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

Re: Using an image map for long described image links [Was: Revert Request]

From: Matthew Turvey <mcturvey@gmail.com>
Date: Thu, 2 Feb 2012 19:45:25 +0000
Message-ID: <CAFp5+AqYHkOidpe9r5Hir3_EY5s7tqvUcMeysmdjxhuLKP-rNg@mail.gmail.com>
To: John Foliot <john@foliot.ca>
Cc: Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com>, Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>, Silvia Pfeiffer <silviapfeiffer1@gmail.com>, Laura Carlson <laura.lee.carlson@gmail.com>, Sam Ruby <rubys@intertwingly.net>, Paul Cotton <Paul.Cotton@microsoft.com>, Maciej Stachowiak <mjs@apple.com>, HTML WG <public-html@w3.org>
On 2 February 2012 04:44, John Foliot <john@foliot.ca> wrote:
> Matthew Turvey wrote:
>>
>> I've already explained why I think a normal link on the image meets
>> these requirements:
>>
>> http://lists.w3.org/Archives/Public/public-html/2012Jan/0184.html
>
> ...and you have been repeatedly refuted on this simplistic claim.

By "these requirements" I'm referring to "the 3 key requirements" from
your previous message. Specifically:

> [...] the 3 key requirements of
> discoverability, choice to consume or not consume, and preservation of HTML
> structured content; requirements that using aria-describedby with hidden or
> off-screen content simply cannot deliver.

Note that I didn't suggest using aria-describedby with hidden or
off-screen content anywhere in my previous message, either.

> However, let's make this real:  On January 17th, 2011 Google celebrated the
> 50th Anniversary of John F. Kennedy's inaugural speech with one of their
> famous Google Doodles (seen here at the Internet Archive "WayBackMachine"
> http://web.archive.org/web/20110120204039/http://www.google.com/)
>
> On that day, clicking on the Doodle would take you to the following URL:
> http://www.google.com/search?q=John+F.+Kennedy+inaugural+address (please
> feel free to examine the source code at the above URL to verify this claim).
>
> Google also used the following code to actually insert the image into their
> home page:
>
>        <img alt="50th Anniversary of JFK&#39;s Inaugural Address"
> title="50th Anniversary of JFK&#39;s Inaugural Address" border=0 height=190
> src="...">
>
> Because the link that surrounded the image took you to the search page
> related to JFK's speech, the alt text was, for the most part correct.
>
> Sadly however, Google apparently doesn't think enough of the time, cost and
> effort they put into those famous doodles to bother sharing them with
> non-sighted users. If they did, they could have provided a long description
> similar to the following:
>
>        <p>Today's Google Doodle continues on the tradition of providing
> alternatives to the traditional G O O G L E logo on the home page. The
> letter G is an ASCII rendition of the G made from numerous words taken from
> <a
> href="http://www.americanrhetoric.com/speeches/jfkinaugural.htm"><abbr>JFK</
> abbr>'s Inaugural Speech</a>, arranged to form the actual letter. Some of
> the key blue words in this letter are ask, what, nation, citizens, greater,
> light, deeds and devotion, all in random order. The first O is composed of a
> similar set of words in red, with key words including you, can, do, globe,
> effort, progress, now and fruitful. The second O again follows this pattern,
> with orange key words including hope, anew, for, peace, help and truly. </p>
>
>        <p>The traditional lower case G is here a blue likeness of
> <abbr>JFK</abbr> himself. At first it, it appears to be a charcoal or pencil
> sketch, but on closer inspection you notice that it too is in fact an ASCII
> artwork, again using key words taken from <abbr>JFK</abbr>'s speech. Few
> actual words are discernible however. The final 2 letters, a green L and red
> E are again created using key words: the L containing your, unite, renew,
> pledge, deeds, and the E containing country, who, tap, sides, both for and
> your.</p>
>
>
> If this prose was put on a separate web page (JFK_doodle.html) how would you
> link it to the doodle?
>
> 1) You can't just simply make a "normal link" as you suggest, as the linking
> function is already reserved for the search result page. This is not some
> "made-up" use case, this is a real world, actual demonstration of the need,
> on arguably one of the most visited web pages on the internet.
>
> 2) You couldn't just park it off screen somewhere, or 'hide' if with
> @hidden, and then link to it with aria-describedby because:
>
>        a) You would lose both the hyperlink to the actual speech, as well
> as the semantic markup of <abbr>, as both would be flattened to string text.
>
>        b) If somehow you could overcome the flattening-to-string-text
> problem, to activate the hyperlink you must put tab-focus on the link - how
> do you focus on something that is hidden? And what of sighted users,
> (perhaps using a tool such as ZoomText Magnifier/Reader, which is both a
> screen reader and screen magnifier -
> http://www.aisquared.com/zoomtext/more/zoomtext_magnifier_reader/) who would
> hit the tab key and not see a visible tab focus (failing WCAG 2.4.7 Focus
> Visible: Any keyboard operable user interface has a mode of operation where
> the keyboard focus indicator is visible. (Level AA)) - what exactly should
> the magnifier magnify?
>
>        c) The aria-describedby attribute (as well as its companion
> aria-labelledby and aria-label attributes) are read as the Accessible Label
> in the Accessibility API's - those APIs do not recognize the fact that this
> is additional and supplemental information that should be served to the user
> on demand - they instead supply their labeling/describing text as part of
> the regular speech flow. The net effect is, even if I were to remove the
> hyperlink and <abbr> markup, the screen reader would announce the following:
>
>        "Link: 50th Anniversary of JFK's Inaugural Address[1] Today's Google
> doodle continues on the tradition of providing alternatives to the
> traditional G O O G L E logo on the home page.  The letter G is an ASCII
> rendition of the G made from numerous words taken from JFK's Inaugural
> Speech, arranged to form the actual letter. Some of the key blue words in
> this letter are ask, what, nation, citizens, greater, light, deeds and
> devotion, all in random order. The first O is composed of a similar set of
> words in red, with key words including you, can, do, globe, effort,
> progress, now and fruitful. The second O again follows this pattern, with
> orange key words including hope, anew, for, peace, help and truly. The
> traditional lower case G is here a blue likeness of JFK himself. At first
> it, it appears to be a charcoal or pencil sketch, but on closer inspection
> you notice that it too is in fact an ASCII artwork, again using key words
> taken from  JFK's speech. Few actual words are discernible however. The
> final 2 letters, a green L and red E are again created using key words: the
> L containing your, unite, renew, pledge, deeds, and the E containing
> country, who, tap, sides, both for and your."
>
> [1 Note, because there is no period at the end of the ALT text, it would
> result in a run-on sentence in the screen reader]
>
> Go ahead, read that out loud to yourself and then ask if that even remotely
> seems to be an appropriate user-experience. Every time.
>
> The problem with your suggestion also has the additional problem: No user,
> sighted or otherwise, would know prior to clicking on the image, where the
> link would be taking them to, and few users would be expecting a textual
> description of an image they just saw and clicked on.
>
> As shown, there are also real-world use cases where doing what you suggest
> as the appropriate solution simply will not work.

Note that the subject of this thread is, and I quote, "Using an image
map for long described image links".

Using an image map here would be a reasonable solution. Previous
Google doodles have had hidden functionality mapped to individual
letters for users to discover, so it's not like this is anything new.

Another solution would be to use multiple text links styled with CSS
to display parts of a background sprite that changes on hover and
focus. Another solution would be to put one link on the logo and then
provide the long description for everyone on the page it links to,
with the search results shown below the image and the description.
Another solution would be to provide a separate link, and optionally
refer to the link in the image alt; this could be positioned
off-screen with CSS and optionally displayed underneath the doodle on
hover and focus and/or displayed with a user option to use an
alternative "accessibility" style sheet, or it could be displayed at
the base of the page etc etc. We don't need longdesc here.

>> I think the poor quality of the longdesc-specific examples in the wild
>> confirms my own real world experience that this technique is never
>> needed.
>
> We can no longer give any credence to this line of argument, as multiple
> actors in the HTML5 evolution have stated that HTML5 is not just a capturing
> of the real internet as it is today (which includes BTW limited but real and
> appropriate use of @longdesc), but that it should also be a road-map to what
> the web should be:
>
> On 2012-01-06 Anne van Kesteren stated: "A standard sets a goal."
> https://www.w3.org/Bugs/Public/show_bug.cgi?id=13531#c8
>
> On 20-01-30 Jonas Sicking wrote: "... so far this has not happened.  So
> let's instead write the specification that will create the most accessible
> web once implemented..."
> http://lists.w3.org/Archives/Public/public-html/2012Jan/0199.html

You're comparing apples and oranges.

>> For the record, "d"-links never worked either, for obvious reasons:
>>
>> http://lists.w3.org/Archives/Public/w3c-wai-ig/1998AprJun/0096.html
>
> Correct, and both Kynn and Geoff acknowledge the problem of forced visual
> encumbrances on the web page in the email you reference, a requirement that
> you seem very confused about, even when you point to a 14 year old
> conversation that covered that very problem nearly a decade and a half ago.
> This is why many of us are still trying to remind people of the  problem of
> "forcing" a visual indication on the web-page.
>
> JF

Image maps do not force a "visual encumbrance" on the web page. That's
the whole point of this thread.

-Matt
Received on Thursday, 2 February 2012 19:45:54 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:17:43 GMT