur Radio. The "Amateur"
part is a statutory requirement and a militant conviction. The
formal bindings in amateur radio are weaker than the bindings in
the Usenet community. The amateur calls CQ and sets up a conversation
on the spot. Usenet requires advance co-operation between the two
parties of a conversation (transport level) and there is usually
a long series of these conversations. There are such long term
relationships among amateurs, but there is a much higher proportion
of one-shot, we'll never talk again, contacts, than there is of
"anonymous UUCP" connections.
The point is, OSCAR found the resources to build a satellite and get
it launched. eight or more times. Those resources came from hobby
beneficiaries of the service. The service is open to all who know
how to use it, regardless of their participation in providing the
resources. It is the realization of the vision of a very small
group of amateurs. (I do not believe there was any "seed money"
to start OSCAR. It was the kind of things you are now doing
producing contributions, in kind, and in funds.)
>If it DOES happen to be true that such benefits COULD be achieved by
>a satellite broadcast then we are ALL suffering significant costs
>for not having an organization capable of doing it. Likewise with
>any major software development that may be necessary or desirable.
>
>Should any such projects be left to the conventional corporate world?
>Or is it possible that the people who have demonstrated such creativity
>and imagination in building Usenet and the freely available software
>that goes with it, and in setting up archives etc, all by voluntary
>effort (or diversion of resources from other jobs), could also
>build larger scale software projects that require greater coordination?
>
>They do it at work. Can it be done for fun too?
Yes. You may find yourself turning into the leader around whom such
an effort forms. But probably not if you preserve the conviction
that only "higher forms of organization" can accomplish great works.
dan herrick d...@ncoast.org
Path: gmdzi!unido!math.fu-berlin.de!ox.com!yale.edu!think.com!samsung!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: Usenet User's Association?
Message-ID: <1991Jul16.210123.7517@newshost.anu.edu.au>
Date: 16 Jul 91 21:01:23 GMT
References: <1991Jul12.120227.5108@iccgcc.decnet.ab.com> <1991Jul14.173626.3010@newshost.anu.edu.au> <1991Jul15.100530.15096@NCoast.ORG>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 255
In article <1991Jul15.100530.15...@NCoast.ORG> d...@NCoast.ORG
(daniel lance herrick) writes:
>There is a long tradition in the United States of foundations
>providing seed money to start a worthy project in the expectation
>that it will become self sustaining in a year or two. I don't
>share your pessimism about the viability of satellite broadcast
>of Usenet.
>
>One audio channel with enough bandwidth for an fm broadcast signal
>for a year would not require the budget of a small country. It is
>well within the resources of a small group to start the broadcasts
>on speculation. They would require one or two year's funding and
>the imagination to work out a payment scheme that the Usenet
>community would buy into. Probably the biggest impediment is the
>militant conviction in some portions of the Usenet that somebody
>else should pay. However, the people holding that conviction are
>not the ones who would subscribe to such a service.
I am not pessimistic about the viability of satellite broadcast,
but frustrated about the fact that it has been viable for so long
and has still not been implemented due to political/organizational
problems such as the difficulties of working out a payment scheme
as you mention DESPITE the fact that it is viable and the total
costs for ALL could be reduced. When such a situation exists,
it NECESSARILY implies some inadequacy of organization in being
unable to make use of opportunities.
I suspect you may be right that it has even become possible to
implement a satellite broadcast as a subscription service
started by some small group on speculation WITHOUT Usenet
having the degree of organization needed to take responsibility
itself. But that merely highlights the fact that several years
have already been wasted, during which it COULD have been viable
if Usenet was sufficiently organized and that we are STILL WAITING
for it to BECOME viable despite Usenet NOT having being sufficiently
organized to implement it earlier.
This has already cost existing sites substantially as well
as unnecessarily excluding additional sites (and countries).
>I believe the OSCAR satellites provide an explicit counterexample
>for your pessimism. The AR stands for Amateur Radio. The "Amateur"
>part is a statutory requirement and a militant conviction. The
>formal bindings in amateur radio are weaker than the bindings in
>the Usenet community. The amateur calls CQ and sets up a conversation
>on the spot. Usenet requires advance co-operation between the two
>parties of a conversation (transport level) and there is usually
>a long series of these conversations. There are such long term
>relationships among amateurs, but there is a much higher proportion
>of one-shot, we'll never talk again, contacts, than there is of
>"anonymous UUCP" connections.
Once again, I am not pessimistic.
It's been a while since I used the amateur licence I qualified for
when I was 15, (VK3YHD), but I don't need to be told that the AR
in Oscar stands for Amateur Radio (and the OSC stands for
Orbital Satellite Carrying). The "one night stand" aspect of
amateur radio that you mention is closer to IRC than to newsgroups,
however you are quite mistaken about the "formal bindings" of
amateur radio being weaker than the bindings in the Usenet community.
Apart from Government supervised licensing and certificate
examinations, there are a whole range of highly organized
amateur radio activities, which include band plans (frequency
assignments) repeater networks and of course OSCAR. In Australia
there is (or was when I was more familiar with things, many years ago),
an amateur radio association called the "Wireless Institute of
Australia". In North America I believe the equivalent is called
the "Amateur Radio Relay League". Both names sound quaint because
they are quite ancient.
As I recall, the ARRL in particular has a LARGE and somewhat
bureaucratic apparatus with many full time staff, divisional
offices, testing laboratories and so forth - far more organized
than would appear to me (and I think to many North American
radio amateurs), to be necessary for that hobby.
Certainly projects like OSCAR were NOT achieved on the basis of
the same kind of organization as Usenet projects, but were
able to obtain funding, and in particular rocket launch space,
on the basis of endorsement by organizations that Governments
and other bodies could take seriously.
In short, your counter-example proves my point. :-)
We don't even have a simple "Usenet User's Association" with
none of the pretensions of the ARRL.
>>Should any such projects be left to the conventional corporate world?
>>Or is it possible that the people who have demonstrated such creativity
>>and imagination in building Usenet and the freely available software
>>that goes with it, and in setting up archives etc, all by voluntary
>>effort (or diversion of resources from other jobs), could also
>>build larger scale software projects that require greater coordination?
>>
>>They do it at work. Can it be done for fun too?
>
>Yes. You may find yourself turning into the leader around whom such
>an effort forms. But probably not if you preserve the conviction
>that only "higher forms of organization" can accomplish great works.
The thing that gets me about advocates of "free market", "libertarianism",
"anarchy" and so forth, and people who are very suspicious of
"organization" and especially "centralized" organization, is how
QUICKLY they start looking for a "leader" to actually get anything done. :-)
A project to enable General Public access to this technology, will
I believe be feasible, whether Usenet gets it's act together or not
(just as the Gnu project, ISODE, C news team etc all exist without
having first transformed Usenet as a precondition for existing). I hope to
contribute to that, but it will depend on involving a number of people
with greater project management and software engineering skills than mine,
in an organization capable of doing it, not just finding a "leader".
I like to think of myself as an "ideas person" and I hope some
of the ideas I am promoting get taken up, but I'm not the one to
take charge of implementing anything.
If Usenet is going to establish any higher form of organization than it
currently has, an initiative to do so will have to be endorsed by a
number of people who are well known for their contributions to the
net, it won't spring forth from a bright idea by me.
All I hope to do is help prepare public opinion, for example
by drawing attention to the existing level of organization and
undermining the conventional fictions about it's non-existance.
Now, returning to my promised reply to the rest of your previous article.
In article <1991Jul12.120227.5...@iccgcc.decnet.ab.com>
herri...@iccgcc.decnet.ab.com writes:
>Usenet, here in the USofA, is the purest marketplace I have ever seen.
>There is continuing discussion of various ideas by the people who are
>interested in those ideas. It is aimless and meandering. Participants
>come and go while the discussion continues. Every once in a while,
>someone gets excited enough to take action.
>
>The action can be to try to topple the RSA patent by freely publishing
>an unfettered alternative.
>
>It can be to gather some corporations into an expensive co-operative
>effort to implement Privacy Enhanced Mail.
>
>It can be to write a tool like Cnews.
>
>And the market responds. RSA shouts, "You're infringing!" and the
>author backs off. PEM grows in the corner, maybe to proliferate,
>maybe to sit in a niche, maybe to fade away. Cnews seems about to
>take over the world, though Bnews lingers on, and at least one
>non-customer is very unhappy with Cnews.
First, Usenet is not "here in the USofA" though your euologies to
the marketplace have that very distinctive national stamp :-).
Second, what you describe is not a "marketplace", pure or otherwise.
Essential to the concept of a marketplace is some form of exchange
of property rights between independent property owners, usually
through exchange of money for goods or services (or financial
instruments which indirectly depend on the former), sometimes
through barter. The Usenet activities you describe involve no
such "market exchange" and are closer in spirit to a communistic
exchange based on "from each according to their ability, to
each according to their need". That of course is especially
difficult to acknowledge in the "USofA" where it is rather safer
to describe things as a free marketplace than a commune :-)
I think the point you are trying to make is not about "markets",
but about "pluralism" and the advantages which pluralism - many
independent centers of decision making, has over any single
center of decision making.
>You look at this and say, "It's not neat. Nobody knows what is
>happening. It is out of control. Nobody is planning the future.
>I know what the future should look like, but there is no one to
>whom I can go and show him my vision and get him to decide that
>this is the future we will build."
No, that isn't what I say. I like what I see, including the aimless
meandering (without which I would have been ruled out of order
long ago :-)
What I'm saying is that there is something MISSING, but it isn't
somebody who can decide what future to build. What's missing is
the kind of organizational structures that can take on large
projects (with the sole exception of the news.groups naming
structure which takes on the very large project of channelling
discussions among tens of thousands of participants to hundreds
of thousands of onlookers at tens of thousands of sites into
a few hundred newsgroup labels - quite efficiently and successfully
and without any need for coercive powers, full-time bureaucracies
etc etc)
>I look at it and say, "Hey! That's neat! There's no out of reach
>authority deciding what my network is going to look like. If I want
>to install a fribblewhiz, all I have to do is persuade the sysadmins
>on the other end of my phone calls to install one, too. And, if
>the sysadmins whose systems I currently call don't want to co-operate,
>I can forget them and call some who do."
But why need any ADDITIONAL form of organization be "out of reach",
and why need it prevent you from installing a fribblewhiz or doing
ANYTHING that can be done the way things are done now (e.g. by
finding neighbours who will cooperate)?
Some things are just larger in scale than what you can do together
with your neighbours. Do we leave those things to conventional
corporations?
>A vision of the future that is currently being applied in an island
>of the net is a transport that allows a reader to subscribe to a
>group that does not currently come to the system on which they
>read news. A few minutes or a few hours later, the traffic shows
>up so they can read it. This will spread, or fade away, or spring
>up in other unconnected islands, or just stay where it is. There
>isn't anyone who can tell us to do it that way.
Fine, but why is it taking so long to permeate when similar technology
(AREAFIX) spread throughout FidoNet years ago?
>In many parts of the US, the local telephone company will soon have
>ISDN on the pole in front of people's houses. They want things on
>ISDN that will make individuals subscribe. This could be the most
>fruitful path for you to sell your vision to someone who can decide
>to realize it.
In Australia ISDN is already available in most places for AUD $839 p.a.
rental (2 x 64Kbs circuit switched B channels and 1 x 16Kbs packet
switched D channel). Certainly the specs are already in place for
the PTTs to deliver public access to email via ISDN (and incidentally,
there are also standards for email addressing via X.400 MTAs at
the cablehead of cable TV distributors).
But their orientation is NOT towards uncontrolled "news" like
Usenet and their pricing structure will not be aimed at ensuring
the widest possible access as quickly as possible. I believe it
should be possible to infiltrate something like "news" onto
most desktop PCs (of which 40% are already connected to LANs,
with 80-90% expected by 1995) and ALSO most domestic PCs (which
will soon be included with domestic Hi Fi sets). The transport
is available, the specs are available, but the PTTs just aren't
particularly oriented towards doing it.
Perhaps there are enough people around Usenet, who have seen the
benefits of unrestricted "news" (including moderated news), that
WE can do it. But it will take more organization than installing
a fribblewhiz does.
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!math.fu-berlin.de!ox.com!yale.edu!qt.cs.utexas.edu!cs.utexas.edu!uunet!munnari.oz.au!bunyip.cc.uq.oz.au!uqcspe!cs.uq.oz.au!rhys
From: r...@cs.uq.oz.au (Rhys Weatherley)
Newsgroups: news.admin
Subject: Re: Usenet User's Association?
Message-ID: <2463@uqcspe.cs.uq.oz.au>
Date: 17 Jul 91 04:23:20 GMT
References: <1991Jul12.120227.5108@iccgcc.decnet.ab.com> <1991Jul14.173626.3010@newshost.anu.edu.au> <1991Jul15.100530.15096@NCoast.ORG> <1991Jul16.210123.7517@newshost.anu.edu.au>
Sender: n...@cs.uq.oz.au
Reply-To: r...@cs.uq.oz.au
Lines: 85
In <1991Jul16.210123.7...@newshost.anu.edu.au> cmf...@anu.oz.au (Albert Langer) writes:
>The thing that gets me about advocates of "free market", "libertarianism",
>"anarchy" and so forth, and people who are very suspicious of
>"organization" and especially "centralized" organization, is how
>QUICKLY they start looking for a "leader" to actually get anything done. :-)
What I find the most frustrating about the current USENET organisation is that
you don't know who to send your ideas to. Sure, we can fight and wrangle all
we want in news.admin, comp.archives.admin, etc, but who's collecting all of
this information with a view to taking the good bits from all the ideas and
putting together a summary of all the good ideas to be posted and wrangled on
about some more until finally a proposal emerges that can be thrown at the
RFC editors or whatever as the "New World Order"?
Albert has a lot of ideas, I have a lot of ideas, everyone has a lot of ideas
about what we'd like USENET to look like, but who's the guy we send this
to as a view to changing things? I've seen a lot of ideas, but barring RFC's,
I haven't seen any concrete followups to the ideas in the form of proposals.
Maybe all of this is going on behind our back in task force meetings or
whatever? If so, where's the minutes of these task forces? If they're on an
FTP site, which one is it, and why aren't they being posted to be debated?
Those of us in other countries (Australia in Albert's and my case) cannot
just get on the next bus to one of these meetings. Where's the suggestion
box for the meetings?
If there was just a handful of designated people on USENET that moderated
a newsgroup like "news.new.world.order" :-) and who saved everything away
to be summarised and reposted at the end of each month for further comment,
then maybe we'd get some work done. (It has to be moderated to stop open
slather junk like in the current groups where proposals soon get lost). The
problem is, these people would need to do something like this more or less
full-time - not as an adjunct to some other job. Heck, universities could
collect an extra $5 from each of their students every year (a pittance
really) in the student fees or whatever, and pay for such a group of people
to live comfortably while they look after the management of combining
proposals and start pushing us forward. It could also go towards funding
other services.
I don't know what the average sort of wage for such a person in the USA is,
but if we were to say $30,000 (Australian), then that's only 6000 students
(we have 3 times that many here at my university). So, add up over all
students in the world and you'll be able to pay these people quite well, even
if means that every student in the world will have to see one less movie a
year :-) . Getting the unis to agree with this is left as an exercise for
the reader :-) . Such "distributed cost" doesn't hurt the individual person,
but allows funding for more net-people and better networks to take place
quite easily.
I could see a situation like the following: Someone gets a "trendy" idea
for how the basic structure of things could be improved. They write a 3-page
proposal which is sent to the moderators and subsequently posted to the group
"news.new.world.order". After a week to a month or so of (heated) debate in
the group "news.new.world.order.d", either that person starts a new proposal
(or goes away entirely :-), or it's adopted into the scheme of things by the
moderators (by public "vote"), whereby it goes into the bi-monthly summary
of things achieved and still to be done.
These "moderators" would also be working in the background to "wordprocess"
proposals into better and better forms, based on the ensuing discussion, and
with consultation with the originators who can help out as well, reposting at
strategic times until finally we all agree on what needs to be done. At that
time, it's off to the RFC editors and other standards bodies, maybe even with
a prototype or two in tow that have been sketched out along the way. This
would be mainly an extension of the moderation duties of a person (where that
person is dedicated to that moderation and summarisation and nothing else),
and won't entail a great overhaul of the organisation of USENET. And, it's
not a big clunky body with the dirty name "central organisation", but a
distributed network of ordinary USENET people gradually refining specifications
until everyone is satisfied, with a few people keeping the process from
degenerating. Don't get me wrong. These people won't be glorified
secretaries: they'd need to be up on all of the current network stuff as well.
But probably the people who should really read this because they can do
something about it don't read news. :-(
Cheers,
Rhys.
+=====================+==================================+
|| Rhys Weatherley | The University of Queensland, ||
|| r...@cs.uq.oz.au | Australia. G'day!! ||
|| "I'm a FAQ nut - what's your problem?" ||
+=====================+==================================+
Path: gmdzi!unido!math.fu-berlin.de!ox.com!caen!sdd.hp.com!cs.utexas.edu!utgpu!utzoo!henry
From: he...@zoo.toronto.edu (Henry Spencer)
Newsgroups: news.admin
Subject: Re: Usenet User's Association?
Message-ID: <1991Jul17.052648.24883@zoo.toronto.edu>
Date: 17 Jul 91 05:26:48 GMT
References: <1991Jul12.120227.5108@iccgcc.decnet.ab.com> <1991Jul14.173626.3010@newshost.anu.edu.au> <1991Jul15.100530.15096@NCoast.ORG> <1991Jul16.210123.7517@newshost.anu.edu.au> <2463@uqcspe.cs.uq.oz.au>
Organization: U of Toronto Zoology
Lines: 28
In article <2...@uqcspe.cs.uq.oz.au> r...@cs.uq.oz.au writes:
>What I find the most frustrating about the current USENET organisation is that
>you don't know who to send your ideas to. Sure, we can fight and wrangle all
>we want in news.admin, comp.archives.admin, etc, but who's collecting all of
>this information with a view to taking the good bits from all the ideas and
>putting together a summary of all the good ideas to be posted and wrangled on
>about some more until finally a proposal emerges that can be thrown at the
>RFC editors or whatever as the "New World Order"?
I'll let you in on one of the deep, dark secrets of the Internet world.
The way to establish a New World Order is not to try to assemble random
trash into a proposal in hopes that some deity will bless it. The way
you do it is to come up with a coherent idea and then *implement it*.
The deities tend to be a whole lot more impressed by experience with
real software than by committee-generated proposals.
>Albert has a lot of ideas, I have a lot of ideas, everyone has a lot of ideas
>about what we'd like USENET to look like, but who's the guy we send this
>to as a view to changing things? ...
Ask not for whom the compiler waits; it waits for thee. :-) To change
things, first you have to prove that it's worth changing them and that
it's possible to do so. You do that with software, not arguments.
Oh, by the way, look before you leap. It's harder than it looks.
--
Lightweight protocols? TCP/IP *is* | Henry Spencer @ U of Toronto Zoology
lightweight already; just look at OSI. | he...@zoo.toronto.edu utzoo!henry
Path: gmdzi!unido!unidui!math.fu-berlin.de!ox.com!caen!uflorida!gatech!mcnc!duke!wolves!ggw
From: g...@wolves.uucp (Gregory G. Woodbury)
Newsgroups: news.admin,comp.org.eff.talk
Subject: Re: What is Usenet?
Summary: Is there an echo in here?
Keywords: Usenet history anecdotes telephones
Message-ID: <1991Jul18.042054.13265@wolves.uucp>
Date: 18 Jul 91 04:20:54 GMT
References: <1991Jul8.191839.8793@newshost.anu.edu.au> <1991Jul10.030757.6007@wolves.uucp> <1991Jul11.194153.11310@newshost.anu.edu.au>
Followup-To: news.admin,comp.org.eff.talk
Organization: Wolves Den UNIX
Lines: 205
X-Checksum-Snefru: a5b35c75 c5c6a563 5ea8c0e1 7be95fe8
In article <1991Jul11.194153.11...@newshost.anu.edu.au>
cmf...@anu.oz.au (Albert Langer) writes:
>
>In article <1991Jul10.030757.6...@wolves.uucp> g...@wolves.uucp
>(Gregory G. Woodbury) quotes me and writes:
>
>I am saying there is OBVIOUSLY some highly centralized, and democratic
>organization for taking decisions about new newsgroups and EQUALLY
>OBVIOUSLY no similar organization for taking decisions about other
>matters.
This is getting ridiculous. We finally seem to agree that there
is NO formal organization to Usenet. The there is a minimally organized
(minarchist - minimum of hierarchy) community of folks providing a
service that is well ordered and highly cooperative - without any
central authority.
Now Albert says that there is "OBVIOUSLY" a highly centralized
organization that does all this marvelous work for us! Really, where
Albert, where? Please point me to this central organization! I have a
few questions I want to ask whoever is in charge around here! Also, I
want some refunds for the services that they haven't provided!
This is EXACTLY why I say that you need to "make up your mind."
It seems that no matter what anyone says out here to point out to you
the obvious error of your viewpoint (in terms of a "highly centralized,
and democratic organization") you cling tighter and tighter to this
fantasy you have worked up. Who is this organization? The CIA, the
NSA, the FBI, maybe ACM, or even the FCC and/or the BBC? Come on, who
are the central controllers of Usenet?
> If further organization was established it might well use
>the same mechanisms as "news.groups" etc but a glance shows that
>is NOT happening at the moment. There is no contradiction in what
>I am saying and no amount of demands that I make up my mind, with
>or without exclamation marks, will create one.
That same "glance" should show you that there is no central
authority to this organization in the first place.
>Usenet has been very SLOW to adapt, and is still essentially using a
>1970s technology, BECAUSE the lack of any organization that could
>take decisions about upgrading technology RESTRAINS developers to
>maintaining the old services and protocols (as distinct from ensuring
>gateways and conversion for backward compatability and transition,
>which any competent decision making body would specify as part of
>an upgrade). All sorts of good and interesting things are happening,
>but it's no longer at the leading edge of computer communications
>(and it's still at the bleeding edge as regards needing continuous
>maintenance by admin staff).
1970's technology? Are you stuck using pure RS-232 serial
cables and 110 baud modems from point to point? Or are you even older
and using current loop (EIA). Or is UNIX a 1970s technology to you?
Netnews (Version A) was released in 1980. Bnews rewrite was
1985, and Cnews is 1989. NNTP and the current internet are hardly
1970's technology (ethernet is 1984, fiber is even newer!) If all you
see is 1970s technology, I guess things are quite conservative where
ever you are sitting.
In any community, there is a natural conservatism to maintain
compatibility with the elders of the community, but in this particular
situation, this is not nearly as strong a need as you seem to suggest.
There are one or two A news sites hanging on for dear life, but they are
not long for the world, I don't think anyone is actively supporting
Anews anymore.
Another point, a well designed protocol or standard has lots of
life left in it. Simply because it is older than your graduation from
college or secondary school is no reason to consigne it to the dustbin.
I just thought, perhaps you are calling characters of text
displayed on a CRT 1970s technology (which is basically is). The
question then is, what would you replace it with? Are you declaring
that the human penchant and development of written language should be
thrown out the window? Perhaps you want all of us to digitally record
the sound of our voices with our arguments and shipping them out over
the network? (This is foolish, the telephone is much more efficient for
that sort of thing. Besides, how do you propose that the unsophiscated
user edit their voice recordings?)
No, that not it, you want us to ship gigabytes of full-motion
video signals around! Yeah, thats it. But again, how is the
unsophisticated user going to edit this video signal?
Or is it the store and forward nature of the net? You want
interactive direct communications? Try IRC or call some telephone
"party line" and try getting a well considered discussion of the future
of Usenet going. It will be an interesting experience.
>> [...] (this is theoretical - my point is
>>that certain formal networks can AND DO suppress certain topics which
>>the virtual network of Usenet still carries.) In the USA, while some
>>specific sites don't supply certain groups, the "free market" principles
>>still generally apply, and the traffic can be obtained elsewhere if
>>the desire can support the cost.
>
>Fine. If anybody proposes organization for the purpose of suppressing
>certain topics, let them be lashed ten thousand times with wet noodles,
>AS WELL AS failing ignominuously in so patently foolish an endeavour,
>given the world-wide interconnectivity of the PSTN. But why are the wet
>noodles (which are superfluous anyway, given the devastating effect of
>PSTN reality), brought out during discussions of ORGANIZATION rather than
>just during discussions of SUPPRESSION?
Your appeal to the wonder of the PSTN (Public Switched Telephone
Network for those just joing us :-) is just a lot of baloney. The
"public" is, more often than not, a pseudonym for the government. And
when the government controls, then some kind of suppression will occur.
(Gee, that sounds aweful paranoid!) In a free market, the communication
systems will be duplicated somewhat so that competition can uge the
advance of technology and keep the cost to a lower level.
Also, are you prepared to accept you lashes? You are proposing
an organization for your WorldNet that is going to function to suppress
some information. It can't be helped, a formal organization is going to
function in ways that will tend to suppress information that is
detrimental to its own functioning and survival. Only a
non-hierarchical organization (commune or co-operative) is going to be
flexible enough to allow the freedom of expression you seem to really be
wanting.
>Seems we are agreed that IBM and company aren't going to provide
>the sort of INTERACTIVE and "UNCONTROLLED" news and mail that the
>general public should have.
The more general point is that any FORMAL organization is not
going to provide an uncontrolled situation. Tied to that is an
observation that without a formal organization, you are not going to get
universal access - no one is going to pay for that by hiding it in the
overhead.
As for interactive, there seems to be some indecision or muddled
definitions around that word. To you mean to imply that network
communication will be like a "coctail party" where you simply sample the
data stream until you hear/see something you like and step into the
discussion (welcome or not)? Or do you mean some sort of moderated
forum where you listen to the discussion and send comments to the
moderator(s) to be included in the stream (if deemed appropriate). Or
is there some other model you have in mind? Please clarify this
particular point.
It should be noted, in the past few months and years, several
sites and organizations have dropped out of participation in Usenet and
Internet. Among reasons give have been the cost of participation, the
worsening signal-to-noise levels, and the risks of having too many
people having access to personell.
I am willing to bet that as more and more FIDOnet sites and
Public Access sites join the net, and as more and more non-technical
folk become involved, some sites are going to drop into read-only modes
of participation, or just drop out. Additionally, some sites are going
to severely limit the amount and kind of traffic they pass.
This IS a "death of Usenet predicted" kind of bet, with a
difference....
Usenet will actually continue to grow and adapt, it just won't
be the same entity it is now (whatever that is). The current Usenet
type audience will find their own net (perhaps as a subnet of the
WorldNet) and the tech folks will continue to talk among themselves
without the great masses reading over their shoulder and sticking their
fingers in unwelcomly.
I more likely see a large number of different nets, any and all
of which COULD exchange data with any other net, each with a different
emphasis and clientele.
I am disappointed that you ignored my comments about the reality
of the technical infrastructure underlying the PSTN. Once again you
have held up the telephone as the ideal of how you want to interact, and
claim that it is done without operators. Once again, I point out that
that is just simply a fantasy.
A lot of technology, operated by lots of people, make it appear
that the telephone is one bit univerally connected network, with
idiot-proof consumer-grade controls. Additionally, there is the fact
that it has taken the telco's 60 years to get dial telephones nearly
everywhere. (There was quite a bit of publicity about the recent
de-installation of the last non-direct-dial exchanges in the USA.) It
is probably going to take the telcos another 10 or so to get universal
international dialling to the point where the user doesn't need to get
someone to assist or look it up in some manual! (Don't even try to tell
me that a busy executive trying to reach the manager of a field office
in some remote corner of the world [at least to her] is going to take
the time currently to dig out a phone book and lookup all the numbers
and dial them herself. The executive is going to tell an assistant or
secretary to "get the manager in Bumphuc on the telephone for me.")
This is hardly and operatorless situation. The skill to do what
is necessary to communicate has simply been moved around.
The "generic information consumer" of your WorldNet is not going
to care to know how his machine connects to the net. Nor is he going to
care how the message he types and polishes (with whatever tools his
machine provides) gets to the machine of its intended recipient. He
will expect that the "network company will take care of that." You can
bet that the "network company" will have a number of people who worry
about machine connection technologies, and mail routing, and all the
stuff that we (the pioneers) have to deal with on our own today.
--
Gregory G. Woodbury @ The Wolves Den UNIX, Durham NC
UUCP: ...dukcds!wolves!ggw ...duke!wolves!ggw [use the maps!]
Domain: g...@cds.duke.edu ggw%wol...@duke.cs.duke.edu
[The line eater is a boojum snark! ] <standard disclaimers apply>
Path: gmdzi!unido!math.fu-berlin.de!ox.com!caen!zaphod.mps.ohio-state.edu!swrinde!cs.utexas.edu!utgpu!cunews!micor!latour!ecicrl!clewis
From: cle...@ferret.ocunix.on.ca (Chris Lewis)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access (Was: Re: What is Usenet?
Message-ID: <2319@ecicrl.ocunix.on.ca>
Date: 18 Jul 91 06:44:42 GMT
References: <1991Jul14.185420.3553@newshost.anu.edu.au> <TV26ZI9@taronga.hackercorp.com>
Organization: Elegant Communications Inc., Ottawa, Canada
Lines: 78
In article <TV26...@taronga.hackercorp.com> pe...@taronga.hackercorp.com (Peter da Silva) writes:
>cmf...@anu.oz.au (Albert Langer) writes:
>> If you have achieved it already, even though at only two small sites and
>> with a skilled C programmer as sysadmin, the possibility of doing it
>> more widely with just current RFC based software must be taken seriously.
>Thanks for the kudos, but I'm just running straight C-news under UNIX.
>I haven't written any special code, just used what's out there on the net.
Even B-news is capable of operating with virtually zero maintenance.
Stock B 2.11.19 works very well, just needs an incoming throttle to
make it bullet proof from floods. The key to success is knowing a few
of the unpublished ins and outs of news, which come thru experience -
there are lots of useful configuration hints that I've not seen documented
anywhere, but I've picked up from experience (B-news since '84, C-news
since '86). I doubt that I put in any more than 5 minutes a month
on my B-news - and most of that is stuff that doesn't really need to be done.
Most of the systems I've set up (probably a few dozen) can and have run for
months without any interference at all. The biggest mistake people make is
running too close to their disk capacity limits. The vast majority of
news-related problems I encounter are modem lockups.
On the other hand, if I had to reinstall news on this system, I'd probably
install cnews instead of bnews.
>> Is there any chance of you packing up a complete copy of your configuration,
>> including ALL the relevant files as well any software you have customized,
>> so I can have a look at how you did this and try it myself?
>Just get a copy of:
> cnews
> elm
or mush. Mush is particularly problem-free.
> smail 2.5
Plus Ron Zeeff's lmail package (with our bug fixes) if you need mail-to-pipe
and mail-to-file.
> pathalias
You forgot a news reader (rn, nn etc.)
>To build these, you'll also need
> unshar
I'd recommend Rich Salz's version (it's non-/bin/sh based).
> patch
At least patch level 8.
>For map unpacking you'll have to wait. I'll see what Karl has set up at
>sugar... I didn't do that part.
Both uuhosts and unpackmaps are "set-and-forget". The advantage to unpackmaps
is that it does more of the work to get your paths file setup. Another
advantage is that I wrote it. ;-) Uuhosts leaves a bit more (including the
pathalias invocation) as an exercise to the reader. I've never had to diddle
with either of them once they're config'd. I don't recommend rolling your
own map unpacker until you completely understand the security issues
(both uuhosts and unpackmaps are pretty secure. Most quick hacks aren't.
There are many other unpackers out there, but since I've not audited them,
so I won't recommend them). uuhosts is available from c.s.u archives,
unpackmaps from c.s.m archives and is, I believe, more recent.
You'll also need compress 4.
Currently available sources from the net are quite solid, and need
very little or no maintenance.
--
Chris Lewis; UUCP: ...!{cunews,uunet,latour}!ecicrl!clewis;
DOMAIN: cle...@ferret.ocunix.on.ca; Phone: Canada 416 832-0541
Psroff info: psroff-requ...@ferret.ocunix.on.ca
Ferret mailing list: ferret-requ...@ferret.ocunix.on.ca
Path: gmdzi!unido!fauern!math.fu-berlin.de!ox.com!yale.edu!think.com!samsung!olivea!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: Usenet User's Association?
Message-ID: <1991Jul18.173645.23873@newshost.anu.edu.au>
Date: 18 Jul 91 17:36:45 GMT
References: <1991Jul16.210123.7517@newshost.anu.edu.au> <2463@uqcspe.cs.uq.oz.au> <1991Jul17.052648.24883@zoo.toronto.edu>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 124
In article <1991Jul17.052648.24...@zoo.toronto.edu> he...@zoo.toronto.edu
(Henry Spencer) writes:
>I'll let you in on one of the deep, dark secrets of the Internet world.
>The way to establish a New World Order is not to try to assemble random
>trash into a proposal in hopes that some deity will bless it. The way
>you do it is to come up with a coherent idea and then *implement it*.
>The deities tend to be a whole lot more impressed by experience with
>real software than by committee-generated proposals.
Hmm, Rhys specifically asked about whether there were any task forces
and where their minutes could be obtained. A not so deep and dark
secret of the Internet world is that they do exist, that their minutes
are published as RFCs and FYIs and ietf drafts etc and these are available
at many ftp sites. The Internet does have funding, and staff and a much
tighter organization than Usenet (currently still responsible to the
United States Department of Defence, which is not generally reputed to
be a hotbed of "minanarchy").
I can sympathize with some of the paranoia about organization, when one
sees examples such as Rhys drifting straight into a $5 levy on all students
(why students?!?) to pay for full-time staff to do jobs that could and
should be done by volunteer moderators. But that is ground for caution
about how one should go about getting more organized. It should not
justify ignoring the consequences of failing to do so.
The technical environment that Usenet depends on (for technology, not
just some convenient communications channels) was not established,
and could not be established without a GREAT DEAL more organization
than Usenet itself possesses. The result of course is that the
destiny and evolution of Usenet is greatly influenced by decisions
made elsewhere, by people who very tolerantly allow the minarchists to
play in their sandpits.
I agree the general "technical culture" in the Internet has
favoured the "American" approach of evolving standards afterwards,
on the basis of experience gained from what has been proved workable
in practice, rather than the "European" approach of adopting standards
before there are any implementations. But even that has become
something of a "fiction".
In practice, while still uttering rugged macho platitudes in their
spare time, about how unimpressed they are with committee-generated
proposals, the "deities" have had little choice but to go along
with such proposals, since they are mandated by U.S. Government
regulations (GOSIP), which are binding on the DoD and DARPA as
much as any other agency.
Sure they are dragging their feet, and one sees very little OSI
and X.400 or X.500 implementation in the U.S.A. compared with
widespread use of X.25 instead of TCP/IP for equivalent European
networks and slightly less widespread use of X.400 instead of
RFC based mail in Europe.
But the simple reality is that the Internet is committed to OSI
migration, whether rightly or wrongly. (I happen to think rightly,
but that is irrelevant).
Usenet can of course limp along with the old technology until
somebody ELSE demonstrates a fully implemented working model
of the latest stuff, but it seems kind of a shame when the
RFCs used to trail behind Usenet.
Once upon a time, the philosophy of "implement first, ask for
comments later" (and call the request for comments a kind of
"standard") was part of being at the leading edge of technical
development, so proposals just had to catch up later.
Today, the same philosophy has exactly opposite implications.
It means trailing behind with "improvements" to fundamentally
obsolete technology, while the leading edge is working from
specs that were published SEVERAL YEARS AGO but have been
given very little attention here.
We should be ADVANCING on X.400 (e.g. defining how to implement
if for news as well as just mail, and extending that to more
sophisticated groupware applications), not lagging behind it.
>Ask not for whom the compiler waits; it waits for thee. :-) To change
>things, first you have to prove that it's worth changing them and that
>it's possible to do so. You do that with software, not arguments.
I'm sure that, and the similar advice from Brad Templeton and Joe Buck,
accurately reflects the current situation in Usenet.
To some extent I'll even take that advice, because I know it will be
more productive to produce some software (or at least some RFCs :-) than
to argue very much more here about why it's required.
But consider this. That advice flies DIAMETRICALLY in opposition to
EVERYTHING being taught to the next generation of software engineers,
who are being told, over and over, that requirements come first, then
specification, then design, then coding and testing (until we figure
out how to automate those), and repeat the cycle for incremental
builds.
Sure there are enough people around who don't hold with that
new fangled stuff and sure some of them can churn out brilliant code
faster than anybody could write the specs for it. But when it isn't
a hobby, most people would rather just use properly engineered
software products than spend their time monitoring news groups for
advice on how to fiddle with things to keep them working.
"Prove things with software, not arguments" is fine for a network
of programmers - which Usenet used to be. But not allowing users
to define the requirements has simple and obvious consequences -
like the news and mail tools are designed to be operated by
programmers, not by computer novices - and the nodes need sysadmins
to keep them operating.
Changing that, whether for Usenet or for some other network that
provides access for the general public, MUST involve requirements
definition before coding, because the requirements for general
public access are different from the requirements of a programmer's
network.
>Oh, by the way, look before you leap. It's harder than it looks.
Yep. (And it looks hard too :-)
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!math.fu-berlin.de!ox.com!caen!zaphod.mps.ohio-state.edu!rpi!news-server.csri.toronto.edu!utzoo!henry
From: he...@zoo.toronto.edu (Henry Spencer)
Newsgroups: news.admin
Subject: Re: Usenet User's Association?
Message-ID: <1991Jul19.170716.9012@zoo.toronto.edu>
Date: 19 Jul 91 17:07:16 GMT
References: <1991Jul16.210123.7517@newshost.anu.edu.au> <2463@uqcspe.cs.uq.oz.au> <1991Jul17.052648.24883@zoo.toronto.edu> <1991Jul18.173645.23873@newshost.anu.edu.au>
Organization: U of Toronto Zoology
Lines: 77
In article <1991Jul18.173645.23...@newshost.anu.edu.au> cmf...@anu.oz.au (Albert Langer) writes:
>Hmm, Rhys specifically asked about whether there were any task forces
>and where their minutes could be obtained. A not so deep and dark
>secret of the Internet world is that they do exist...
However, he was asking about ones working on *Usenet*, not the Internet.
And he wanted to know about places where he could send his ideas to be
implemented, not places that would bless already-implemented ideas.
>The technical environment that Usenet depends on (for technology, not
>just some convenient communications channels) was not established,
>and could not be established without a GREAT DEAL more organization
>than Usenet itself possesses...
Very true. It took a lot of organization to set up the telephone network.
Oh, are you under the impression that Usenet vitally depends on the Internet?
I hate to shatter anyone's fond preconceptions, but Usenet existed without
the Internet and would continue to exist (albeit with a lot of pain and some
changes) if the Internet vanished. For Usenet, the Internet *is* just some
convenient communications channels. Usenet's technology was not invented
or established by the Internet.
>But the simple reality is that the Internet is committed to OSI
>migration, whether rightly or wrongly...
I don't understand what this has to do with Usenet. There is no OSI/ISO
standard that even vaguely addresses news. (News is not mail, despite
some overenthusiastic claims that X.400 could cover news too.) Usenet
neither knows nor cares how its data gets from point A to point B. When
it gets there, it is processed by software that has little or nothing
to do with most RFCs and no detectable relationship to OSI migration.
>Usenet can of course limp along with the old technology until
>somebody ELSE demonstrates a fully implemented working model
>of the latest stuff...
Usenet will continue to function happily with the old technology even if
somebody demonstrates the latest stuff, unless the latest stuff offers
some real advantages. Backward compatibility is immensely more powerful
in an anarchy like Usenet than in even a loosely-administered organization
like the Internet. Besides, there is nothing much wrong with Usenet's
current technology and no need to replace it (*improve* it, yes, but
the foundations are sound).
>... while the leading edge is working from
>specs that were published SEVERAL YEARS AGO but have been
>given very little attention here.
The leading edge of *what*? Certainly not technical progress. I detect
little or no technical merit in most of those several-year-old specs.
The leading edge of Political Correctness, maybe.
>We should be ADVANCING on X.400 ...
With rifles and flamethrowers.
>>Ask not for whom the compiler waits; it waits for thee. ...
>But consider this. That advice flies DIAMETRICALLY in opposition to
>EVERYTHING being taught to the next generation of software engineers,
No, it flies diametrically in opposition to everything taught to the
*previous* generation of software alchemists.
>who are being told, over and over, that requirements come first, then
>specification, then design, then coding and testing (until we figure
>out how to automate those), and repeat the cycle for incremental
>builds.
The next generation of software engineers are learning, at least in places
that have their heads screwed on right, that requirements, specifications,
and design are almost predictably wrong, and that the crucial step is to
get prototypes going quickly so that you can discover the mistakes before
casting them in concrete. Pity OSI/ISO doesn't know this.
--
Lightweight protocols? TCP/IP *is* | Henry Spencer @ U of Toronto Zoology
lightweight already; just look at OSI. | he...@zoo.toronto.edu utzoo!henry
Path: gmdzi!unido!mcsun!uunet!zaphod.mps.ohio-state.edu!think.com!snorkelwacker.mit.edu!world!geoff
From: ge...@world.std.com (Geoff Collyer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Keywords: Usenet history anecdotes telephones
Message-ID: <1991Jul21.072057.12515@world.std.com>
Date: 21 Jul 91 07:20:57 GMT
References: <1991Jul10.030757.6007@wolves.uucp>
<1991Jul11.194153.11310@newshost.anu.edu.au>
<1991Jul18.042054.13265@wolves.uucp>
<1991Jul21.011400.10117@newshost.anu.edu.au>
Organization: Software Tool & Die Netnews Research Center
Lines: 62
someone unattributed:
>> Netnews (Version A) was released in 1980. Bnews rewrite was
>>1985, and Cnews is 1989.[...]
These dates are off. B News 2.3 was available in 1982, when I started
running it, which means that B News had been out for a while before that.
C News was written in 1985, initially released in the fall of 1987, with
the "production release" following in the summer of 1989. Plenty of folks
happily ran the initial release, so the name "production release" may be
somewhat misleading; it means that we hadn't been entirely happy with the
initial release, which was marked "Alpha". (We have a stricter view of
what constitutes production-quality software than some people.)
Albert Langer:
>Netnews A, B and C are all based on the same fundamental design, which
>is based on ARPA mail formats dating back to RFC561 in September 1973.
Wrong. A News used a message format that had absolutely nothing to do
with ARPAnet standards. Here is the description from news(5):
Aarticle-ID
newsgroups
path
date
title
Body of article
Note that the first character is a literal "A" and note the utter lack of
header keywords. This was a very sensible format, but was probably doomed
because it looked nothing like mail and was inextensible (though
extensibility of news headers is vastly overrated). Note that RFC 561
suggests 822-like headers with header keywords.
In addition, A News was one big program, so I wouldn't call it the same
fundamental design as B News.
B News used a subset of RFC 822's message format, which is extensible,
but which has its share of problems, including excess generality. C News
uses the same format as B News because I didn't want to write a protocol
converter and the B News format seemed tolerable if not wonderful.
>RFC850 and the current RFC1036 are more extensible so that implementations
>of them can expect a longer life, however they have NOT been designed with
>the same care devoted to future extensions as X.400.
(a) Expending vast caring is no guarantee of a worthwhile result. Let us
judge the results and not the tender loving care that produced them.
(b) X.400 is a joke: an unnecessary, *binary* format for even textual
mail, apparently designed by European phone companies to maximise their
revenue by inserting bulk. The addressing is ridiculous and the less
said about X.500, the better. The specifications are too large to carry,
let alone read. This we should emulate?
If you really think the ISO protocols are so superior to this ancient
cruft, go ahead and implement netnews over X.400; just don't expect
anyone to actually use it. And if you can't implement it, don't come
whining to Usenet at large about how you have an inalienable right to
have someone else implement your proposal. (Or are you trying to be this
month's mathew@mantis?)
--
Geoff Collyer world.std.com!geoff, uunet.uu.net!geoff
Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Keywords: Usenet history anecdotes telephones
Message-ID: <1991Jul22.072111.17743@newshost.anu.edu.au>
Date: 22 Jul 91 07:21:11 GMT
References: <1991Jul18.042054.13265@wolves.uucp> <1991Jul21.011400.10117@newshost.anu.edu.au> <1991Jul21.072057.12515@world.std.com>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 180
In article <1991Jul21.072057.12...@world.std.com> ge...@world.std.com
(Geoff Collyer) quotes me and writes:
>>Netnews A, B and C are all based on the same fundamental design, which
>>is based on ARPA mail formats dating back to RFC561 in September 1973.
>
>Wrong. A News used a message format that had absolutely nothing to do
>with ARPAnet standards. Here is the description from news(5):
>
> Aarticle-ID
> newsgroups
> path
> date
> title
> Body of article
>
>Note that the first character is a literal "A" and note the utter lack of
>header keywords. This was a very sensible format, but was probably doomed
>because it looked nothing like mail and was inextensible [...]
I agree that A news departed from the mail format and I mentioned that
in a response to Henry where it was relevant. Your statement that it
was probably doomed because it looked nothing like mail and was
inextensible confirms the point I was making there.
In the present context, I stand by the statement that A, B and C are
all based on the same fundamental design. In the case of B and C they
are even based on the same specifications. For all 3, the design is
based on broadcast by flooding of plain ASCII text files which have
a series of header lines followed by a body. The headers include
a message ID and path to control duplicates, newsgroups to define
which subnet(s) they are broadcast to and a date and title for similar
reasons to ordinary mail (plus automatic expiry by date).
The (important) differences in format of header lines between A
and B are differences in specification within the same fundamental
design. X.400 on the other hand involves a quite different design,
despite also defining a means for store and forward broadcasting
of text messages.
>(b) X.400 is a joke: an unnecessary, *binary* format for even textual
>mail, apparently designed by European phone companies to maximise their
>revenue by inserting bulk.
It is an indisputable matter of fact that the binary encoding rules
for ASN.1 occupy LESS space than the keyword based header lines of
RFC1036, therefore:
1) They could not POSSIBLY have been "apparently designed by
European phone companies to maximise their revenue by inserting
bulk."
2) You are demonstrably ignorant of the subject you are discussing,
but despite this (or perhaps because of it) you quite gratuitously
assert devious and vicious motives to other people instead of
discussing technical issues on their merits.
Merely because you use an encoding system that is bulkier
than the one used by X.400, neither I nor any rational person would
even think of remarking that your software is "apparently designed
by American phone companies to maximise their revenue by inserting
bulk."
An ASN.1 binary encoding typically has an overhead of two bytes -
one for the tag (instead of an entire keyword, plus colon and space),
and one for the length (same as the one required
for a newline, but permitting the use of arbitrary 8 bit data and
not requiring ridiculous kludges for continuation lines etc).
Among the many advantages of this encoding are:
1. No lexical analysers or parsers are required to extract fields
- you always know the type of each field and how long it is
before you have to process it or skip it. Consequently buffers
can be easily handled on input (and also for output). The result
is GREATER implementation efficiency than when parsing header
keywords. (Even though this is a relatively unimportant side
effect of a notation primarily intended to permit comprehension
and implementation of more complex designs).
2. Arbitrary nested structures can be represented (and skipped over)
efficiently.
3. As a result of the above, a "Presentation Layer" can make use
of its knowledge of the application syntax to encode each field
efficiently according to the data it contains. For example tags
indicating the language of body text can be used to encode with an
optimal compression algorithm acceptable to both of two
neighbouring sites and header fields that are repeated between
articles can be replaced with pointers. Thus as well as the
inherently less bulky data than RFC1036 it is also feasible to
get much more compression than the 16 bit LZW applied to header
and body text indiscriminately that is typical of Usenet news
transmission.
4. A side benefit is that since the encoding isn't just human
readable text, nobody would be silly enough to design a
reader that doesn't interpret each field and display it
appropriately. While it isn't the fault of RFC1036 there is
an amazing tendency to use the availability of readable
keywords as an excuse to just display them with an ordinary
pager and enter them from a draft header through an ordinary
editor. (For debugging and maintenance purposes of course
special software must be used instead of just a pager, which
is about as inconvenient as typing "asn" instead of "more").
> The addressing is ridiculous and the less
>said about X.500, the better.
Since you obviously know nothing about the matter, your advice
about how much you should say on the subject is sound. Pity
you did not follow it.
>The specifications are too large to carry,
>let alone read. This we should emulate?
I found them intimidating too. But I managed to read them.
If I had found myself unable to do so, whether due to lack of
time or lack of the necessary background information and skills,
I hope I would not have been either so arrogant or else so insecure
as to claim that something I had been unable to read was therefore worthless.
I find C code as difficult, and Chinese even more difficult to read
than you apparently find ISO/CCITT standards. Should I describe
your C code, or some Chinese manuscript as "a joke" when I have
not read it?
BTW, when I say I find C code "as difficult", I mean I can plough
my way through it if I have to, but even the simplest program always
looks alien to me, whereas a competent C programmer would find it
perfectly NATURAL for example, to open a file as a side effect of
not printing an error message about being unable to open it.
Likewise I am quite sure that you could read an ISO spec if you had to.
That doesn't oblige you to do so - but don't pretend it's the fault
of the specs if you choose not to.
>If you really think the ISO protocols are so superior to this ancient
>cruft, go ahead and implement netnews over X.400; just don't expect
>anyone to actually use it.
Whether you like it or not, before embarking on such an ambitious
project I shall naturally make inquiries and see who else is interested
and whether anybody who actually knows what they are talking
about has some sound criticisms which might show it isn't a good
idea - and of course I will first check whether it would be likely
to be used.
However I won't be asking you until you HAVE read the specs.
Please feel free not to bother sending your opinion until then.
>And if you can't implement it, don't come
>whining to Usenet at large about how you have an inalienable right to
>have someone else implement your proposal. (Or are you trying to be this
>month's mathew@mantis?)
I said nothing whatever that could conceivably be interpreted as
asserting an inalienable right to have someone else implement my
proposal. Nor have I been "whining" about anything. The only
"whining" I have noticed in this discussion so far is your
remarks about ISO specs being too heavy for you - as though
you have some "inalienable right" to have specs written in ways
that suit you.
Your resort to childish rudeness would be unnecessary if you
had anything of interest to say.
Your contribution to the net in developing C news is valued.
I would be very happy to treat your opinions on the practical
problems of news design and implementation with the respect
that is due to somebody who obviously has far more practical
experience and skills in the matter than I do.
I hope you make it possible to do so by adopting a different
tone, and sharing the knowledge you DO have, instead of making
snide remarks about things you DON'T know.
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!mcsun!uunet!elroy.jpl.nasa.gov!usc!samsung!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <1991Jul23.122947.25080@newshost.anu.edu.au>
Date: 23 Jul 91 12:29:47 GMT
References: <1991Jul14.185420.3553@newshost.anu.edu.au> <TV26ZI9@taronga.hackercorp.com> <2319@ecicrl.ocunix.on.ca>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 57
In article <2...@ecicrl.ocunix.on.ca> cle...@ferret.ocunix.on.ca
(Chris Lewis) writes:
>Even B-news is capable of operating with virtually zero maintenance.
>Stock B 2.11.19 works very well, just needs an incoming throttle to
>make it bullet proof from floods. The key to success is knowing a few
>of the unpublished ins and outs of news, which come thru experience -
>there are lots of useful configuration hints that I've not seen documented
>anywhere, but I've picked up from experience (B-news since '84, C-news
>since '86). I doubt that I put in any more than 5 minutes a month
>on my B-news - and most of that is stuff that doesn't really need to be done.
>Most of the systems I've set up (probably a few dozen) can and have run for
>months without any interference at all. The biggest mistake people make is
>running too close to their disk capacity limits. The vast majority of
>news-related problems I encounter are modem lockups.
>
>On the other hand, if I had to reinstall news on this system, I'd probably
>install cnews instead of bnews.
Well, that's at least ONE corroborating witness for Peter's 10 seconds per
day. I'm still skeptical, but I've made a note of the software both of you
have recommended (thanks) and will just have to see how it goes.
I'll be looking to see how hard it is to design something for a fairly
naive user to automatically install and then maintain a news system with
the recommended software. (BTW if you can publish any of those "useful
configuration hints" that would be great - maybe a "hint list" could
be included with future software releases and/or referred to in news.*
FAQ answers.)
>Currently available sources from the net are quite solid, and need
>very little or no maintenance.
Well, I have to setup a system based on currently available sources
anyway, so I'll keep careful records of maintenance time and see.
But I still have a suspicion that the lack of loud cries to the
contrary is related to the "Emperor's New Clothes" syndrome. I mean
who wants to admit that they spend 10 or 100 times as much time
maintaining their news system as Peter and Chris do :-)
I'm convinced (from experience) that just leaving the damn machine
alone does WONDERS for stability, but the traffic in news.* and comp.mail.*
suggests there are quite a few maintenance and operations problems rather
than just installation (although those are the biggest).
Not long ago somebody was explaining how much time was involved in providing
a feed and how it therefore wasn't unreasonable to refuse feeds unless
to related organizations etc.
Don't forget the concept I am putting forward for general public access
implies (naive) users as sites providing feeds to each other as well,
not just being leaf nodes.
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!math.fu-berlin.de!ox.com!caen!uakari.primate.wisc.edu!samsung!uunet!snorkelwacker.mit.edu!world!geoff
From: ge...@world.std.com (Geoff Collyer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Keywords: Usenet history anecdotes telephones
Message-ID: <1991Jul23.143645.2002@world.std.com>
Date: 23 Jul 91 14:36:45 GMT
References: <1991Jul18.042054.13265@wolves.uucp>
<1991Jul21.011400.10117@newshost.anu.edu.au>
<1991Jul21.072057.12515@world.std.com>
<1991Jul22.072111.17743@newshost.anu.edu.au>
Organization: The World @ Software Tool & Die
Lines: 15
_ 23 July 1991
Dear Mr. Langer:
Thank you for submitting your application for the job of
mathew@mantis.
You certainly have demonstrated an impressive set of skills: inability
to read or understand simple written material; inability to reason
clearly; extremely weak memory; inaccurate remote psychoanalysis; and
ability to flame pointlessly and at great length.
Goodbye, Mr. Langer.
--
Geoff Collyer world.std.com!geoff, uunet.uu.net!geoff
Path: gmdzi!unido!math.fu-berlin.de!ox.com!yale.edu!qt.cs.utexas.edu!zaphod.mps.ohio-state.edu!rpi!news-server.csri.toronto.edu!smoke.cs.toronto.edu!moraes
From: mor...@cs.toronto.edu (Mark Moraes)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <91Jul24.112416edt.1530@smoke.cs.toronto.edu>
Date: 24 Jul 91 15:24:27 GMT
References: <1991Jul18.042054.13265@wolves.uucp> <1991Jul21.011400.10117@newshost.anu.edu.au> <1991Jul21.072057.12515@world.std.com> <1991Jul22.072111.17743@newshost.anu.edu.au>
Organization: Department of Computer Science, University of Toronto
Lines: 36
Albert,
Those of us interested in viable[%] implementations of communications
standards find short, comprehensible specs appealing. You see, some
people do some of this implementation as a spare time activity because
they find it interesting and/or challenging. Sometimes, they even
convince their employers that part of what they do is useful work,
though that's usually after they spend many evenings implementing it.
Perhaps you now understand why some people don't get excited by X.?00.
It'll take too long to read the standards, let alone implement them.
A significant problem in the design of certain standards it that
implementing them is hard -- question is, which is superior: a robust
available implementation of a simple non-standard that suffices for
most people's needs, or a fragile, real-soon-now implementation of an
incredibly complex standard that attempts to please everyone?
Pointing people at a working implementation is the way to prove that
the specs you espouse may be superior to existing specs; hysterical
handwaving in news.admin may feel very satisfying, but it isn't very
convincing.
Again, those of us interested in viable communications have discovered
that it's very difficult to take a completely incompatible standard
and force it down everyone's throats. Backward compatibility may be
confining, but it's necessary if you want to talk to a reasonable
number of people.
Your statement that ASN.1 encoded data is easy to parse brightened up
my morning (again!). Thank you.
Mark.
% To me, "viable" means I should be able to get the programs without
too much hassle, and run them on relatively small machines, where I
also have to get other "real" work done.
Path: gmdzi!unido!mcsun!uunet!zaphod.mps.ohio-state.edu!samsung!news.cs.indiana.edu!purdue!spaf
From: s...@cs.purdue.EDU (Gene Spafford)
Newsgroups: news.announce.newusers,news.admin
Subject: What is Usenet?
Message-ID: <15396@ector.cs.purdue.edu>
Date: 25 Jul 91 23:15:52 GMT
Expires: 23 Oct 91 23:15:52 GMT
Followup-To: news.announce.newusers
Organization: Dept. of Computer Sciences, Purdue Univ.
Lines: 265
Approved: s...@cs.purdue.EDU
Supersedes: <14...@ector.cs.purdue.edu>
Original from: c...@count.tct.com (Chip Salzenberg)
[Most recent change: 23 Jul 1991 by s...@cs.purdue.edu (Gene Spafford)]
The first thing to understand about Usenet is that it is widely
misunderstood. Every day on Usenet, the "blind men and the elephant"
phenomenon is evident, in spades. In the opinion of the author, more
flame wars arise because of a lack of understanding of the nature of
Usenet than from any other source. And consider that such flame wars
arise, of necessity, among people who are on Usenet. Imagine, then,
how poorly understood Usenet must be by those outside!
Any essay on the nature of Usenet cannot ignore the erroneous
impressions held by many Usenet users. Therefore, this article will
treat falsehoods first. Keep reading for truth. (Beauty, alas, is
not relevant to Usenet.)
WHAT USENET IS NOT
------------------
1. Usenet is not an organization.
Usenet has no central authority. In fact, it has no central
anything. There is a vague notion of "upstream" and "downstream"
related to the direction of high-volume news flow. It follows
that, to the extent that "upstream" sites decide what traffic
they will carry for their "downstream" neighbors, that "upstream"
sites have some influence on their neighbors. But such influence
is usually easy to circumvent, and heavy-handed manipulation
typically results in a backlash of resentment.
2. Usenet is not a democracy.
A democracy can be loosely defined as "government of the people,
by the people, for the people." However, as explained above,
Usenet is not an organization, and only an organization can be run
as a democracy. Even a democracy must be organized, for if it
lacks a means of enforcing the peoples' wishes, then it may as
well not exist.
Some people wish that Usenet were a democracy. Many people
pretend that it is. Both groups are sadly deluded.
3. Usenet is not fair.
After all, who shall decide what's fair? For that matter, if
someone is behaving unfairly, who's going to stop him? Neither
you nor I, that's certain.
4. Usenet is not a right.
Some people misunderstand their local right of "freedom of speech"
to mean that they have a legal right to use others' computers to
say what they wish in whatever way they wish, and the owners of
said computers have no right to stop them.
Those people are wrong. Freedom of speech also means freedom not
to speak; if I choose not to use my computer to aid your speech,
that is my right. Freedom of the press belongs to those who own
one.
5. Usenet is not a public utility.
Some Usenet sites are publicly funded or subsidized. Most of
them, by plain count, are not. There is no government monopoly
on Usenet, and little or no control.
6. Usenet is not a commercial network.
Many Usenet sites are academic or government organizations; in
fact, Usenet originated in academia. Therefore, there is a Usenet
custom of keeping commercial traffic to a minimum. If such
commercial traffic is generally considered worth carrying, then it
may be grudgingly tolerated. Even so, it is usually separated
somehow from non-commercial traffic; see "comp.newprod."
7. Usenet is not the Internet.
The Internet is a wide-ranging network, parts of which are
subsidized by various governments. The Internet carries many
kinds of traffic; Usenet is only one of them. And the Internet is
only one of the various networks carrying Usenet traffic.
8. Usenet is not a UUCP network.
UUCP is a protocol (some might say "protocol suite," but that's a
technical point) for sending data over point-to-point connections,
typically using dialup modems. Usenet is only one of the various
kinds of traffic carried via UUCP, and UUCP is only one of the
various transports carrying Usenet traffic.
9. Usenet is not a UNIX network, nor even an ASCII network. It is
also most certainly not just an American network.
Don't assume that everyone is using "rn" on a UNIX machine. There
are Vaxen running VMS, IBM mainframes, Amigas, and MS-DOS PCs
reading and posting to Usenet. And, yes, some of them use
(shudder) EBCDIC. Ignore them if you like, but they're out there.
Some sites use special character sets for non-English postings,
too, and even if they use the same character set, realize that
your words might mean different things in other cultures.
10. Usenet is not software.
There are dozens of software packages used at various sites to
transport and read Usenet articles. So no one program or package
can be called "the Usenet software."
Software designed to support Usenet traffic can be (and is) used
for other kinds of communication, usually without risk of mixing
the two. Such private communication networks are typically kept
distinct from Usenet by the invention of newsgroup names different
from the universally-recognized ones.
Well, enough negativity.
WHAT USENET IS
--------------
Usenet is the set of machines that exchange articles tagged with one
or more universally-recognized labels, called "newsgroups" (or
"groups" for short).
(Note that the term "newsgroup" is correct, while "area," "base,"
"board," "bboard," "conference," "round table," "SIG," etc. are
incorrect. If you want to be understood, be accurate.)
DIVERSITY
---------
If the above definition of Usenet sounds vague, that's because it is.
It is almost impossible to generalize over all Usenet sites in any
non-trivial way. Usenet encompasses government agencies, large
universities, high schools, businesses of all sizes, home computers of
all descriptions, etc, etc.
CONTROL
-------
Every administrator controls his own site. No one has any real
control over any site but his own.
The administrator gets his power from the owner of the system he
administers. As long as the owner is happy with the job the
administrator is doing, he can do whatever he pleases, up to and
including cutting off Usenet entirely. Them's the breaks.
PROPAGATION
-----------
In the old days, when UUCP over long-distance dialup lines was the
dominant means of article transmission, a few well-connected sites had
real influence in determining which newsgroups would be carried where.
Those sites called themselves "the backbone."
But things have changed. Nowadays, even the smallest Internet site
has connectivity the likes of which the backbone admin of yesteryear
could only dream. In addition, in the U.S., the advent of cheaper
long-distance calls and high-speed modems has made long-distance
Usenet feeds thinkable for smaller companies. There is only one
pre-eminent UUCP transport site today in the U.S., namely UUNET. But
UUNET isn't a player in the propagation wars, because it never refuses
any traffic -- it gets paid by the minute, after all; and besides, to
refuse based on content would jeopardize its legal status as an
enhanced service provider.
All of the above applies to the U.S. In Europe, different cost
structures favored the creation of strictly controlled hierarchical
organizations with central registries. This is all very unlike the
traditional mode of U.S. sites (pick a name, get the software, get a
feed, you're on). Europe's "benign monopolies," long uncontested, now
face competition from looser organizations patterned after the U.S.
model.
NEWSGROUP CREATION
------------------
As discussed above, Usenet is not a democracy. Nevertheless, the
current most popular way to create a new newsgroup involves a "vote"
to determine popular support for (and opposition to) a proposed
newsgroup. The document that describes this procedure is entitled
"How To Create A New Newsgroup." Its common name, however, is "the
guidelines."
If you follow the guidelines, it is probable that your group will be
created and will be widely propagated.
HOWEVER: Because of the nature of Usenet, there is no way for any user
to enforce the results of a newsgroup vote (or any other decision, for
that matter). Therefore, for your new newsgroup to be propagated
widely, you must not only follow the letter of the guidelines; you
must also follow its spirit. And you must not allow even a whiff of
shady dealings or dirty tricks to mar the vote.
So, you may ask: How is a new user supposed to know anything about the
"spirit" of the guidelines? Obviously, he can't. This fact leads
inexorably to the following recommendation:
>> If you are a new user, don't try to create a new newsgroup. <<
If you have a good newsgroup idea, then read the "news.groups"
newsgroup for a while (six months, at least) to find out how things
work. If you're too impatient to wait six months, then you really
need to learn; read "news.groups" for a year instead. If you just
can't wait, find a Usenet old hand to run the vote for you.
Readers may think this advice unnecessarily strict. Ignore it at your
peril. It is embarrassing to speak before learning. It is foolish to
jump into a society you don't understand with your mouth open. And it
is futile to try to force your will on people who can tune you out
with the press of a key.
IF YOU ARE UNHAPPY...
---------------------
Property rights being what they are, there is no higher authority on
Usenet than the people who own the machines on which Usenet traffic is
carried. If the owner of the machine you use says, "We will not carry
alt.sex on this machine," and you are not happy with that order, you
have no Usenet recourse. What can we outsiders do, after all?
That doesn't mean you are without options. Depending on the nature of
your site, you may have some internal political recourse. Or you
might find external pressure helpful. Or, with a minimal investment,
you can get a feed of your own from somewhere else. Computers capable
of taking Usenet feeds are down in the $500 range now, and
UNIX-capable boxes are going for under $2000, and there are at least
two UNIX lookalikes in the $100 price range.
No matter what, though, appealing to "Usenet" won't help. Even if
those who read such an appeal are sympathetic to your cause, they will
almost certainly have even less influence at your site than you do.
By the same token, if you don't like what some user at another site is
doing, only the administrator and/or owner of that site have any
authority to do anything about it. Persuade them that the user in
question is a problem for them, and they might do something (if they
feel like it).
If the user in question is the administrator or owner of the site from
which he or she posts, forget it; you can't win. Arrange for your
newsreading software to ignore articles from him or her if you can,
and chalk one up to experience.
WORDS TO LIVE BY #1:
USENET AS SOCIETY
--------------------
Those who have never tried electronic communication may not be aware
of what a "social skill" really is. One social skill that must be
learned, is that other people have points of view that are not only
different, but *threatening*, to your own. In turn, your opinions may
be threatening to others. There is nothing wrong with this. Your
beliefs need not be hidden behind a facade, as happens with
face-to-face conversation. Not everybody in the world is a bosom
buddy, but you can still have a meaningful conversation with them.
The person who cannot do this lacks in social skills.
-- Nick Szabo
WORDS TO LIVE BY #2:
USENET AS ANARCHY
--------------------
Anarchy means having to put up with things that really piss you off.
-- Unknown
--
Gene Spafford
NSF/Purdue/U of Florida Software Engineering Research Center,
Dept. of Computer Sciences, Purdue University, W. Lafayette IN 47907-1398
Internet: s...@cs.purdue.edu phone: (317) 494-7825
Path: gmdzi!unido!fauern!math.fu-berlin.de!ox.com!yale.edu!think.com!sdd.hp.com!mips!pacbell.com!att!cbnewse!danj1
From: Dan_Jacob...@ATT.COM
Newsgroups: news.admin
Subject: "What is Usenet?" kind of sexist
Message-ID: <1991Jul26Fr022216.Dan_Jacobson@ATT.COM>
Date: 26 Jul 91 06:22:19 GMT
References: <15396@ector.cs.purdue.edu>
Sender: da...@cbnewse.cb.att.com (daniel.jacobson)
Reply-To: Dan_Jacob...@ihlpz.ATT.COM
Organization: AT&T-BL, Naperville IL, USA
Lines: 18
In-Reply-To: spaf@cs.purdue.EDU's message of 25 Jul 91 23: 15:52 GMT
Not to be too picky, but in "What is Usenet?" of 25 Jul 91 23:15:52
GMT, I found,
(Lines matching
"\\<him\\>\\|\\<her\\>\\|\\<he\\>\\|\\<she\\>\\|\\<his\\>") [Also the
command "sexist" can be used on some UNIX systems to process this
input.]
56:Gene> someone is behaving unfairly, who's going to stop him? Neither
147:Gene> Every administrator controls his own site. No one has any real
148:Gene> control over any site but his own.
150:Gene> The administrator gets his power from the owner of the system he
152:Gene> administrator is doing, he can do whatever he pleases, up to and
201:Gene> "spirit" of the guidelines? Obviously, he can't. This fact leads
though near the end of the file it is non-sexist:
245:Gene> which he or she posts, forget it; you can't win. Arrange for your
246:Gene> newsreading software to ignore articles from him or her if you can,
Path: gmdzi!unido!math.fu-berlin.de!ox.com!emory!wupost!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <1991Jul30.141403.24269@newshost.anu.edu.au>
Date: 30 Jul 91 14:14:03 GMT
References: <1991Jul21.072057.12515@world.std.com> <1991Jul22.072111.17743@newshost.anu.edu.au> <91Jul24.112416edt.1530@smoke.cs.toronto.edu>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 107
In article <91Jul24.112416edt.1...@smoke.cs.toronto.edu> mor...@cs.toronto.edu
(Mark Moraes) writes:
>Those of us interested in viable[%] implementations of communications
>standards find short, comprehensible specs appealing. You see, some
>people do some of this implementation as a spare time activity because
>they find it interesting and/or challenging. [...]
>Perhaps you now understand why some people don't get excited by X.?00.
>It'll take too long to read the standards, let alone implement them.
Not many people implement mailers like sendmail and MMDF as a spare
time activity. The only one I can think of is smail (was that spare
time, are there others - I don't know?). To tackle an implementation
of X.400 and X.500 as a spare time activity would require a large
scale highly coordinated project embracing many people - not just
a single individual. That comes back to the points I was making
about organization - without it, no such spare-time project is
practical, even though complete source code is available for pilot
implementations that show how to do it.
Similarly with news transport software - C news seems to be the
only current project and I gather it also had some funding from
uunet so it may not be entirely "spare-time" either (though I
imagine more time would have been spent on it than has been paid for).
On the other hand LOTs of people have worked on mail User Agents and
news readers as a spare-time activity. Using the X/OPEN API for an
X.400 Message Store should be no harder than using any other API,
and would allow production of FAR more powerful mail User Agents
and news readers with concentration on the actual user interface
rather than on accessing messages.
>Pointing people at a working implementation is the way to prove that
>the specs you espouse may be superior to existing specs; hysterical
>handwaving in news.admin may feel very satisfying, but it isn't very
>convincing.
I point to isode-7.0 which includes full source code for an X.500
Directory Server and pp-5.2 which includes full source code for
an X.400 MTA, as well as the X/OPEN and X.400 API Association
series of APIs.
As for "hysterical handwaving", the most MODERATE hostile reactions
in news.admin have ranged from wanting to run away with rocket
assist to wanting to advance on it with rifles and flamethrowers.
The guy who ended up unable to say anything AT ALL about the issues
and just resorted to a personal attack on me may be an extreme case,
but surely your remarks about hysteria should be aimed in that
direction.
>Again, those of us interested in viable communications have discovered
>that it's very difficult to take a completely incompatible standard
>and force it down everyone's throats. Backward compatibility may be
>confining, but it's necessary if you want to talk to a reasonable
>number of people.
True in general, though I would argue that backward compatability
through gateways is an adequate measure.
But don't forget that X.400 IS going to be forced down everyone's
throats. Whether that is a good thing or not, the GOSIP requirements
mean that within a few years there WILL be FAR more people using
X.400 than RFC822 based mail. Failing to take the initiative does
NOT mean that X.400 will fade away from lack of support - it simply
means the initiative will be taken elsewhere and people comfortable
in their "backward compatability" will suddenly find themselves
unprepared, facing an urgent need for external compatability with
a mail system on which they never had any influence.
>Your statement that ASN.1 encoded data is easy to parse brightened up
>my morning (again!). Thank you.
You are welcome. In return for that small favour I would like to ask
you to help me brighten up other people's day by explaining exactly
what was amusing about it.
I'm reluctant to start a discussion on the binary encoding of ASN.1
in view of the unfortunate incident last time (when it was claimed
that it added bulk for the benefit of European telephone companies
and as soon as I explained that a type length value encoding is
ALWAYS less bulky than the header keywords used by RFC822 the
poor guy's brain apparently snapped).
However as you seem to be in a good humor and unlikely to resort
to personal abuse I will try one more time.
The binary encoding of ASN.1 results in a simple type length value
for each field. I maintain not only that this is easy to parse, but
also that it is utterly IMPOSSIBLE to devise any encoding that
could be EASIER to parse. No matter what other kind of binary
encoding you use, the output from a lexical scanner would consist
of a series of type tokens with the length of data they apply to,
and the data itself. So what could possibly be simpler than having
that provided as the encoding?
>% To me, "viable" means I should be able to get the programs without
>too much hassle, and run them on relatively small machines, where I
>also have to get other "real" work done.
Looking up isode and pp with archie and using ftp does not strike me
as a great deal of hassle. They DO take up a lot of disk space, but
disk space is cheap these days ($2000 for 1 Gb) and so does X windows
to which you have contributed, so I don't think you could mean that.
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!mcsun!uunet!sugar!taronga!peter
From: pe...@taronga.hackercorp.com (Peter da Silva)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <56L6I0E@taronga.hackercorp.com>
Date: 31 Jul 91 13:27:36 GMT
References: <1991Jul30.141403.24269@newshost.anu.edu.au>
Organization: A corner of our bedroom
Lines: 49
cmf...@anu.oz.au (Albert Langer) writes:
> To tackle an implementation
> of X.400 and X.500 as a spare time activity would require a large
> scale highly coordinated project embracing many people
Isn't that the point? RFC822 has small implementations at the routing
level. Smail 2 runs in small model on an 80286.
> Similarly with news transport software - C news seems to be the
> only current project and I gather it also had some funding from
> uunet so it may not be entirely "spare-time" either (though I
> imagine more time would have been spent on it than has been paid for).
C News was basically *done* before UUNET got involved.
> Using the X/OPEN API for an
> X.400 Message Store should be no harder than using any other API,
"All APIs are equal", eh? In that case we might as well dump UNIX and go
back to OS/360.
> As for "hysterical handwaving", the most MODERATE hostile reactions
> in news.admin have ranged from wanting to run away with rocket
> assist to wanting to advance on it with rifles and flamethrowers.
Reasonable reactions given what's been proposed. Previous ISO efforts
have been less than successfull. Hand up everyone using OSI instead of
TCP/IP! We're using a proprietary network that's been moving towards
OSI, and even so we've had to add TCP/IP to connect to third-party
systems.
> But don't forget that X.400 IS going to be forced down everyone's
> throats.
Like OSI?
> They DO take up a lot of disk space, but
> disk space is cheap these days
Disk space is cheap! RAM is cheap! Who cares about practical considerations
like getting it to run on PCs?
> and so does X windows
Perfect. So this mail system will be comparable to X? It'll be a cold day
in heck before I run either of them.
--
Peter da Silva. `-_-' Taronga Park BBS +1 713 568 0480 2400/n/8/1
Taronga Park. 'U` "Have you hugged your wolf, today?"
Path: gmdzi!unido!math.fu-berlin.de!ox.com!caen!sol.ctr.columbia.edu!samsung!uunet!mcsun!ukc!warwick!cudep
From: cu...@warwick.ac.uk (Ian Dickinson)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <D5H}PLC@warwick.ac.uk>
Date: 31 Jul 91 15:48:26 GMT
References: <TV26ZI9@taronga.hackercorp.com> <2319@ecicrl.ocunix.on.ca> <1991Jul23.122947.25080@newshost.anu.edu.au>
Sender: n...@warwick.ac.uk (Network news)
Organization: Team Limpid's Death Mollusc From Hell - Kunst und Wahnsinn
Lines: 20
Nntp-Posting-Host: thistle
In article <1991Jul23.122947.25...@newshost.anu.edu.au> cmf...@anu.oz.au (Albert Langer) writes:
>Well, that's at least ONE corroborating witness for Peter's 10 seconds per
>day. I'm still skeptical, but I've made a note of the software both of you
>have recommended (thanks) and will just have to see how it goes.
>
>But I still have a suspicion that the lack of loud cries to the
>contrary is related to the "Emperor's New Clothes" syndrome. I mean
>who wants to admit that they spend 10 or 100 times as much time
>maintaining their news system as Peter and Chris do :-)
I often spend more than 10 times as long as Peter does.
2 minutes a day is hardly a killer, is it? :-)
Cheers,
--
\/ato /'\ /`\
Ian Dickinson TED KALDIS FOR PRESIDENT! /^^^\/^^^\
v...@warwick.ac.uk /TWIN/TEATS\
@c=GB@o=University of Warwick@ou=Computing Services@cn=Ian Dickinson / \
Path: gmdzi!unido!unidui!math.fu-berlin.de!ox.com!yale.edu!think.com!samsung!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <1991Jul31.170600.5588@newshost.anu.edu.au>
Date: 31 Jul 91 17:06:00 GMT
References: <1991Jul30.141403.24269@newshost.anu.edu.au> <56L6I0E@taronga.hackercorp.com>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 57
In article <56L6...@taronga.hackercorp.com> pe...@taronga.hackercorp.com
(Peter da Silva) quotes me and writes:
>> To tackle an implementation
>> of X.400 and X.500 as a spare time activity would require a large
>> scale highly coordinated project embracing many people
>
>Isn't that the point? RFC822 has small implementations at the routing
>level. Smail 2 runs in small model on an 80286.
And FidoNet software runs on 8088 XTs - so what?
I see no reason why an X.400 MTA and X.500 DSA could not be written for
an 8088 XT or a Mac se and I would like to see that done. But the existing
freely available implementations are for BSD unix and I believe
they are complex enough that to port them as a spare time activity
would require several people. So what? Should we only use software
that can be developed by one person alone in their spare time?
>> Using the X/OPEN API for an
>> X.400 Message Store should be no harder than using any other API,
>
>"All APIs are equal", eh? In that case we might as well dump UNIX and go
>back to OS/360.
What I meant was that while I agree that an MTA or DSA would be a
major project, I believe that using the APIs to write superior
X.400 UAs and DUAs could easily be done by individuals working
alone in their spare time - just as people who could not possibly
implement X window servers and might even have great difficulty
porting them can nevertheless write useful client applications -
is that so silly?
>Disk space is cheap! RAM is cheap! Who cares about practical considerations
>like getting it to run on PCs?
I would estimate that CCITT and ISO do not care about this and further
that this is a reasonable attitude from their perspective (only UAs
and DUAs need to run on PCs for their design aims to be achieved).
Likewise RFC822 mail and RFC1036 news was not designed by people who
cared whether it would run on PCs and it was not until quite a lot
later that PC implementations were developed.
I do care about getting MTAs and DSAs that can run on PCs because I would like
to see this technology available to much wider circles than those with
access to corporate and educational LANs - people who need dial-up access
from home computers. Therefore I would like to accelerate the development
of suitable software rather than waiting for it to trickle down.
If you disagree with either the feasability or desirability of that,
it is no reason to treat it as though it represents some reprehensible
attack on your most vital interests. Life goes on - people do different things.
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!math.fu-berlin.de!ox.com!yale.edu!qt.cs.utexas.edu!zaphod.mps.ohio-state.edu!samsung!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <1991Aug1.174141.15227@newshost.anu.edu.au>
Date: 1 Aug 91 17:41:41 GMT
References: <2319@ecicrl.ocunix.on.ca> <1991Jul23.122947.25080@newshost.anu.edu.au> <D5H}PLC@warwick.ac.uk>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 42
In article <D5H}...@warwick.ac.uk> cu...@warwick.ac.uk
(Ian Dickinson) quotes me and writes:
>>Well, that's at least ONE corroborating witness for Peter's 10 seconds per
>>day. I'm still skeptical, but I've made a note of the software both of you
>>have recommended (thanks) and will just have to see how it goes.
>>
>>But I still have a suspicion that the lack of loud cries to the
>>contrary is related to the "Emperor's New Clothes" syndrome. I mean
>>who wants to admit that they spend 10 or 100 times as much time
>>maintaining their news system as Peter and Chris do :-)
>
>I often spend more than 10 times as long as Peter does.
>
>2 minutes a day is hardly a killer, is it? :-)
Well, I've had another corroborating witness for 10 seconds per day
by email.
But yes, 2 minutes a day IS a killer.
The discussion was about the feasability of using current news and
mail software (enhanced) for naive users to operate their own single
user nodes that pass news feeds on to each other.
As little as 2 minutes per day maintenance time would require a
full-time sysadmin for about 150 users. I don't think that's acceptable.
But if Peter is right and one tenth of that can be achieved, I think
one sysadmin per 1500 users would be worth investigating.
Unlike X.400 and X.500, this is an area in which Peter presumably
DOES know what he is talking about. But I would still be interested
in other opinions.
Any others reckon they spend as little as 10 seconds per day on
news and mail maintenance? Any others reckon they spend more than
2 minutes per day?
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!math.fu-berlin.de!ox.com!yale.edu!think.com!samsung!uunet!pdn!tct!chip
From: c...@tct.com (Chip Salzenberg)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <28995A6E.4686@tct.com>
Date: 2 Aug 91 13:13:18 GMT
References: <1991Jul22.072111.17743@newshost.anu.edu.au> <91Jul24.112416edt.1530@smoke.cs.toronto.edu> <1991Jul30.141403.24269@newshost.anu.edu.au>
Organization: TC Telemanagement, Clearwater, FL
Lines: 67
According to cmf...@anu.oz.au (Albert Langer):
>Not many people implement mailers like sendmail and MMDF as a spare
>time activity. The only one I can think of is smail (was that spare
>time, are there others - I don't know?).
I know of Sendmail, MMDF, Smail 2.5, Smail 2, Smail 3, Zmailer, and
PMDF. (I wrote Deliver, but it doesn't know how to parse address
headers, so it doesn't count.) To my knowledge, all these mail
transports were developed in spare time (or at least unpaid).
>X.400 and X.500 as a spare time activity would require a large
>scale highly coordinated project embracing many people - not just
>a single individual.
That's a bug, not a feature.
>Similarly with news transport software - C news seems to be the
>only current project ...
It came into existence entirely due to spare-time work. So did B News
(which is workable, if outdated) and InterNetNews (currently in beta
test).
The basics of news transport aren't difficult to write. The hard part
of writing a news or mail transport -- the people involved will please
correct me if I am wrong -- is careful optimization so as to provide
satisfactory performance in the face of high traffic volume.
X.400 implementors will face the same performance issues. However,
the complexity of the standard will make correctness a significant
sink for creative effort. Acceptable performance will of necessity
have to wait for later. This is progress?
>Using the X/OPEN API for an X.400 Message Store should be no harder
>than using any other API ...
The fallacy of "all APIs are equal" should be obvious.
>I point to isode-7.0 which includes full source code for an X.500
>Directory Server and pp-5.2 which includes full source code...
I believe that I shall find these programs and read them. They should
be good for something, if only entertainment.
>Failing to take the initiative does NOT mean that X.400 will fade away
>from lack of support - it simply means the initiative will be taken
>elsewhere ...
You can always tell the pioneers by the arrows in their backs. Unless
something drastic happens, I plan to wait for X.400 to be proved
viable in large networks, and for it to make the inevitable
accomodations to experience, before I write an X.400 anything.
>... a type length value encoding is ALWAYS less bulky than the header
>keywords used by RFC822 ...
I agree with this X.400 design decision, and I think that Usenet can
benefit from it.
I see no reason why a type length value encoding could not be
introduced into news transport between cooperating sites. Such a
feature could be used in a way analogous to batch SMTP between UUCP
sites: those who upgrade reap the benefits; those who do not upgrade
continue to work without breakage.
--
Chip Salzenberg at Teltronics/TCT <c...@tct.com>, <uunet!pdn!tct!chip>
"You must make this fire very big, so I can find you."
Path: gmdzi!unido!math.fu-berlin.de!ox.com!yale.edu!think.com!samsung!uunet!pdn!tct!chip
From: c...@tct.com (Chip Salzenberg)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <28995BA2.1829@tct.com>
Date: 2 Aug 91 13:18:26 GMT
References: <1991Jul30.141403.24269@newshost.anu.edu.au> <56L6I0E@taronga.hackercorp.com> <1991Jul31.170600.5588@newshost.anu.edu.au>
Organization: TC Telemanagement, Clearwater, FL
Lines: 15
According to cmf...@anu.oz.au (Albert Langer):
>Should we only use software that can be developed by one person alone
>in their spare time?
The question is whether the loss of spare-time projects is worth other
gains. I'd say that the jury is still out on that question, though I
prefer software that is comprehensible to a single human being.
>Likewise RFC822 mail and RFC1036 news was not designed by people who
>cared whether it would run on PCs ...
This claim is interesting. Is there any evidence for it?
--
Chip Salzenberg at Teltronics/TCT <c...@tct.com>, <uunet!pdn!tct!chip>
"You must make this fire very big, so I can find you."
Path: gmdzi!unido!fauern!math.fu-berlin.de!ox.com!caen!zaphod.mps.ohio-state.edu!mips!pacbell.com!att!linac!midway!msuinfo!sharkey!fmsrl7!art-sy!news
From: c...@art-sy.detroit.mi.us (j chapman flack)
Newsgroups: news.admin,soc.women,soc.men
Subject: What is English? (was Re: "What is Usenet?" kind of sexist)
Message-ID: <9108021924.aa12654@art-sy.detroit.mi.us>
Date: 2 Aug 91 23:24:23 GMT
References: <15396@ector.cs.purdue.edu>
<1991Jul26Fr022216.Dan_Jacobson@ATT.COM>
<5253@lib.tmc.edu>
Reply-To: c...@art-sy.detroit.mi.us (j chapman flack)
Followup-To: soc.women,soc.men
Organization: Appropriate Roles for Technology
Lines: 48
In <1991Jul26Fr022216.Dan_Jacob...@ATT.COM> Dan_Jacob...@ihlpz.ATT.COM wrote:
>Not to be too picky, but in "What is Usenet?" ... I found,
[ several lines featuring 'he' or 'him' as a generic pronoun ]
In <5...@lib.tmc.edu> jmayn...@oac.hsc.uth.tmc.edu (Jay Maynard) writes:
>This one set the bullshit flag.
^^^^^^^^ er, "difference of opinion"?
>
>It is proper English usage to use the masculine pronoun to refer to
>...
>considered sexist. The article is correct and should be left alone.
I've seen several responses of this type, resting on some assertion of
"correct" or "proper" English. Human languages are dynamic, and the gender-
unspecified third-person pronoun is one area where change is taking place in
English. He/him and various alternative forms are all in widespread use
these days, all by educated speakers of English. Many speakers, regardless of
which form they use in their own speech, don't give a hoot which form others
choose. Those who choose to complain about others' use of he/him do so
because of perceived social implications, not ideas of "correct English," and
to respond to them with a "correct English" argument is to miss the point.
Those who respond with surprising emotional fervor ("bullshit"?) seem to be
really stating their own positions on a *social* question, dressed up as
"objective" with reference to a "correct English" which is, however, illusory.
"...the puristic attitude towards language--the idea that there is an
absolute standard of correctness which should be maintained--has its
origin in a natural nostalgic tendency in man [sic ;-)], supplemented
and intensified by social pressures. It is illogical, and impossible
to pin down to any firm base. Purists behave as if there was a vintage
year when language achieved ... excellence .... In fact, there never
was such a year. The language of Chaucer or Shakespeare's time was no
better and no worse than that of our own--just different."
-Jean Aitchison, _Language Change: Progress or Decay_, 1981, 0-87663-456-0
I agree that the most constructive action for the original complainer would
probably have been to write to Mr. Spafford and offer to rewrite the affected
passages.
If anyone really *wants* to continue the thread, might I suggest soc.women &/|
soc.men? I don't think the issue is really language, except in a superficial
sense, so the sci.lang'ers probably aren't interested. News administrators
almost certainly aren't. I've changed the Followup-To:.
--
Chap Flack Their tanks will rust. Our songs will last.
c...@art-sy.detroit.mi.us -MIKHS 0EODWPAKHS
Nothing I say represents Appropriate Roles for Technology unless I say it does.
Path: gmdzi!unido!fauern!math.fu-berlin.de!ox.com!caen!zaphod.mps.ohio-state.edu!qt.cs.utexas.edu!cs.utexas.edu!utgpu!nrcnet0!cunews!micor!latour!ecicrl!clewis
From: cle...@ferret.ocunix.on.ca (Chris Lewis)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <2397@ecicrl.ocunix.on.ca>
Date: 3 Aug 91 06:47:07 GMT
References: <2319@ecicrl.ocunix.on.ca> <1991Jul23.122947.25080@newshost.anu.edu.au> <D5H}PLC@warwick.ac.uk> <1991Aug1.174141.15227@newshost.anu.edu.au>
Organization: Elegant Communications Inc., Ottawa, Canada
Lines: 52
In article <1991Aug1.174141.15...@newshost.anu.edu.au> cmf...@anu.oz.au (Albert Langer) writes:
>In article <D5H}...@warwick.ac.uk> cu...@warwick.ac.uk
>(Ian Dickinson) quotes me and writes:
>>I often spend more than 10 times as long as Peter does.
>>2 minutes a day is hardly a killer, is it? :-)
>Well, I've had another corroborating witness for 10 seconds per day
>by email.
>But yes, 2 minutes a day IS a killer.
>The discussion was about the feasability of using current news and
>mail software (enhanced) for naive users to operate their own single
>user nodes that pass news feeds on to each other.
>As little as 2 minutes per day maintenance time would require a
>full-time sysadmin for about 150 users. I don't think that's acceptable.
If "150 users" have a single sysadmin, then they have no business
each running a full blown news system, they should be running a
news reader from a LAN-based article server (ala NNTP, rrn etc.).
That isn't much more sysadmin load than a single news system.
(presuming that the LAN support is needed independently of news and
is already budgetted).
If "150 users" are *really* independent single user nodes that really
feed each other, that means that each of the 150 users has to spend the
daily time; there is no sysadmin - whether it's 10 seconds or 2 minutes/day
it's not a killer because each user does it.
If "150 users" were really single user nodes that actually feed each
other in a traditional sense, but rely is on one sysadmin person,
the sysadmin would be smart (we'd hope) and make all of the configurations
identical, and so his/her work factor would not scale linearly with the
number of systems.
The real way to get news easy to install in a global sense is to build
preconfigured binary versions of news for each type of system out there.
No installation options, no optional features - as one might say, "you can
get any colour you want, as long as it's black"
It doesn't matter how good the configuration utilities or the software
is, if the user can affect the installation in any way, there will
be significant numbers of broken installs. Just witness the number of
times Henry has to say "RTFM" w.r.t. the DBZ/Xenix optimizer bug.
95% of the bug reports on psroff have RTFM answers too. Sigh.
--
Chris Lewis; UUCP: ...!{cunews,uunet,latour}!ecicrl!clewis;
DOMAIN: cle...@ferret.ocunix.on.ca; Phone: Canada 416 832-0541
Psroff info: psroff-requ...@ferret.ocunix.on.ca
Ferret mailing list: ferret-requ...@ferret.ocunix.on.ca
Path: gmdzi!unido!fauern!math.fu-berlin.de!ox.com!yale.edu!think.com!spool.mu.edu!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <1991Aug3.193043.1862@newshost.anu.edu.au>
Date: 3 Aug 91 19:30:43 GMT
References: <D5H}PLC@warwick.ac.uk> <1991Aug1.174141.15227@newshost.anu.edu.au> <2397@ecicrl.ocunix.on.ca>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 110
In article <2...@ecicrl.ocunix.on.ca> cle...@ferret.ocunix.on.ca
(Chris Lewis) writes:
>>The discussion was about the feasability of using current news and
>>mail software (enhanced) for naive users to operate their own single
>>user nodes that pass news feeds on to each other.
>
>>As little as 2 minutes per day maintenance time would require a
>>full-time sysadmin for about 150 users. I don't think that's acceptable.
>
>If "150 users" have a single sysadmin, then they have no business
>each running a full blown news system, they should be running a
>news reader from a LAN-based article server (ala NNTP, rrn etc.).
>That isn't much more sysadmin load than a single news system.
>(presuming that the LAN support is needed independently of news and
>is already budgetted).
No problem there - I regard that as a normal site. Problem under
discussion is access for general public who are not on supported
institutional or corporate LANs but accessing from home computers etc.
To keep it cheap, while allowing flexible sorting through messages,
they pickup their own batched newsfeeds overnight, thus requiring an
absolute minimum number of phone lines (hopefully, only their own
normal domestic phone lines). If they had to be online to a central
server to read news, far more dedicated phone lines would be required
and the system would not be so cheap. Likewise if it turns out that
in order to run the full blown news systems required for this mode,
too much sysadmin support is needed, then again it would not be cheap.
Can we design cheap access for the general public with enhanced
versions of current software? (My suspicion is that it would
require the sort of completely specified software available from
X.400, with Directory Services and a Management system rather
than just an enhancement of existing news and mail software. But
if the workload for a single user node is only 10 seconds per day
with a skilled sysadmin on site, then maybe something feasible can
be designed with a naive local user but a remote sysadmin.)
>If "150 users" are *really* independent single user nodes that really
>feed each other, that means that each of the 150 users has to spend the
>daily time; there is no sysadmin - whether it's 10 seconds or 2 minutes/day
>it's not a killer because each user does it.
Not if they are "naive users" - they CAN'T do it. That's the model, but
they have to "really feed each other" with ZERO local user action per
feed (and absolutely minimal networked commands or directory and routing
table entries from a remote sysadmin). Managing a feed has to be
equivalent to managing an extra user on a single multi-user system or
at worst equivalent to managing an extra node on a LAN (but WITHOUT
the continuous access one has via a LAN).
>If "150 users" were really single user nodes that actually feed each
>other in a traditional sense, but rely is on one sysadmin person,
>the sysadmin would be smart (we'd hope) and make all of the configurations
>identical, and so his/her work factor would not scale linearly with the
>number of systems.
Certainly the configurations must be identical - that would be
determined by the software developers (who would be developing this
for MANY such sysadmins each of whom looks after many users).
But after eliminating all problems that CAN be predicted in
advance, it may be that remaining "10 seconds per day" requirements
are essentially random events per node. Any estimates of workload
are pretty arbitrary but linear may be more accurate than assuming
one could have only twice as much sysadmin work with 1500 users instead of
150 users.
>The real way to get news easy to install in a global sense is to build
>preconfigured binary versions of news for each type of system out there.
>No installation options, no optional features - as one might say, "you can
>get any colour you want, as long as it's black"
>
>It doesn't matter how good the configuration utilities or the software
>is, if the user can affect the installation in any way, there will
>be significant numbers of broken installs. Just witness the number of
>times Henry has to say "RTFM" w.r.t. the DBZ/Xenix optimizer bug.
>95% of the bug reports on psroff have RTFM answers too. Sigh.
Absolutely. We are talking about "naive users" who could not install
or configure software anyway but are only capable of using a newsreader
(perhaps a more friendly one).
It might be that a cheap version of unix is booted for news purposes
on their machine that normally runs DOS, or the news software might
be specially ported (like Waffle). Assume every reasonable measure,
short of a complete redesign of the underlying news and mail software,
is taken so it can be installed and configured automatically and will
run itself without operator intervention. (Including things like
Brad Templeton's dynafeed and sophisticated expiry routines to avoid
the disk getting full).
A full-time competent system administrator somewhere "looks after"
a sub-network of these single user nodes that are passing feeds
on to each other - e.g. adding new nodes and defining their links
(which should also be automatic as far as possible) AND FIXING
PROBLEMS WHEN SOMETHING GOES WRONG. (Normally via networked commands).
The question is then, how many such single user nodes could that
system administrator look after? Things WILL go wrong (apart from
hardware failures) so the answer won't be "any number you like".
An alternative question, which may be less accurate if the workload
is not linear in the number of systems, but may help to visualize
the problem, is how many seconds per day of sysadmin attention
would each node add?
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!math.fu-berlin.de!ox.com!yale.edu!qt.cs.utexas.edu!cs.utexas.edu!utgpu!news-server.csri.toronto.edu!utzoo!henry
From: he...@zoo.toronto.edu (Henry Spencer)
Newsgroups: news.admin,soc.women,soc.men
Subject: Re: What is English? (was Re: "What is Usenet?" kind of sexist)
Message-ID: <1991Aug3.232241.4138@zoo.toronto.edu>
Date: 3 Aug 91 23:22:41 GMT
References: <15396@ector.cs.purdue.edu> <1991Jul26Fr022216.Dan_Jacobson@ATT.COM> <5253@lib.tmc.edu> <9108021924.aa12654@art-sy.detroit.mi.us>
Organization: U of Toronto Zoology
Lines: 23
In article <9108021924.aa12...@art-sy.detroit.mi.us> c...@art-sy.detroit.mi.us (j chapman flack) writes:
>... Those who choose to complain about others' use of he/him do so
>because of perceived social implications, not ideas of "correct English," and
>to respond to them with a "correct English" argument is to miss the point.
>Those who respond with surprising emotional fervor ("bullshit"?) seem to be
>really stating their own positions on a *social* question, dressed up as
>"objective" with reference to a "correct English" which is, however, illusory.
Actually, some of them (me, for example) just object to replacing clear
English with clumsy, obscure, even obfuscated English on the assumption
that this trendy, superficial nonsense will somehow solve difficult and
important social problems.
>I agree that the most constructive action for the original complainer would
>probably have been to write to Mr. Spafford and offer to rewrite the affected
>passages.
Were I Spaf, receiving such an offer, my reply would be "if you can do it
without convoluting the sentence structure or otherwise impairing readability,
go ahead, otherwise forget it".
--
Arthritic bureaucracies don't tame new | Henry Spencer @ U of Toronto Zoology
frontiers. -Paul A. Gigot, WSJ, on NASA | he...@zoo.toronto.edu utzoo!henry
Path: gmdzi!unido!mcsun!uunet!elroy.jpl.nasa.gov!usc!apple!snorkelwacker.mit.edu!world!geoff
From: ge...@world.std.com (Geoff Collyer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <1991Aug4.202830.26944@world.std.com>
Date: 4 Aug 91 20:28:30 GMT
References: <1991Jul22.072111.17743@newshost.anu.edu.au>
<91Jul24.112416edt.1530@smoke.cs.toronto.edu>
<1991Jul30.141403.24269@newshost.anu.edu.au> <28995A6E.4686@tct.com>
Organization: Software Tool & Die Netnews Research Center
Lines: 14
Chip Salzenberg:
>The basics of news transport aren't difficult to write. The hard part
>of writing a news or mail transport -- the people involved will please
>correct me if I am wrong -- is careful optimization so as to provide
>satisfactory performance in the face of high traffic volume.
I guess it depends upon how you define "hard". To me, optimisation was
one of the fun parts. The parts that caused us by far the most work and
aggravation have invariably been human-interface components (primarily
inews, message-id lengths, configuration, mail in response to control
messages, documentation and date parsing). The hard parts were political,
not technical.
--
Geoff Collyer world.std.com!geoff, uunet.uu.net!geoff
Path: gmdzi!unido!mcsun!uunet!spool.mu.edu!think.com!snorkelwacker.mit.edu!world!geoff
From: ge...@world.std.com (Geoff Collyer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <1991Aug4.204354.27373@world.std.com>
Date: 4 Aug 91 20:43:54 GMT
References: <1991Jul30.141403.24269@newshost.anu.edu.au>
<56L6I0E@taronga.hackercorp.com>
<1991Jul31.170600.5588@newshost.anu.edu.au> <28995BA2.1829@tct.com>
Organization: Software Tool & Die Netnews Research Center
Lines: 20
>>Likewise RFC822 mail and RFC1036 news was not designed by people who
>>cared whether it would run on PCs ...
I'm not sure if the claim is that the RFCs or the implementations were
designed by people who ...
>This claim is interesting. Is there any evidence for it?
If the above claim is about implementations, then there is counter
evidence: C News has been run on PCs (under MS DOS, even) and was
developed on PDP-11s, which are more restrictive than PCs in per-process
address space, by people who cared very much that it run and run well on
PDP-11s. A and B News ran on PDP-11s too. But ultimately it doesn't
matter what the designers *cared* about; what matters is how good the
resulting specification or implementation is. Implementations of both
RFC822 mail (e.g. sendmail) and RFC1036 news ran on the PDP-11, so those
RFCs clearly did not require implementations to be enormous. Can the
same be said of any mail-related ISO standard?
--
Geoff Collyer world.std.com!geoff, uunet.uu.net!geoff
Path: gmdzi!unido!math.fu-berlin.de!ox.com!yale.edu!qt.cs.utexas.edu!zaphod.mps.ohio-state.edu!wupost!uunet!munnari.oz.au!manuel!cmf851
From: cmf...@anu.oz.au (Albert Langer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <1991Aug4.203948.8257@newshost.anu.edu.au>
Date: 4 Aug 91 20:39:48 GMT
References: <91Jul24.112416edt.1530@smoke.cs.toronto.edu> <1991Jul30.141403.24269@newshost.anu.edu.au> <28995A6E.4686@tct.com>
Sender: n...@newshost.anu.edu.au
Organization: Computer Services Centre, Australian National University, Canberra, Australia.
Lines: 174
In article <28995A6E.4...@tct.com> c...@tct.com
(Chip Salzenberg) quotes me and writes:
>>X.400 and X.500 as a spare time activity would require a large
>>scale highly coordinated project embracing many people - not just
>>a single individual.
>
>That's a bug, not a feature.
It's neither. As Peter remarks the KISS principle should be applied
whenever possible, but some projects just ARE big. Perhaps as software
technology matures it will become possible for individuals to tackle
projects as complex as Directory Services, just as you say individuals
have produced mailers like Zmailer and Sendmail as well as the variants
of smail - but then there will be still larger projects that will
require coordinated efforts by many. We use operating systems that
were not written by individuals in their spare time - and add many
improvements to them - why not do the same with a more sophisticated
mail system (especially since the hard part will soon be included
with the operating system)?
>The basics of news transport aren't difficult to write. The hard part
>of writing a news or mail transport -- the people involved will please
>correct me if I am wrong -- is careful optimization so as to provide
>satisfactory performance in the face of high traffic volume.
>
>X.400 implementors will face the same performance issues. However,
>the complexity of the standard will make correctness a significant
>sink for creative effort. Acceptable performance will of necessity
>have to wait for later. This is progress?
Yes, the initial implementations of X.400 MTAs would have had most
of the effort put into just complying with the standard and would
not have adequate performance for news transport (though PP is
handling high volumes at mail gateways). Later implementations can
improve performance and add features. That IS progress - as with
B news followed by C news. I'm not suggesting there is a mature
X.400 "product" available for news transport now - I'm saying the
opportunity is there to build one. See below for why it will be
easier to improve performance beyond C news using X.400 than by
re-designing current news software.
>>I point to isode-7.0 which includes full source code for an X.500
>>Directory Server and pp-5.2 which includes full source code...
>
>I believe that I shall find these programs and read them. They should
>be good for something, if only entertainment.
Excellent! I recommend you start with the docs though (in *ps.tar.Z files)
- 20 MB or so of source code isn't light reading! The Directory Service
stuff is probably the best place to start (volume 5 of the isode manual
and in pilot-ps.tar.Z). Unfortunately the standards themselves and
the XOPEN APIs aren't available online but you need to read them too.
Don't get put off by the bulk - the isode is a research tool
with a developer's kit for OSI applications in general, not just an
implementation. Also most of the complexity of PP arises from it's role
as a gateway between X.400, RFC822 and Janet Greybook mail - a
standalone X.400 MTA would be much simpler (but useless at present).
>>Failing to take the initiative does NOT mean that X.400 will fade away
>>from lack of support - it simply means the initiative will be taken
>>elsewhere ...
>
>You can always tell the pioneers by the arrows in their backs. Unless
>something drastic happens, I plan to wait for X.400 to be proved
>viable in large networks, and for it to make the inevitable
>accomodations to experience, before I write an X.400 anything.
Fair enough, except that one should be appreciative of the pioneers
at the bleeding edge while the prevailing mood seems to be hostile
towards them. It really is SEVERAL YEARS down the track now and
X.400 HAS been proved viable in large networks (especially in Europe).
Incidentally the Internet pilot implementation of a free X.500 Directory
Service covering users as well as hosts, which Peter was so dubious
about, is about 2 years old now and covered more than a quarter of a
million entries last time I looked.
>>... a type length value encoding is ALWAYS less bulky than the header
>>keywords used by RFC822 ...
>
>I agree with this X.400 design decision, and I think that Usenet can
>benefit from it.
>
>I see no reason why a type length value encoding could not be
>introduced into news transport between cooperating sites. Such a
>feature could be used in a way analogous to batch SMTP between UUCP
>sites: those who upgrade reap the benefits; those who do not upgrade
>continue to work without breakage.
Yes, I agree that form of encoding can be useful independently of
X.400 itself (and is used for example in SNMP). It could be interesting
to introduce it to RFC style news and mail, but some of the issues
that would arise in doing so may help to highlight the advantages of
the OSI approach.
1. The reduced bulk and easier parsing are merely SIDE EFFECTS of
a notation (ASN.1) that was primarily designed for handling more
complex specs. They only came up here because Geoff Collyer and Mark
Moraes made the extraordinary claims that the binary encoding was
MORE bulky and HARDER to parse than the header keywords of RFC822.
I doubt that the improvement would be worth the trouble for bilateral
links (though at least it IS something that could be introduced
between pairs of sites without requiring changes elsewhere).
2. With an OSI presentation layer there is one place where you
can redefine an encoding method bilaterally and also negotiate
which encodings are available for any session, but there is
no equivalent to a presentation layer in current mail and news
software. So you would effectively be adding a conversion stage
to the new encoding as each message is exported and again back
to RFC822 style as each message is imported, with only a LOSS
in efficiency for a fairly slight reduction in bulk (since
LZW compression would tend to include header keywords in the
compression dictionaries anyway).
3. The ASN.1 style encoding would be much more useful if it was
used throughout the news and mail system including news readers
and NNTP as well as rnews and inews equivalents. THEN everything
would be pre-parsed and easy to get at. That unfortunately would
require re-writing many different programs that currently access
/usr/spool/news.
4. This shows the advantage of the seemingly "academic" OSI
definitions of "abstract services" with defined input and
output "ports". There is a consistent "open" way of accessing and
operating on messages through the network whereas current news readers
and news transport software are tied to each other through the
specific implementation of /usr/spool/news files. That will make
it quite difficult to achieve the next breakthrough in performance
since to bypass the unix file system bottleneck you have to rewrite
news transport and newsreaders and the interface between them at once
- which is not a spare time individual project.
5. A reasonable approach might be to define a new version of the NNTP
protocol that uses the binary encoding. This could be used bilaterally
for news transport and also for news readers. But it is pretty
inefficient having to call up all the messages, or at least their
headers through NNTP in order to filter and select those of interest
and one would want to add NNTP commands that retrieve messages or
selected attributes of messages by filters. One could then replace
the /usr/spool/news format in a later implementation of that revised NNTP
which runs more efficiently, while still using the same newsreaders and
the same interface. This would also provide the benefit of bilateral
selective newsfeeds deferring or delaying GIFs etc. Dial-up sites
with no Internet connection could use it over SLIP or PPP instead
of UUCP.
6. The nice thing about using the XOPEN Message Store API is that
it allows newsreader designers to concentrate on designing the
newsreader to that API (which can be a 1 person, spare time
project), while others concentrate on designing the colocated
MTA and Message Stores that implement the API (a bigger project).
I gather that libraries implementing the API will soon be available
for the next releases of unix so I would be more inclined to focus
on learning to use the API. One then also gets formatted text,
multi-media body parts, privacy and authentication etc all thrown
in.
7. The other big advantage of X.400 for news would be use of the X.500
Directory Service to define Distribution Lists and thus allow
unlimited creation, filtering and combination of newsgroups with the selectivity
of mailing lists but the broadcast efficiency of news. Anybody could
create their own filtered combination of existing newsgroups or start their
own and advertize it in the Directory, with others easily locating and
subscribing to it, without messages travelling over any route more than once.
(Implementation of Clifford Adams "post-moderation" concept could simply
use moderators Distribution Lists to distribute message-IDs and keywords
which are then used for selecting which messages to pull from Message
Stores).
--
Opinions disclaimed (Authoritative answer from opinion server)
Header reply address wrong. Use cmf...@csc2.anu.edu.au
Path: gmdzi!unido!mcsun!uunet!zaphod.mps.ohio-state.edu!think.com!snorkelwacker.mit.edu!world!geoff
From: ge...@world.std.com (Geoff Collyer)
Newsgroups: news.admin
Subject: Re: What is Usenet?
Message-ID: <1991Aug5.233901.23679@world.std.com>
Date: 5 Aug 91 23:39:01 GMT
References: <91Jul24.112416edt.1530@smoke.cs.toronto.edu>
<1991Jul30.141403.24269@newshost.anu.edu.au> <28995A6E.4686@tct.com>
<1991Aug4.203948.8257@newshost.anu.edu.au>
<91Aug5.190250edt.1528@smoke.cs.toronto.edu>
Organization: Software Tool & Die Netnews Research Center
Lines: 19
>>would not have adequate performance for news transport (though PP is
>>handling high volumes at mail gateways).
Mailing each article of a full news feed (or anything approaching a full
news feed) separately is madness. Utter madness. Try it and watch your
machine grind itself into dust.
>>See below for why it will be easier to improve performance beyond C
>>news using X.400 than by re-designing current news software.
Sorry, C News throughput is limited by disk access times, not header
parsing time. X.400 won't improve matters at all. RAIDs might, lighting
a fire under disk manufacturers might (halving average access times in a
decade is not that impressive, particularly when compared with CPU and
memory speed increases in that decade). Alas, with the demand for
cheeeeap, beeeg SCSIs for the desktop driving the disk market, I'm not
hopeful. Sigh, the tyranny of the desktop.
--
Geoff Collyer world.std.com!geoff, uunet.uu.net!geoff
Path: gmdzi!unido!mcsun!uunet!cs.utexas.edu!utgpu!nrcnet0!cunews!micor!ecicrl!clewis
From: cle...@ferret.ocunix.on.ca (Chris Lewis)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <2413@ecicrl.ocunix.on.ca>
Date: 6 Aug 91 18:33:24 GMT
References: <D5H}PLC@warwick.ac.uk> <1991Aug1.174141.15227@newshost.anu.edu.au> <2397@ecicrl.ocunix.on.ca> <1991Aug3.193043.1862@newshost.anu.edu.au>
Organization: Elegant Communications Inc., Ottawa, Canada
Lines: 96
In article <1991Aug3.193043.1...@newshost.anu.edu.au> cmf...@anu.oz.au (Albert Langer) writes:
>In article <2...@ecicrl.ocunix.on.ca> cle...@ferret.ocunix.on.ca
>(Chris Lewis) writes:
>>>The discussion was about the feasability of using current news and
>>>mail software (enhanced) for naive users to operate their own single
>>>user nodes that pass news feeds on to each other.
After the last go round, *now* I see what you're driving at. USENET/Mail
as an "appliance". Actually, I'm a more horrified at the class
of users that that approach would bring to the net than of the
technical difficulties.
The biggest technical difficulty is that of each of these appliances
shipping news to each other. For if the "appliance" is to behave
anyways reasonably, given the current network load, each appliance
would have to work with a partial feed matching their owner's interests,
and that leads to extreme difficulties in trying to match everybody's
wants with everyone else's newsgroup list. That would require
some sort of dynamic news feeding mechanism that not only
decides what newsgroups to feed over a particular link (ie: Brad's
dynafeed), but what feed sites to select. Ugh. Not to mention
the delays by assuming nightly-only downloads and single lines.
>Can we design cheap access for the general public with enhanced
>versions of current software? (My suspicion is that it would
>require the sort of completely specified software available from
>X.400, with Directory Services and a Management system rather
>than just an enhancement of existing news and mail software. But
>if the workload for a single user node is only 10 seconds per day
>with a skilled sysadmin on site, then maybe something feasible can
>be designed with a naive local user but a remote sysadmin.)
As far as building an USENET/Mail "appliance" that is only a leaf,
it is my firm belief that the current RFC-compliant software is
perfectly adequate for the job. The work needed to build this
appliance is mainly an integration and bullet-proofing task:
1 stripped down cnews that has no administrator interfaces
(all control messages are enabled, only cancel passes thru
main hub automatically, others need to be vetted by sysadmin)
Followed by a little bit of bullet-proofing (eg: auto
nuke of stale locks)
2 space-based expire (eg: existing reap program, or even an
iterative find) invoked both by cron (or equivalent) plus
as part of the incoming news unpacking.
(already exists, can be kludged quickly enough)
3 some reader (rn/nn)
4 smail 2.5 whose paths file consists of the single line:
smart-host <admin'd site>!%s
5 some mail reader (mush et. al.)
6 Dynafeed or its equivalent (mail message saying "please send me
x.y.z")
7 A better editor than vi or emacs... (several do exist)
The above software is actually quite reliable by itself. Aside from
system/disk/power failures there's no reason to believe that these
cannot run indefinitely with little if any intervention. The latest
Cnews is essentially bullet-proof to malformed articles. Most of the
remaining problems people encounter would be eliminated by making
the "appliance" a binary package for specific hardware/OS platforms.
One need only ask of people running current cnews - are there
problems aside from configuration/integration or newsflood/disk overflows?
The one thing you've left out, and I've omitted (because I don't think
it's particularly necessary) in the list above is automated reporting
and reaction facilities for use by the system administrator. In
a network of nightly-only batch transfers, it's not going to be
practical for the admin to "log into" remote machines. Most users
probably wouldn't like it either.
I'm not going to say too much about remote monitoring or reaction
(because my company markets something that can be used for this ;-) other
than to say that news is an ideal transport mechanism for it (especially
with cnew's ability to drop in new control messages fairly easily)
The worst problems would be due to unexpected power failures (whether
intentional or not ;-). All of the above software (given a little
attention to stale lock nuking, and occasional full history rebuilds)
will survive abrupt stoppages provided that the filesystem itself does.
After all is said and done, though, I still tend to believe that
building something this way is asking for immediate obsolescence
(whether X.400 or not). Ie: ISDN is coming, two-way CATV and that I
would think that most telephone companies are doing the same thing ours
is - all new homes for the last several (10?) years are being wired for a
second telephone line anyways. That makes it easier to adopt the
compuserve model (central node, people login). With this, individual
central sites can support more users, the increment per-user on admin
time is greatly reduced, and the entry cost for a user is MUCH
lower.
--
Chris Lewis; UUCP: ...!{cunews,uunet,latour}!ecicrl!clewis;
DOMAIN: cle...@ferret.ocunix.on.ca; Phone: Canada 613 832-0541
Psroff info: psroff-requ...@ferret.ocunix.on.ca
Ferret mailing list: ferret-requ...@ferret.ocunix.on.ca
Path: gmdzi!unido!mcsun!uunet!cis.ohio-state.edu!pacific.mps.ohio-state.edu!linac!midway!machine!chinet!les
From: l...@chinet.chi.il.us (Leslie Mikesell)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <1991Aug07.154644.21114@chinet.chi.il.us>
Date: 7 Aug 91 15:46:44 GMT
References: <2397@ecicrl.ocunix.on.ca> <1991Aug3.193043.1862@newshost.anu.edu.au> <2413@ecicrl.ocunix.on.ca>
Organization: Chinet - Chicago public access UNIX
Lines: 66
In article <2...@ecicrl.ocunix.on.ca> cle...@ferret.ocunix.on.ca (Chris Lewis) writes:
>>>>The discussion was about the feasability of using current news and
>>>>mail software (enhanced) for naive users to operate their own single
>>>>user nodes that pass news feeds on to each other.
>After the last go round, *now* I see what you're driving at. USENET/Mail
>as an "appliance". Actually, I'm a more horrified at the class
>of users that that approach would bring to the net than of the
>technical difficulties.
I see this as a very interesting possibility too, but I would approach
it from a completely different angle. If you are going out to single
user nodes, (actually even if you don't...) the bandwidth is almost entirely
one way. How about setting up a broadcast mechanism which allows unlimited
numbers of receivers with no extra management requirements, then layer
email and news replies into a common user interface that uses a different
(point-to-point) transport for the much lower volume that needs the
return path. I suspect that there are many more people that read news
than ever post anything, so in many cases the broadcast reception would
be all that is needed (and the receive-only users wouldn't cause any
problems at all).
>The biggest technical difficulty is that of each of these appliances
>shipping news to each other.
And the solution is to not require that.
>For if the "appliance" is to behave
>anyways reasonably, given the current network load, each appliance
>would have to work with a partial feed matching their owner's interests,
>and that leads to extreme difficulties in trying to match everybody's
>wants with everyone else's newsgroup list.
Using a broadcast approach, everyone just sets their machine up to
ignore unwanted groups.
>After all is said and done, though, I still tend to believe that
>building something this way is asking for immediate obsolescence
>(whether X.400 or not). Ie: ISDN is coming, two-way CATV and that I
>would think that most telephone companies are doing the same thing ours
>is - all new homes for the last several (10?) years are being wired for a
>second telephone line anyways. That makes it easier to adopt the
>compuserve model (central node, people login). With this, individual
>central sites can support more users, the increment per-user on admin
>time is greatly reduced, and the entry cost for a user is MUCH
>lower.
I'd guess that several times the current volume of news posting and
email could be handled by existing equipment at uunet, attmail or
compuserve. The phone time for sending postings from any individual
would be almost negligable (on the average...) and certainly wouldn't
require a dedicated phone line for most people.
However it all boils down to economics. Based on a little experience
I'd say that it would currently be possible to set up a satellite feed
with receiving equipment costing around $1000 (dish, RF, and box with
some RAM). With a user base of a thousand or so it would probably
require a $25/month fee to support the administration and uplink costs.
Mass-market electronic items seem to take off at about the $500 point
so maybe we are still a few years away but it's time for someone to
have the design ready. Maybe something like this could be funded
for public libraries to start it off.
Les Mikesell
l...@chinet.chi.il.us
Path: gmdzi!unido!mcsun!uunet!cs.utexas.edu!utgpu!nrcnet0!cunews!micor!latour!ecicrl!clewis
From: cle...@ferret.ocunix.on.ca (Chris Lewis)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <2453@ecicrl.ocunix.on.ca>
Date: 12 Aug 91 04:17:47 GMT
References: <2397@ecicrl.ocunix.on.ca> <1991Aug3.193043.1862@newshost.anu.edu.au> <2413@ecicrl.ocunix.on.ca> <1991Aug07.154644.21114@chinet.chi.il.us>
Organization: Elegant Communications Inc., Ottawa, Canada
Lines: 40
In article <1991Aug07.154644.21...@chinet.chi.il.us> l...@chinet.chi.il.us (Leslie Mikesell) writes:
>In article <2...@ecicrl.ocunix.on.ca> cle...@ferret.ocunix.on.ca (Chris Lewis) writes:
>How about setting up a broadcast mechanism which allows unlimited
>numbers of receivers with no extra management requirements, then layer
>email and news replies into a common user interface that uses a different
>(point-to-point) transport for the much lower volume that needs the
>return path. I suspect that there are many more people that read news
>than ever post anything, so in many cases the broadcast reception would
>be all that is needed (and the receive-only users wouldn't cause any
>problems at all).
Well, I did mention 2-way CATV, which is actually a better solution -
You broadcast the signal during vertical retrace on a satellite channel
(CNN perhaps?), and everybody who receives that station via dish
or CATV (don't need the dish or receiver) can get a feed. What an idea!
Cheap! They might even do it for free because it's not being used
for anything else! Can I patent this idea?
"Stargate", the Oracle intones.
What?
"Stargate"...
Sigh. Stargate actually implemented all of this. Unfortunately, it
died due to the usual political bickering. They felt it should be
moderated groups only, which caused a ruckus. Then there was the
"no further distribution" rule (ala EUNET) that eventually killed it.
[Needed to cover the overhead costs]
Two-way CATV has the ability to feed the much smaller back-feed into
the network without relying on telephones.
Perhaps it's time to try the idea again. Stargate.com is still
out there.
--
Chris Lewis; UUCP: ...!{cunews,uunet,latour}!ecicrl!clewis;
DOMAIN: cle...@ferret.ocunix.on.ca; Phone: Canada 613 832-0541
Psroff info: psroff-requ...@ferret.ocunix.on.ca
Ferret mailing list: ferret-requ...@ferret.ocunix.on.ca
Path: gmdzi!unido!ira.uka.de!smurf.sub.org!urlichs
From: urli...@smurf.sub.org (Matthias Urlichs)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <9qfb=!#@smurf.sub.org>
Date: 14 Aug 91 13:02:07 GMT
References: <2413@ecicrl.ocunix.on.ca> <1991Aug07.154644.21114@chinet.chi.il.us> <2453@ecicrl.ocunix.on.ca>
Organization: University of Karlsruhe, FRG
Lines: 40
In news.admin, article <2...@ecicrl.ocunix.on.ca>,
cle...@ferret.ocunix.on.ca (Chris Lewis) writes:
< "Stargate", the Oracle intones.
< What?
< "Stargate"...
<
< Sigh. Stargate actually implemented all of this. Unfortunately, it
< died due to the usual political bickering. They felt it should be
< moderated groups only, which caused a ruckus. Then there was the
< "no further distribution" rule (ala EUNET) that eventually killed it.
< [Needed to cover the overhead costs]
<
Something like this seems to be up&running in Germany.
Unfortunately, it's not used for broadcasting NetNews --
it's uninteresting stuff like PD+Share-ware for Amiga/Atari/MeSs-DOS and
information from a few computer-related publishers, direct marketing people,
news and stock exchange wires, et al.
Most of this (except for the direct marketing of course...) is sent
encrypted.
The stuff runs on top of Pro7, one of Germany's private TV stations.
Allegedly, most of the cost is paid by the electronic equivalent of
commercials, and the commercial services using the channel.
The hardware costs DM 400 (about $200).
I assume that a big chunk of that hardware is dedicated to decryption of the
commercial stuff. You'd also need a TV tuner (VCR, whatever).
I don't have detailed info on transmissionmethod and speed; they say it's
more than one GByte/week, which translates to about 13300 bps.
I don't yet know what they do about error correction.
There probably aren't enough Usenet people in Germany (yet?) to start
thinking about using this service for News, assuming that they have free
bandwidth available... I'll ask them next week.
--
Matthias Urlichs -- urli...@smurf.sub.org -- urli...@smurf.ira.uka.de /(o\
Humboldtstrasse 7 -- 7500 Karlsruhe 1 -- Germany -- +49-721-9612521
Path: gmdzi!unido!fauern!ira.uka.de!sol.ctr.columbia.edu!spool.mu.edu!agate!forney.berkeley.edu!jbuck
From: jb...@forney.berkeley.edu (Joe Buck)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <1991Aug14.230402.28311@agate.berkeley.edu>
Date: 14 Aug 91 23:04:02 GMT
References: <1991Aug07.154644.21114@chinet.chi.il.us> <2453@ecicrl.ocunix.on.ca> <9qfb=!#@smurf.sub.org>
Sender: use...@agate.berkeley.edu (USENET Administrator)
Organization: University of California, Berkeley
Lines: 18
In article <9qfb...@smurf.sub.org> urli...@smurf.sub.org (Matthias Urlichs) writes:
>< Sigh. Stargate actually implemented all of this. Unfortunately, it
>< died due to the usual political bickering. They felt it should be
>< moderated groups only, which caused a ruckus. Then there was the
>< "no further distribution" rule (ala EUNET) that eventually killed it.
>< [Needed to cover the overhead costs]
While the two things you mention caused a lot of screaming, that's not
what killed it. It just wasn't economically feasible -- the price that
they needed to charge simply wasn't competitive, especially after uunet
came along -- uunet was offering a full feed for less money than
Stargate could offer moderated groups only for. Effectively, uunet
killed Stargate.
--
--
Joe Buck
jb...@galileo.berkeley.edu {uunet,ucbvax}!galileo.berkeley.edu!jbuck
Path: gmdzi!unido!fauern!ira.uka.de!sol.ctr.columbia.edu!zaphod.mps.ohio-state.edu!qt.cs.utexas.edu!cs.utexas.edu!utgpu!utzoo!henry
From: he...@zoo.toronto.edu (Henry Spencer)
Newsgroups: news.admin
Subject: Re: Requirements for General Public Access
Message-ID: <1991Aug14.234841.11304@zoo.toronto.edu>
Date: 14 Aug 91 23:48:41 GMT
References: <1991Aug07.154644.21114@chinet.chi.il.us> <2453@ecicrl.ocunix.on.ca> <9qfb=!#@smurf.sub.org> <1991Aug14.230402.28311@agate.berkeley.edu>
Organization: U of Toronto Zoology
Lines: 16
In article <1991Aug14.230402.28...@agate.berkeley.edu> jb...@forney.berkeley.edu (Joe Buck) writes:
>... uunet was offering a full feed for less money than
>Stargate could offer moderated groups only for. Effectively, uunet
>killed Stargate.
A common error. I made it too, but Rick set me straight: Stargate was
essentially dead well before UUNET got going. As I recall -- and my
memory of this is pretty rusty, mind you -- the ultimate problem was
that it relied on a sweetheart deal from a satellite-broadcasting outfit
that was willing to take a long-term view of the idea. Nobody else
would sell vertical-interval space at reasonable prices for a peculiar
startup venture. Something went wrong with either the company or the
deal and that was that.
--
Any program that calls itself an OS | Henry Spencer @ U of Toronto Zoology
(e.g. "MSDOS") isn't one. -Geoff Collyer| he...@zoo.toronto.edu utzoo!henry
|