Re: Font-style vs. phrase elements

Daniel W. Connolly (connolly@beach.w3.org)
Wed, 22 May 1996 08:10:20 -0400


Message-Id: <m0uMCk8-0002UPC@beach.w3.org>
To: "James K. Tauber" <jtauber@library.uwa.edu.au>
Cc: www-html@w3.org
Subject: Re: Font-style vs. phrase elements 
In-Reply-To: Your message of "Wed, 22 May 1996 17:34:20 +0800."
             <31A2DF9C.77BB@library.uwa.edu.au> 
Date: Wed, 22 May 1996 08:10:20 -0400
From: "Daniel W. Connolly" <connolly@beach.w3.org>

In message <31A2DF9C.77BB@library.uwa.edu.au>, "James K. Tauber" writes:
>Daniel W. Connolly wrote:
>> It could be called "HTML 3.2 Structured"
>> >
>> >Would anybody be interested?
>> 
>> Yes: especially with respect to interoperability among HTML
>> authoring tools.
>
>I'm not sure I understand what you mean. If authoring tools are 
>DTD-driven there's no problem.

Sure there is: I don't want DTD-driven tools writing <center>
when they could be writing <div align=center>. Or paragraphs
without <p> at the beginning. or <listing> or <xmp> ...

> If they aren't, well, I'd be willing to 
>guess that the majority of those tools don't conform to any DTD let 
>alone the existing HTML 3.2 draft.

You might be surprised. I routinely validate the output of GNNpress,
for example. With minor tweaks (which I report as bugs) it does
a pretty good job.

The DTD can give those folks something to shoot at. It allows
them to, for example, do automated regression testing to be sure
that every time their tool saves a document, it conforms.

After folks get burned and educated a bit more (like the netscape
quote bug fix, and the messes when loading hand-written HTML into
tools like GNNpress and Navigator Gold), the DTD may come to
serve as a way for authoring tools to say "that document doesn't
conform. I can't guarantee to load it properly. I'll do my
best to fix it, but you'd better review it closely."

Dan