W3C home > Mailing lists > Public > public-html-a11y@w3.org > January 2010

RE: Discussion on Change Proposal for ISSUE-66

From: John Foliot <jfoliot@stanford.edu>
Date: Fri, 22 Jan 2010 00:21:34 -0800 (PST)
To: "'Maciej Stachowiak'" <mjs@apple.com>
Cc: "'Matt May'" <mattmay@adobe.com>, "'HTML WG'" <public-html@w3.org>, <public-html-a11y@w3.org>
Message-ID: <047b01ca9b3b$e75f5200$b61df600$@edu>
Maciej Stachowiak wrote:
> I totally agree. The question is not at all about whether proper
> alternative text is required. It's just about repair techniques when
> alternative text is missing. That's why I suggested that the spec
> should be clear that any info gathered is an emergency repair technique

Actually Maciej, I'd agree to "...any info gathered is _available_ to an
emergency technique..." - but that the data alone is not a repair

> Let's say you actually have a page with an <img> that's missing alt
> text, and a blind user browses to it using their favorite AT+browser
> combo of choice. Are you saying you wouldn't want the browser or
> assistive technology to use the above techniques?

Guessing is not a technique. Heuristics is simply "best guess", or 
"... of or relating to exploratory problem-solving techniques that utilize
self-educating techniques (as the evaluation of feedback) to improve

...which requires cognition - no machine today can do this.

> Nobody is saying "substitute for". The question is just whether repair
> techniques are allowed in the face of broken content.

I think that perhaps I've not explained myself correctly, of course they
should be allowed.  I don't think they should be prescribed in the
specification, especially given that the root of this discussion, the
removal of language that referenced non-existent techniques, simply sought
to remove reference to said non-existent techniques. Nothing in the Change
Proposal suggests that *all* repairs techniques be foregone, only the
reference to 'Heuristics'. 

> I ask you again:
> are you saying that emergency repair techniques by the UA or AT should
> be explicitly banned? Note that this would make *all* existing AT
> noncompliant. (The most common current emergency repair for missing alt
> is to read the filename.)

I am not. But I have grave concerns that we reference in a Standard a
technique that is a non-technique, as it can foster a sense of complacency
in authors - it is but wishful thinking today.  

Suggestions such as Larry's to seek out DC metadata is actually a good
idea, and/but more importantly, how to do that, and the relevant
specifications exist today. That's the basis of a technique, not
suggesting that someday maybe it will all be done like magic, so we'll
allow that too.

I'm more skeptical of EXIF, but accept the possibility because it has
already been specified elsewhere (my impression is that most EXIF data
today is machine written, and thus also lacks the ability to express
intent - but I am hardly an EXIF expert and stand to be corrected if I am
wrong).  Keep the language of the specification in the present and not the
future and let's move on.  That's all.

> >
> > I think any statement that suggests futuristic magic as a potential
> > technical possibility inside a technical specification and standard
> is
> > wrong, and should be avoided.
> I believe what I suggested was exactly the opposite of that.

Then I think we are essentially in agreement.

Received on Friday, 22 January 2010 08:22:12 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:55:28 UTC