[6252] in www-talk@info.cern.ch
Re: Format Negotiation in Practice [Was: Versioning HTML at the server]
daemon@ATHENA.MIT.EDU (Brian Behlendorf)
Wed Oct 19 16:36:13 1994
Date: Wed, 19 Oct 1994 21:34:10 +0100
Errors-To: postmaster@www0.cern.ch
Errors-To: postmaster@www0.cern.ch
Reply-To: brian@wired.com
From: Brian Behlendorf <brian@wired.com>
To: Multiple recipients of list <www-talk@www0.cern.ch>
On Wed, 19 Oct 1994, Earl Hood wrote:
> > 2) A stylesheet definition language (DSSSL? I don't know anything
> > about it).
>
> How would the language support differing display devices: text,
> monochrome, color, displays for the deaf and blind?
For stylesheets, we don't really have a choice - all the content is in
the HTML itself, so conceivably one could even chuck the style sheet and
still get all the content in a reasonable format. Obviously
mapping <H1> to Helvetica 12 point isn't going to mean anything to a
blind person or someone on Lynx, but in this case I don't think that's
bad, as long as the HTML content is properly semantically marked up.
> > 3) A parser in every client.
>
> Alot of data to transmit when I connect to other-side of the world. I
> think supporting SGML directly will be nice, but extra data is required
> during transmission (the declaration, the DTD, the stylesheet, and the
> document instance). For big DTDs (and subsequently big stylesheets)
> this can be a pain. I'd hate having to wait for the HTML DTD to
> be sent to me evertime I retrieved an HTML document.
I don't think that's necessary - DTD's should be cached, even
persistantly disk-cached, and hopefully there'd be a common HTML 2.0
DTD, HTML 3.0 DTD, etc. Sites that had a different DTD for every
document had better have a good reason. Style sheets could be similarly
persistant.
Brian
