No subject

when more than one language as to be used on a page.</font>
<br>
<br><font size=3D2 face=3D"sans-serif">Tim, I take your point about the lack
of a decent Irish-speaking synthesiser for a screen reader. My example
was certainly more theoretical than practical. I was wondering about the
problem from the point of view of a (hypothetical) web developer who is
required to use two languages for an image but wants to use valid mark-up
and follow the accessibility guidelines.</font>
<br>
<br><font size=3D2 face=3D"sans-serif">Barry and Andrew, I agree that the l=
ong
description attribute option is the best. Before posting my question, I
had dismissed the use of that attribute as a solution. It seems to be the
ideal solution. </font>
<br>
<br><font size=3D2 face=3D"sans-serif">I can't imagine that many web develo=
pers
will go to the bother of adding an extra attribute and page to a website
just to enable languages to be read out on synthesisers that don't yet
exist for screen readers - but that's not the point. The point is that
we can continue to tell web developers that existing web technologies *can*
cope with a requirement like the one I mentioned, and that there's no excuse
for dismissing WCAG or valid (X)HTML.</font>
<br>
<br><font size=3D2 face=3D"sans-serif">I wonder if anyone wants to discuss
how to mark up the title of this page http://www.pobail.ie/en/AnGhaeltacht/
on a Friday afternoon? Anyone? Anyone?! :)</font>
<br><font size=3D2 face=3D"sans-serif">__________<br>
Alan Dalton<br>
Accessibility Development Advisor, Excellence through Accessibility http://=
www.nda.ie/eta<br>
National Disability Authority, 25 Clyde Road, Dublin 4.<br>
E-mail: adalton at nda.ie<br>
Direct Line: 01-6080 406<br>
</font>
<br>
<br>
<br>
<table width=3D100%>
<tr valign=3Dtop>
<td width=3D40%><font size=3D1 face=3D"sans-serif"><b>&quot;Simius Puer&quo=
t;
&lt;simius.puer at googlemail.com&gt;</b> </font>
<br><font size=3D1 face=3D"sans-serif">Sent by: irl-dean-bounces at list.eeng.=
dcu.ie</font>
<p><font size=3D1 face=3D"sans-serif">13/08/2008 10:41</font>
<table border>
<tr valign=3Dtop>
<td bgcolor=3Dwhite>
<div align=3Dcenter><font size=3D1 face=3D"sans-serif">Please respond to<br>
Irish Design-for-all and e-Accessibility Network &nbsp; &nbsp; &lt;irl-dean=
@list.eeng.dcu.ie&gt;</font></div></table>
<br>
<td width=3D59%>
<table width=3D100%>
<tr valign=3Dtop>
<td>
<div align=3Dright><font size=3D1 face=3D"sans-serif">To</font></div>
<td><font size=3D1 face=3D"sans-serif">&quot;Irish Design-for-all and e-Acc=
essibility
Network&quot; &lt;irl-dean at list.eeng.dcu.ie&gt;</font>
<tr valign=3Dtop>
<td>
<div align=3Dright><font size=3D1 face=3D"sans-serif">cc</font></div>
<td>
<tr valign=3Dtop>
<td>
<div align=3Dright><font size=3D1 face=3D"sans-serif">Subject</font></div>
<td><font size=3D1 face=3D"sans-serif">Re: [Irl-dean] Identifying a language
change in an image's alternate text</font></table>
<br>
<table>
<tr valign=3Dtop>
<td>
<td></table>
<br></table>
<br>
<br>
<br><font size=3D3>I think you are spot on there Barry.<br>
<br>
It is not always necessary or even sensible to provide different language
versions, indeed the examples you cite would prohibit such an approach.
&nbsp;I often use small snippets of foreign languages such as Latin (e.g.
- e.g., etc, pro bono, etc...) or French (e.g. en route) and various others.
&nbsp;An appropriate &nbsp;&lt;span&gt; or &lt;abbr&gt; with both title=3D&=
quot;&quot;
and lang=3D&quot;&quot; &nbsp;attributes helps greatly - especially if you
add a class so that visual users can see that additional info is available
(we have to presume those using assistive technologies can access these)
by hovering the mouse over the text.<br>
<br>
Sticking to the core coding methods is a <b>must</b>. &nbsp;If people use
&quot;work-arounds&quot; or &quot;hacks&quot; then all they are doing is
creating confusion. &nbsp;If everyone stuck to core coding methods then
developers and users of assistive technologies would have a much easier
time.<br>
<br>
In the specific example given (mixed languages inside an attribute) HTML/XH=
TML
is limited in its ability to apply two separate lang=3D&quot;&quot; so the
longdesc=3D&quot;&quot; is the next best thing.<br>
<br>
We do have to remember that the W3C guidelines on accessibility should
be treated as best practice and that <i>every reasonable effort</i> is
made towards accessibility - not simply approached with the old &quot;tick-=
box
mentality&quot;. &nbsp;So, far from being a &quot;slightly contrary opinion=
&quot;,
I think you have suggested the perfect (and simplest) solution!<br>
<br>
</font><font size=3D2 face=3D"Verdana"><b><br>
_____________________<br>
Andrew Hart</b></font><font size=3D2 color=3D#666666 face=3D"Verdana"><b><b=
r>
Simius Web Ltd<br>
<br>
tel: +353 (0)91 395 586<br>
mob: +353 (0)871 264 728<br>
</b></font><font size=3D2 color=3Dblue face=3D"Verdana"><b><u><br>
</u></b></font><a href=3Dhttp://www.simiusweb.ie/><font size=3D2 color=3Dbl=
ue face=3D"Verdana"><b><u>http://www.simiusweb.ie</u></b></font></a><font s=
ize=3D3><br>
<br>
<br>
</font>
<br><font size=3D3>2008/8/13 Barry McMullin &lt;</font><a href=3Dmailto:bar=
ry.mcmullin at dcu.ie><font size=3D3 color=3Dblue><u>barry.mcmullin at dcu.ie</u>=
</font></a><font size=3D3>&gt;</font>
<br><font size=3D3>On Tue, 12 Aug 2008, ADalton at NDA.ie wrote:<br>
<br>
&gt; Is it possible within HTML to identify a language change that occurs<b=
r>
&gt; within an image's alternate text, e.g., &quot;&lt;img src=3D'council.p=
ng'<br>
&gt; alt=3D'County Council - Comhairle Chontae' /&gt;&quot;?<br>
</font>
<br><font size=3D3>Hi Folks -<br>
<br>
Just to throw in one slightly contrary opinion?<br>
<br>
The particular example raised by Alan may seem a bit trivial<br>
(appropriately conveying the bi-lingual text associated with a<br>
logo); but I think the general problem he raises is a genuine, if<br>
slightly subtle, one.<br>
<br>
Firstly, serveral responses have suggested that content authors<br>
should simply never mix languages in a single page. I agree that<br>
this mixing of languages is sometimes done inappropriately; but I<br>
can't agree that (on accessibility grounds) there should be a<br>
blanket prohibition on this. Indeed, I think there are many<br>
situations in which it is entirely appropriate to mix different<br>
languages in a single web page (language education materials<br>
being an obvious example, but far from the only such case).<br>
<br>
But simply mixing languages in not a problem in general (from the<br>
content author's point of view). &nbsp;One simply includes the<br>
appropriate lang attributes on relevant tags (possibly adding<br>
dedicated span or div tags if no other element already brackets<br>
the relevent text). &nbsp;Of course, as some responses have also<br>
pointed out, the support for this by assistive technologies is<br>
fairly variable; and even where the assistive technology has<br>
adequate support, specific users may not be proficient in using<br>
it (and may actually be hindered rather than helped as a<br>
result). But I don't think those issues should inhibit content<br>
authors from &quot;doing the right thing&quot; in terms of properly marking=
<br>
up language changes. &nbsp;It is a chicken and egg situation, but<br>
things can only improve if we start somewhere.<br>
<br>
So the problem (for content authors) is not with language changes<br>
in general; but with the specific case where the language change<br>
occurs *inside* an attribute value, specifically an alt<br>
attribute. &nbsp;Then we run into a genuine, if rather esoteric,<br>
design limitation of HTML, which is that it simply has no<br>
mechanism to represent this.<br>
<br>
Alan pointed out that some workarounds (or &quot;hacks&quot; if you prefer)=
<br>
have been proposed to still manage to encode the language change<br>
in such a way that there is some chance that assistive<br>
technologies may pick it up, and (hopefully) without doing any<br>
other harm. &nbsp;Basically this involves introducing otherwise<br>
unnecessary elements and splitting the original multi-lingual alt<br>
across several alts in those different elements, so that they can<br>
be associated, in turn, with different lang attributes. &nbsp;Yes,<br>
this is ugly. &nbsp;But no, it is not really comparable to the<br>
old-school &quot;spacer gifs&quot; and related hacks. &nbsp;Whereas those<b=
r>
techniques were basically contaminating the HTML with<br>
presentational information that could and should properly be<br>
represented in CSS, this multi-lingual technique is simply trying<br>
to represent genuine semantic information in the HTML, which is<br>
precisely where it ought to be - except that HTML is (arguably)<br>
deficient in supporting this.<br>
<br>
All that said, where do I stand on the specific example cited by<br>
Alan?<br>
<br>
I think the practical accessibility impact of not marking up this<br>
particular language change is as close to zero as makes no<br>
difference. So if the only concern is practical impact on<br>
accessibility, I would not spend any more time even thinking<br>
about this - there are far more substantive accessibility<br>
barriers to worry about.<br>
<br>
On the other hand, this would formally represent a failure of<br>
conformance with WCAG; so if the issue is that the client wants<br>
to be &quot;squeeky clean&quot; in claiming conformance then *some* sort
of<br>
workaround is needed. &nbsp;(I should say that, perhaps surprisingly,<br>
I actually have a lot of sympathy with the idea of being &quot;squeeky<br>
clean&quot; in this way, but that is a quite different discussion!)<br>
<br>
In that case, I think it would be reasonable to use the<br>
workaround mentioned by Alan (ugly as it is); though my own<br>
preferred workaround would be slighly more &quot;purist&quot;. &nbsp;I
would put<br>
only the English text (or whatever is the &quot;main&quot; language of
the<br>
page) in the alt attribute but add a longdesc attribute which<br>
points to somewhere with the full bi-lingual text; as the latter<br>
is plain HTML text, as opposed to an attribute value, it can use<br>
all normal HTML functionality, including div and span elements,<br>
and lang attributes to represent the language change. &nbsp;Now some<br>
may argue that this is not really a &quot;long description&quot;, as it
is<br>
still very short; but I would argue that the &quot;long&quot; in &quot;long=
desc&quot;<br>
is not solely about length but also about &quot;complexity&quot;; that
is,<br>
if one needs to include anything in a text alternative which<br>
cannot be represented in an alt attribute, then that already<br>
mandates the use of longdesc to get access to the full expressive<br>
power of HTML. Yes, this covers the case of a language change;<br>
but also some other cases, such as wanting to emphasise or to<br>
hypertext link some fragments of an alt text. Note that the<br>
latter is quite different from wanting to wrap a complete alt<br>
inside a hypertext link; this is most obvious if, say, one wanted<br>
to have two different hypertext links inside a single text<br>
alternative.<br>
<br>
In any case, I would certainly argue that this longdesc technique<br>
is no longer a &quot;hack&quot;, and should not be subject to the various<b=
r>
criticisms about reverting to the &quot;bad old days&quot;.<br>
<br>
Just my contrarian 2c worth...<br>
<br>
Best regards - Barry.</font><font size=3D3 color=3D#888888><br>
<br>
--<br>
Barry McMullin, Dublin City University<br>
 &nbsp;phone: +353-1-700-5432<br>
 &nbsp;web: </font><a href=3Dhttp://www.eeng.dcu.ie/%7Emcmullin/ target=3D_=
blank><font size=3D3 color=3Dblue><u>http://www.eeng.dcu.ie/~mcmullin/</u><=
/font></a>
<br><font size=3D3>_______________________________________________<br>
Irl-dean mailing list</font><font size=3D3 color=3Dblue><u><br>
</u></font><a href=3D"mailto:Irl-dean at list.eeng.dcu.ie"><font size=3D3 colo=
r=3Dblue><u>Irl-dean at list.eeng.dcu.ie</u></font></a><font size=3D3 color=3D=
blue><u><br>
</u></font><a href=3D"http://list.eeng.dcu.ie/mailman/listinfo/irl-dean" ta=
rget=3D_blank><font size=3D3 color=3Dblue><u>http://list.eeng.dcu.ie/mailma=
n/listinfo/irl-dean</u></font></a>
<br><font size=3D3><br>
<br>
<br>
-- <br>
_____________________<br>
Andrew Hart<br>
Simius Web Ltd<br>
<br>
tel: +353 (0)91 395 586<br>
mob: +353 (0)871 264 728<br>
</font><font size=3D3 color=3Dblue><u><br>
</u></font><a href=3Dhttp://www.simiusweb.ie/><font size=3D3 color=3Dblue><=
u>http://www.simiusweb.ie</u></font></a><font size=3D2><tt>________________=
_______________________________<br>
Irl-dean mailing list<br>
Irl-dean at list.eeng.dcu.ie<br>
http://list.eeng.dcu.ie/mailman/listinfo/irl-dean<br>
</tt></font>
<br>
<p><span style=3D"font-family:'Arial';font-size:8pt;">&quot;The information=
 transmitted is intended only for the person or entity to which it is addre=
ssed and may contain confidential and/or privileged material. Any review, r=
etransmission, dissemination or other use of, or taking of any action in re=
liance upon, this information by persons or entities other than the intende=
d recipient is prohibited. If you received this in error, please contact th=
e sender and delete the material from any computer. It is the policy of the=
 National Disability Authority to disallow the sending of offensive materia=
l and should you consider that the material contained in this message is of=
fensive you should contact the sender immediately and also mailminder at nda.g=
ov.ie&quot;.</span></p>
<p><span style=3D"font-family:'Arial';font-size:8pt;">&nbsp;</span></p>

--=_alternative 0040B346802574A6_=--