Why do people go for these patents?
More prior art: (but older) I'll write more than necessary just to make sure
people cool their jets when applying for patents...
At Digital in Valbonne, France, we in the CASEE group built and demonstrated
to a number of customers (including CERN....) a system almost identical to
the WWW in 1986 and 1987. The system, first called "Memex" and later
"Sunshine" supported links to embedded procedural objects from documents in
a variety of formats that were retrieved over DECnet. In fact, from a
document, you could link to anything that was runnable on a VMS machine. The
documents we typically demonstrated related to software development. We
often demonstrated a "page" of code which contained embedded links that
would bring up Digital's source code analyzer in a second window in the same
browser -- links from a requirements document to the code which implemented
the requirement. Links to and from images, with image maps (hotspots), etc.
were also supported.
This Memex system had almost the same concept of URL as is used in the WWW
today. This was important to us in being able to do the embedding discussed
above. Like a URL, the name of a thing started by specifing its processor
(program to run) and that was followed by opaque strings which were
processor specific. So, the "URL" for a VaxNotes entry looked something like
this: "notes atfab::sys$notes:worldwide.not:42.1" for the first reply to the
42nd note in the worldwide.not notesfile... The first version of Memex was
written as a combination of BLISS code and TPU macros. One of the processor
types we supported was, in fact, TPU code. Thus, it was possible to link to
a page which was actually TPU code that would get interpretted and you could
interact with (i.e. what JAVA does today).
Whatever we did with Memex, Sunshine, etc. it should also be noted that the
folk at MCC were working on Plaintext at about the same time (mid to late
80's). Although their system was much more limited in function than our work
in Valbonne (we supported wide area networks, for example, -- they didn't)
The MCC effort demonstrated most of the interesting hypertext/media concepts
. What wasn't built in code was described in reports -- and those reports
made most of what has followed "obvious to one practiced in the art."
Independent of "web-like stuff," Digital was building "compound documents"
back in 81 or so that would embed procedural objects inside normal documents
. This was much like OLE on Windows today. ie: you would have a text
document that had a "figure" that was actually a live spreadsheet. (I first
saw this at Digital as part of the OFIS project. Anton Chernof wrote the
fist serious demo of it I think -- using code from his "Digicalc", a
Visicalc ripoff.) This compound document stuff later got formalized into
various things like CDA, etc. Anyway, anytime you used something like Memex
to link to one of these compound documents, you were doing what the EOLAS
patents seems to claim they invented... Rubbish...
bob wyman