Re: Project Gutenberg's Roget's Thesaurus
Tim Berners-Lee <timbl@www3.cern.ch>
Date: Wed, 14 Apr 93 09:28:17 +0100
From: Tim Berners-Lee <timbl@www3.cern.ch>
Message-id: <9304140828.AA06019@www3.cern.ch>
To: "William M. Perry" <wmperry@papaya.ucs.indiana.edu>
Subject: Re: Project Gutenberg's Roget's Thesaurus
Cc: Nathan.Torkington@vuw.ac.nz, www-talk@nxoc01.cern.ch
Reply-To: timbl@nxoc01.cern.ch
| Date: Tue, 13 Apr 1993 11:14:42 -0500
| From: "William M. Perry" <wmperry@papaya.ucs.indiana.edu>
| Tim Berners-Lee writes:
| [...]
| >One possibility is we feed back the hypertext to Michael Hart.
| >It could just be an overlay of files which have pointers to
| >the actual text, or a smart server which can #include the
| >actual text into a wrapper document.
| [...]
|
| Are there any plans to have an html tag to include a file? That
way you
| wouldn't have to depend on a server being smart enough to include
the file.
| Something akin to the IMG tag now. ie: <INCLUDE SRC="XXXXXX">, and
| possibly <INCLUDE SRC="HTTP://fqdn:80/XXXXXX">.
If we were to adopt it, then the we should use the SGML entity
scheme. Basically, this involves declaring a new entity
(thing which is inserted with a & sign) as being an
external object with a given URL, and then making a reference
to that entity. But we're talking about something slightly different
from what I was suggesting, in that the these tags or entities
would be picked up by the client, whereas I was thinking
of something implemented on the server which would be
invisible to the client. It would be faster that way, and not
require any extensions either, like you could do it tomorrow with cpp
for example.
Tim