W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > July to September 1998

Re: Consistency with HTML attributes

From: Al Gilman <asgilman@access.digex.net>
Date: Thu, 10 Sep 1998 14:55:35 -0400 (EDT)
Message-Id: <199809101855.OAA26398@access2.digex.net>
To: marja@w3.org (Marja-Riitta Koivunen)
Cc: w3c-wai-gl@w3.org
Thank you, Marja-Riita!

I want to re-affirm some of the things that you have said.

	It is hard for authors to follow the thread of what they
	need to do when the text backup for an image goes in
	different syntactic locations for images that arise in
	different places in the syntax.

	For the future we want to better capture the essence
	of the requirement into the formalism that defines
	web documents so that the casewise variation between
	different HTML elements (and yet more XML elements)
	is hidden from the people who have the natural
	language understanding to supply appropriate text.

In the WAI-CG we are discussing with the rest of the W3C
management how to build sensitivity to access drivers into the
future Web Document development process.

One approach that I can see for doing this is to define
abstract relationships (image to ALT-equivalent, image to
LONGDESC-equivalent) and have some form of binding of this
abstract semantic relationships to multiple 'concrete'
relationships as defined in the syntax of diverse DTDs. 

One open question is whether architectural forms are the right
formalism for capturing abstract relationships and binding them
to XML or SGML DTDs.  Another idea is that RDF assertions may
suffice.  It is going to take more work to figure this out.  We
are trying to make this this work as integral to the W3C
development of future formats as possible.  But do keep our feet
to the flames.

Thank you for the good summary.


to follow up on what Marja-Riitta Koivunen said:
> From w3c-wai-gl-request@w3.org  Thu Sep  3 15:35:56 1998
> Resent-Date: Thu, 3 Sep 1998 15:32:55 -0400 (EDT)
> Resent-Message-Id: <199809031932.PAA12367@www19.w3.org>
> Message-Id: <199809031933.PAA11003@www10.w3.org>
> X-Authentication-Warning: www10.w3.org: Host tundra.w3.org [] claimed to be tundra
> X-Sender: marja@
> X-Mailer: QUALCOMM Windows Eudora Pro Version 4.0
> Date: Thu, 03 Sep 1998 15:33:43 -0400
> To: w3c-wai-gl@w3.org
> From: Marja-Riitta Koivunen <marja@w3.org>
> In-Reply-To: <001601bdd1e4$8b3fba00$9d62fea9@gregg>
> Subject: Consistency with HTML attributes
> Resent-From: w3c-wai-gl@w3.org
> X-Mailing-List: <w3c-wai-gl@w3.org> archive/latest/780
> X-Loop: w3c-wai-gl@w3.org
> Sender: w3c-wai-gl-request@w3.org
> Resent-Sender: w3c-wai-gl-request@w3.org
> Precedence: list

> I like the new version, especially in the table form :-).
> One thing that bothers me still is that HTML does not use alt-text
> long-desc and maybe also the other attributes consistently. It would be so
> easy to say that with every object with an image or a video or animation
> use alt-text for purpose X and long-desc for purpose Y.
> Now we need to have separate techniques for OBJECTS that are not explained
> in guideline 2 techniques and with FRAME we need to explain that you have
> to use a HTML file just because src attribute does not take alt-text
> (guideline 5). And a FRAME does take a long-desc attribute.
> I think it is important to try to make these more consistent. Now these are
> harder to explain and there is more chance that different authoring tools
> interprete the attributes differently. If we later have other similar
> attributes things just become more and more complicated.
> As there are probably some things we cannot change right now, should we
> have an additional column in the table that explains how be wish these
> things would be done in the future versions of HTML?
>   Marja
Received on Thursday, 10 September 1998 14:54:53 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:27 UTC