Re: color: NCSA Mosaic, Netscape, and HTML3

Ian S. Graham (igraham@alchemy.chem.utoronto.ca)
Tue, 18 Jul 95 17:58:31 EDT


From: "Ian S. Graham" <igraham@alchemy.chem.utoronto.ca>
Message-Id: <199507182158.RAA17078@alchemy.chem.utoronto.ca>
Subject: Re: color: NCSA Mosaic, Netscape, and HTML3
To: C.J.Tilbury@estate.warwick.ac.uk
Date: Tue, 18 Jul 95 17:58:31 EDT
Cc: www-html@www10.w3.org
In-Reply-To: <1CFB4747FA@forest.estate.warwick.ac.uk>; from "Chris Tilbury" at Jul 18, 95 4:23 pm

> 
> On 18 Jul 95 at 10:27, Ian S. Graham wrote:
> 
> [someone else wrote ...]
> > > Netscape color extensions (like many other netscape extensions)
> > > are hot air and presumptiously defined.
> > > 
> > > Unfortunately, style sheets are still being defined :)
>  
> > Hmm.... so something that is implemented, operational and popular is
> > "hot air", while "still being defined" stylesheets are not???


Considering what Netscape (and others) could have done, I find their
adaptations remarkably tame  - after all, they could have implemented
<ROLODEX> <CARD>.....</ROLODEX>..... ;-)

At the same time, their implementations reflect the slow evolution 
of HTML -- nothing has really changed in over a year -- and the
demands of authors. HTML 2.0 is still being finalized, HTML 3.0 is 
soon set to expire, and only now are issues such as charsets, 
internationaliazion, REL/REV, TABLES and FIGs being reworked for
addition into HTML 2.x.   At the same time, authors are demanding
things like better list design, better layout, and yes -- even
font control, none of which are supported by standard HTML.

ALso, I tend to think of most Web documents are transients,
particularly those for which formatting seems a high priority --
they live and die on rather short timescales. I don't think it 
unreasonable that, once HTML is augmented, and once stylesheet
languages (and support) are established,  that authors will turn to 
these more poweful solutions to attain the desired result.  So, I 
don't see Netscape as some anarchic force, but rather as a kick in the 
butt that just might push forward the development of a better HTML.

> No; Netscape's color extensions are hot air, and presumptiously 
> defined, because they [Netscape] seem to have (or have had) the idea 
> that by implementing these elements in their own browser - elements, 
> I might add, which are entirely platform-dependent and conflict 
> /wildly/ with some of the fundamentals principles guiding the 
> development of HTML, version 3, those being "Platform 
> Independence"[1], "Content not Presentation Markup"[2], and "Support 
.............
>
> On the other hand, stylesheets are not any of these precisely because 
> they are still being defined - no any one current implementation is 
> making any claims of being the definitive "standard", and in fact, by 
> their nature, every implementation could be completely different 
> without it mattering one iota. (The stylesheet for a graphical 
> browser is hardly going to be appropriate for a browser which uses 
> speech synthesis, for example).

Well, it would matter in that a U/A needs to be able to parse they
stylesheet, so there still need to be well-defined languages. Without
a defined standard, no-one will use it.  Better a funny netscape 
tag that works some of the time, than a stylesheet that works none
of the time.

> The fact that something is implemented, operational and popular does 
> not necessarily make it either appropriate or correct.

This could also apply to stylesheet languages......

Ian