W3C home > Mailing lists > Public > wai-xtech@w3.org > May 2008

RE: Is Flickr an Edge Case? (was Re: HTML Action Item 54)

From: John Foliot <foliot@wats.ca>
Date: Tue, 27 May 2008 10:36:57 -0700
To: "'L. David Baron'" <dbaron@dbaron.org>
Cc: "'Maciej Stachowiak'" <mjs@apple.com>, "'Karl Groves'" <karl.groves@ssbbartgroup.com>, "'Andrew Sidwell'" <w3c@andrewsidwell.co.uk>, <public-html@w3.org>, "'W3C WAI-XTECH'" <wai-xtech@w3.org>, <wai-liaison@w3.org>, "'HTML4All'" <list@html4all.org>, "'Matt Morgan-May'" <mattmay@adobe.com>
Message-ID: <00f501c8c020$44a85c70$6d01a8c0@stanford.edu>

L. David Baron wrote:
> 
> Applying all the requirements we apply to mass media to content
> creation for small audiences doesn't make sense.  We have to consider
> the costs and benefits of meeting these requirements.  If we enforce
> them on everyone, one thing we'll do is force a lot of this content
> off of the Web entirely, which would make it accessible to much fewer
> people.     

This is not what is being debated here however.  What is being suggested is
that the technical specification be written to open a loop-hole that so far
has been closed: images must contain @alt if they are to be deemed
conformant.  That millions of images lack @alt, or a valuable @alt value is
not open to discussion - I will concur that they exist.  This alone is not a
reason to reverse the course and suggest that it's somehow OK, so we'll
re-write the spec to say that it is.  It's not.  Since the current penalty
for not having @alt is... NOTHING... I cannot see how the new spec helps
anyone save those who want conformant code without doing all that is
required to ensure conformance.

We are talking about a technical specification here: black and white rules
that establish how to be conformant.  Sites and authors will then chose to
be conformant or be non-conformant.  Sites such as Flickr - if they *want*
to be conformant, will do what they can to ensure that from a "code"
perspective they are outputting correct code: if a code fragment requires a
string from an external author, that is beyond their control, but if the
conformance requirement exists that an attribute must exist, they can at
least ensure that the placeholder exists and a means to provide a value for
that attribute is present.

Today, for a web page to be "conformant" the specification calls for a DTD.
No DTD, not conformant.  Yet Google's pages have no DTD, and their web pages
"work" just fine: Google made a choice and that is theirs to make, but since
arguably *the* most visited webpage on the internet today is non-conformant,
then why are we insisting, even in HTML 5, for a DTD?

There is no "technical" reason to reverse the requirement for a mandatory
@alt save that it makes it easier to have conformant pages.  It does nothing
to improve accessibility, it does nothing to enhance or improve the next
generation of HTML, it does nothing for the very people who most need to
have a textual alternative to an image.  If, as suggested, most photos are
viewed by a very few (your telephone analogy), then what is wrong with
adding alt="" to those millions of images viewed by the very few?  The whole
argument falls flat on it's face.

JF
Received on Tuesday, 27 May 2008 17:37:58 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 13:15:49 GMT