W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > July to September 2006

RE: 24 August 2006 Agenda

From: Bailey, Bruce <Bruce.Bailey@ed.gov>
Date: Wed, 23 Aug 2006 12:14:55 -0400
Message-ID: <CCDBDCBFA650F74AA88830D4BACDBAB5130FA5CE@wdcrobe2m02.ed.gov>
To: <w3c-wai-gl@w3.org>

> It's a very thorough report, Bruce.

Thanks Gez!  Now on-line at long and Tiny URLs:

>> Not represented in this sample are the number of people,
>> probably in the majority, who think using the title attribute
>> is a best practice.
> True, which is quite concerning as the title attribute is for
> supplementary information, and the HTML specification provides no
> guidance on how and when it should be rendered. [snip]

Agreed on all points.  I consider myself neutral on the topic.  The team here at ED (that does Web accessibility reviews) does not specifically consider title attributes except in exceptional circumstances.

>> It is reasonable to anticipate that UA (both browser and AT) 
>> support for title attribute will improve.

> Do you have evidence to back up that assertion?

Only circumstantial.  You might also recall the conviction I have communicated on the importance of syntax validity to accessibility.  I cannot prove that assertion either, but I think you have more sympathy for me on that point!

> I'm just curious why you're sure,

Well, we have seen support for the title attribute improve over time in association with frames and links and even images.  At the same time screen readers have improved their facility with data tables and form elements.  It is a reasonable extrapolation.  But hey, it is just my opinion.

> as there is no proper guidance on how it should be
> exposed to the user, 

The number of keystrokes aside, the Mozilla approach seems sound.  What if that were reduced to a single hot key combination?

> and your source from GW Micro appears to contradict it.

My source does not currently work for GW Micro, but he is reliable, brilliant, and lurks on this list!  That comment was strictly regarding the handling of the title attribute on form elements in a table.  I believe it is a mistake to use title just for the benefit of screen readers.  I am in favor of using title as supplemental information that can be available upon demand to *everyone*.  With regard to form elements in tables, the header information is already programmatically exposed.  The screen reader should be able to pick up on that.  (Not all can.)  Requiring redundant information in the title, to make up for known defects in the AT, does not seem like a viable *long term* solution to me.

Keeping on the theme of data tables, I think a reasonable use of titles could be *supplemental* annotations.  Imagine a web-base expense report.  The title attribute could be a pop-up note (rather than a link or superscript) explaining where a number came from.  The best way for the browser (and AT) to expose the availability of a title is subject to debate, but I don't think title should be there just for an AT user.  I am not sure it is a reasonable expectation for the screen reader to expose the title value all the time.  Since people with disabilities may disproportionately benefit from the supplemental information, I regard it as the correct domain of WCAG to include the technique.

> I thought Windows Eyes exposed the title attribute 
> anyway in the absence of a proper label?

I think it depends on the HTML element.  This is not a detail I have first hand knowledge of.

> Is it the supplementary rendering they're refusing - ie,
> if there's a label, it won't provide a separate
> mechanism to reveal the content of the title attribute?

Are you asking only about form elements?  I am not clear how robust WindowEyes is with the label tag.  I do not believe WindowEyes exposes title attribute values on form elements under *any* circumstances.

> The baseline can only list technologies, not user agents.

Quite right.  I had in mind conformance claims which include specific technology used for testing.  My apologies for using the wrong term, but I believe the whole issue of baseline and scope and conformance claims are still open.
Received on Wednesday, 23 August 2006 16:16:07 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:34:01 UTC