W3C home > Mailing lists > Public > public-html@w3.org > February 2012

Re: Request to Reconsider Alt Guidance Location

From: Janina Sajka <janina@rednote.net>
Date: Fri, 24 Feb 2012 17:34:20 -0500
To: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
Cc: Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com>, Sam Ruby <rubys@intertwingly.net>, Maciej Stachowiak <mjs@apple.com>, Paul Cotton <Paul.Cotton@microsoft.com>, Michael Smith <mike@w3.org>, HTML Accessibility Task Force <public-html-a11y@w3.org>, public-html@w3.org, Judy Brewer <jbrewer@w3.org>, Steven Faulkner <faulkner.steve@gmail.com>
Message-ID: <20120224223420.GM4220@sonata.rednote.net>
Leif Halvard Silli writes:
> Janina Sajka, Thu, 23 Feb 2012 23:07:40 -0500:
> > Leif Halvard Silli writes:
> > But, if it's not specific to HTml 5, how can you justify trying to keep
> > it HTML 5 space?
> I can justify it by pointing out that, though the subject of 
> alternative text is general, Steve and Ian's texts on @alt speak about 
> alternative text in HTML5, and not everywhere else.
> > Do you expect people working with PDF, with SVG, with
> > ODF, with whatever other markup where this guidance is appropriate to
> > somehow know they need to find their way into HTML documents to get this
> > guidance?
> It is possible to create guidance that is specific to those other 
> languages you mention as well. Or even a some document that applies to 
> 'everything'.

This document applies to everything. The human side of the guidance it
provides does not differ. Only the lexical markup differs. It's not
practical to try and have documents for each technology, as the
explanations would be the same, and would constitute most of the text.
The lexical part is by far the smallest part. Furthermore, it's helpful
to both authors and developers to have lexical examples for multiple
languages together in the same place when the author/developer is
already familiar with one ml, the parallel usage, presented in parallel,
will only aid comprehension.

Please consider the job from the human perspective to get a better sense
of this.


> The ARIA spec is quite host language agnostic, but still uses HTML as 
> an example all the time. But even so, the process of getting it into 
> HTML5 took some time. If we had let ARIA live its own life and HTMl5 
> its own life, then I suspect that authors would have ended up not using 
> ARIA inside HTML as much as their presence in HTML5 will lead them to 
> do.
> Likewise, it does not sound like a win, for HTML5, if guidance about 
> how to make HTMl5 accessible is taken out of the HTML5 space and 
> 'generalized'.
> > How does that work when the well-known location of this work
> > has been WAI-WCAG for over 15 years? And, it seems the people willing to
> > continue working on the guidance are actually WCAG people, not HTML
> > people?
> It is well known that HTML speaks about how to use @alt. Are there 
> other formats which, on syntax level, makes it required to include 
> alternative text? Yes, the HTML5 validator do make it required, as 
> specced by HTML5 - it is not always permitted to drop it when the 
> author can't think anything to put there. Thus it is not true, when the 
> change proposal says that, quote: 'the HTML 5 draft long ago made the 
> alt attribute an optional feature of the img element'.
> > May I also point out that the author and editor of the Techniques
> > document himself requests this move? Does that count for nothing?
> To those that voted for note track, the probably yes. Meanwhile Steve's 
> request was conditional: 
> http://www.w3.org/mid/CA+ri+Vmd9w9_dOTHJU=mq42M-s7uD=b05cec7RzJhWjOkQ6OAg@mail.gmail.com
> >> In a way, what you say, is that Ian and Steve's two spec's are so good 
> > 
> > No, I'm sorry. They disagree. Ian's must go.
> They agree about much. But even Steve's draft is colored by the HTML5 
> spec. 
> In Steve's draft, Steve has 4 notes about 'willful violations' of 
> HTML5. [And a fifth about willful violation of HTML the  living 
> standard.]
> * 3 of these violations were using @alt not for alternative text but to 
> form an implicit relationship between an IMG and a caption - I don't 
> there is even a change proposal for this ... ? 
> * For 2 of these 3 implicit relationship violations, the justification 
> for the violation is that it is 'recommended as a bridging technique 
> until the figure and figcaption element semantics are implemented in 
> browsers and assistive technology'. Since it is only a temporary advice 
> the violation does not seem very serious.
> * The fourth violation is about @title as fallback for @alt, for which 
> there is a decision process already. 
> I know there are other disagreement - at least one issue about how to 
> mark up a captcha accessible, but it seems that Steve does not see the 
> disagreement as serious enough to describe it as a willful violation - 
> at least not yet.
> >> with regard to @alt, that what their content should be lifted out of 
> >> HTML5 and be made applicable for — I guess — images in general. So on 
> >> one side the CP focuses on problems within the HTML working group, on 
> >> the other side it says that is has produced good and useful texts on 
> >> @alt.
> >> 
> > Yes, Steve, with the help of other from the accessibility group has
> > produced very valuable documentation that needs to be widely shared and
> > properly maintaned. This is not a minor issue for a11y.
> > 
> > As the CP explains, it was produced in this space because Ian and others
> > here were getting so much about alternative text wrong. But, that's not
> > a reason to hang on to it like some kind of trophy.
> It is fully possible to fork it, I think - no need to ask for 
> permission for that, I think. And I by no means see it as trophy.
> >> At any rate: The document 'HTML5 techniques for providing useful text 
> >> alternatives' remains a HTML5 specific document. Why else should the 
> >> HTML working group be co-responsible for it?
> >>
> > No, it's not HTML specific. Perhaps its examples are HTML specific at
> > this time, because this was its initial audience. But, the examples, and
> > its title, can and must be globalized.
> If one drops 'HTML5' from the title of the document, then it sounds as 
> if applies to any kind of mark-up yes. But there is nothing about PDF, 
> SVG or ODF in that document, as far as I can tell. [HTML is often 
> converted to those formats and vice versa, so HTML5 is likely to impact 
> those formats anyhow.]
> -- 
> Leif Halvard Silli


Janina Sajka,	Phone:	+1.443.300.2200

Chair, Open Accessibility	janina@a11y.org	
Linux Foundation		http://a11y.org

Chair, Protocols & Formats
Web Accessibility Initiative	http://www.w3.org/wai/pf
World Wide Web Consortium (W3C)
Received on Friday, 24 February 2012 23:07:23 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:16:20 UTC