Re: Netscape & New HTML

Brian Behlendorf (brian@wired.com)
Wed, 26 Oct 1994 18:37:07 -0700 (PDT)


Date: Wed, 26 Oct 1994 18:37:07 -0700 (PDT)
From: Brian Behlendorf <brian@wired.com>
To: Bjoern Stabell <bjoerns@acm.org>
Cc: Multiple recipients of list <www-html@www0.cern.ch>
Subject: Re: Netscape & New HTML
In-Reply-To: <199410231447.AA188483647@diplom3.cs.uit.no>
Message-Id: <Pine.BSI.3.91.941026181059.17068A-100000@get.wired.com>

On Wed, 26 Oct 1994, Bjoern Stabell wrote:
> ] One question - just one...
> ] 
> ] > I've stated this myself a couple of times before.  Nothing prevents
> ] > developers of WWW clients to support other document languages
> ] > explicitly like PostScript, PDF, TeX, etc.
> ] 
> ] How do you place WWW 'compatible' hypertext links into a PostScript file or 
> ] PDF or TeX or...??? *Easily*...
> 
> Maybe you should take a look at the Hyper-G project; they're
> including their very own (and very good, anti-aliasing)
> PostScript renderer/viewer which supports hypertext links.

I spent some time looking into Hyper-G this past weekend - not enough to
get deep into the theory of it but enough to know that it definitely
merited closer attention.  Its creators believe it to be the next step in
multimedia after the HTTP/HTML model.  The thing that struck me the most
was the ability to create hypertext links on top of *any* arbitrary
bit of data - they have their own gif viewer and movie viewer to support
this.  It also looked like one could assign other attributes (title,
author, etc) onto arbitrary objects, too.  Anyways, it definitely merited
more investigation. 

> All this bitchering about what goes into HTML and what doesn't
> makes me wonder why we don't create two standards: presentation
> oriented HTML and semantically structured HTML.  With the typing
> capabilities of MIME, a browser can easily distinguish between
> and support multiple formats.

Let's just not call them both HTML.  I've sorta been thinking lately that
maybe there needs to be another division in HTML - that which the author
creates, and that which the server serves.  Server-side includes and
relative URL's are just two examples of a situation where what I write is
much different than what I want the user to see.  Mcom's size attributes
to the IMG tag is another example - it's clear that it's a win in terms of
performance, it's just something that document authors shouldn't have to
worry about.  

	Brian