W3C home > Mailing lists > Public > www-html@w3.org > January 2000

Re: Why DOCTYPE Declarations for XHTML?

From: Murray Altheim <altheim@eng.sun.com>
Date: Tue, 18 Jan 2000 04:16:25 -0800
Message-ID: <38845999.59FACE5A@eng.sun.com>
To: Jan Roland Eriksson <rex@css.nu>
CC: www-html@w3.org
Jan Roland Eriksson wrote: 
[...]
> But basically, what you "have been told", from whatever source that may
> be, goes straight against what the world is saying in the ISO/IEC
> _international_ standard for HTML4.
> 
>   <URL:http://woodworm.cs.uml.edu/~rprice/15445/15445.html1>
> 
> (section 9.2 is of interest here)
> 
> And as both Mr Kimber and Arjun has pointed out on several occasions,
> the effort needed from the WG to create something of the same for XHTML
> is so fantastically simple, so we all can have the same level of
> recognition built into XHTML already from the start of it.
> 
> Let me quote Arjun here, since it seems to me that his request was
> modest and simple enough to be of use for all of us...
> 
>   "Is it too much to ask that the specs (say, the
>    Modularization document) provide for the definition
>    of a FPI with public text class NOTATION to identify
>    the "abstract document type"?  (E.g. in the "Naming Rules"
>    section, something like how replacing 'DTD' with 'NOTATION'
>    should be taken as the "official" definition?)"
> 
> (and you did ask for serious and constructive input, right?)

Well, since almost the very beginning of my involvement with the
W3C HTML WG I've had a module reserved for architectural use 
declarations. This module is still included in the distribution,
but it's empty. I early on asked David Megginson and I believe 
Eliot for assistance in filling in the details but did not receive
any such help -- they were too busy (I'm sure legitimately). I 
don't have my email going back that far but I may not have asked 
Eliot, so my apologies if not. I was asking just about anyone for
help, as I did not at the time feel qualified.

There was a period of time when I thought I could not only lobby 
the WG to include this module in the distribution  but add a 
section on architectures. I don't know now how that would be received.
And of course, this ignores *how* this might be attached to the 
instance unless all that is need is including it in the DTD (or
adding it optionally to an instance through a recognized method). 
The mechanisms that Arjun and Eliot are talking about are not to
my knowledge accepted by the W3C and I don't think it prudent to 
push this. If AF declarations are to be made available, it would 
have to continue to be somewhat a 'stealth' activity.

This week I'm travelling to Alexandria to attend a meeting on topic
maps, specified by a very recent ISO 13250. The ISO spec is pretty
much all about architectures, and is based very much on the same
concepts as what Eliot and Arjun are advocating as a functional 
technology. The spec is written by some real SGML geeks (no pejorative
intended - I have high respect for all of them). But do you think 
my proposal to the W3C
will be using AFs? No, I'll probably have to
hide them under the covers and use namespaced attributes. Hopefully
there is some way to find a middle ground that is politically feasible
and still maintains the necessary hooks into the ISO spec so that the
AF engines can make sense of things. That's what I was initially hoping
to do with XHTML, but I don't know how to pull that off anymore.

> "Dumb-asses" like my self, may have spent numbers of hours on Prof.
> Goldfarb's book, flipping the two built in bookmarks around as well as
> adding a few new once of our own, tried our best to understand good
> input from highly knowledgable people that we have been fortunate enough
> to meet on the web, in NG's and sometimes in private e-mail.

Well, count me among the dumb-asses, then.
 
> Still I don't know/understand more than just a fraction of the full
> story yet, but do you... (Murray and others unknown) really want me to
> go on learning the rest of it, knowing that somewhere along the line
> "they decided to obfuscate the truth for me" ?

The world is moving away from Goldfarb, like it or not. The political
winds in the W3C think namespaces will solve all of our problems, and
I'm mystified by that. Contrary to what one might think, I actually
*agree* with Eliot and Arjun to a great degree (as evidenced by the
fact that XHTML m12n snuck in an AF module), but what they're asking
is to my understanding simply not going to be accepted by the W3C. I'm
trying to be pragmatic about what can be both understood and accepted,
and have endeavoured to play both sides of this fence as best I can.

Standards work is at best a compromise, and I'm trying to provide
a simple architecture for XHTML that can be used in all sorts of 
ways within the limited scope of what can be provided by current
XML parsers: markup validation as according to the XML 1.0 REC. 
I'm already considered an SGML radical by many of them. Go figure.

Murray

...........................................................................
Murray Altheim                                   <mailto:altheim@sonic.net>
Member of Technical Staff, Tools Development & Support
Sun Microsystems, Inc. MS MPK17-102
1601 Willow Rd., Menlo Park, California 94025  <mailto:altheim@eng.sun.com>

   the honey bee is sad and cross and wicked as a weasel
   and when she perches on you boss she leaves a little measle -- archy
Received on Tuesday, 18 January 2000 07:18:10 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:15:40 GMT