> Has much thought gone into features needed to have data be both online
> and used to print books. When you are using a hypertext browser you
> want one view of the data and when you print a book you'll need another.
Thanks for pointing this out. Quite often I am browsing through some part of
the web and would like to print out a whole chunk. Unfortunately, while it is
simplicity itself to print the current document, it is really tough to print
out a collection of related material, involving multiple documents.
If we can address this issue well, it will also make it much easier to put
books and technical documentation etc. onto the web. By addressing the needs
of publishers the Web itself will become more pervasive.
You suggest:
> The most obvious thing to add would be:
> <ONLINE>click the mouse or select item</ONLINE>
> <PRINTED>see page such and such</PRINTED>
> But I have a strong feeling this isn't the correct way to solve the problem.
The issue you mention of how authors refer to hypertext links can be solved
through guidelines to authors on how to write such references in a form
suitable for both online and paper media. Tim has already written on this
in "http://info.cern.ch:80/hypertext/WWW/Provider/Style/Printable.html".
Perhaps we need some structuring elements which determine how a group of
related documents can be sequenced into a form suitable for printing, along
with a way of automatically generating a table of contents, and an index.
Hypertext links could then be converted into the appropriate page/section
references.
The obvious structuring element corresponds to the contents list, e.g.
<CONTENTS>
<A HREF ="doc1.html">Introduction to home brewing</A>
<A HREF ="doc2.html">Equipment Needed</A>
<A HREF ="doc1.html">The basic skills</A>
<A HREF ="doc1.html">Popular recipes</A>
</CONTENTS>
i.e. in SGML terms: <!ELEMENT CONTENTS (A+)>
This would be placed somewhere convenient with links back from other
documents using the <LINK> element, e.g.
"Previous" and "Next"
<LINK REL="CONTENTS" HREF="contents.html"> relationships could
also be useful here.
This would allow the browser to work out how to print the material as a book
from any one of the constituent documents. At worst, The index can be
generated automatically from analysing the document text for key phrases.
Better still, would be to have markup which identifies key words or phrases.
The header and inline emphasis tags are potential candidates, but I think
we should add a new tag to support professional quality indexing.
What do you think?
Regards,
Dave Raggett,
-----------------------------------------------------------------------------
Hewlett Packard Laboratories, +44 272 228046
Bristol, England dsr@hplb.hpl.hp.com