E-mail List Archives
Thread: JAWS and relative positioning
Number of posts in this thread: 24 (In chronological order)
From: St
Date: Tue, May 30 2006 1:40PM
Subject: JAWS and relative positioning
No previous message | Next message →
Hi all,
I've just noticed a weird behaviour:
On http://www.parisweb2006.org/ we're using position:relative in a few
places to nudge blocks a bit to the right (accessibility bar) or a bit
to the top (navbar).
I put up the site and because I wanted to do things right I of course
tested it in JAWS 5, and everything was fine.
But then I tested it this week with JAWS 6.2 and 7, and the link list is
desperately short. It doesn't seem to be able to extract the
relatively-positioned links in the page.
I've done some testing and it seems that he problem really lies with
relative positioning, and I can't begin to understand how *that* would
be a problem in the first place!
What do you think? Is this something anyone's already stumbled upon?
--
St
From: Steven Faulkner
Date: Tue, May 30 2006 5:10PM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
Hi St
From: Stephane Deschamps
Date: Wed, May 31 2006 1:40AM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
<quote who="Steven Faulkner">
> Hi St
From: Michael Stenitzer
Date: Wed, May 31 2006 2:10AM
Subject: Re: JAWS and relative positioning
← Previous message | Next message →
On 5/31/06, Stephane Deschamps < = EMAIL ADDRESS REMOVED = > wrote:
> <quote who="Steven Faulkner">
> >
> > In JAWS 6.10 a new feature was added for JAWS support of style sheets:
> >
> >>From the JAWS documentation:
> >
> > "Control How JAWS Uses Style Sheets
[...]
>
> Thanks for the info. The sad thing is that I'm not too much in favour of
> telling the visitors straight away "hey, this site uses advanced CSS, and
> since you're using a relatively new version of JAWS you should press Ins+V
> otherwise this site will be unusable".
>
> I really find it weird. I'm beginning to consider this as a reportable
> bug, and I'll probably end up writing to the JAWS people asking for
> complementary info.
>
has anybody on this list experienced similar problems? are there
general rules when this problem might occur?
this might be a real set-back for advanced accessible css.
/michael
--
Michael Stenitzer | WIENFLUSS information.design.solutions
www.wienfluss.net | linke wienzeile 178 / 3 / 142 | wien AT
fon ++43 650 wfluss 0 | fax ++43 1 4854200 31
From: Joshue O Connor
Date: Wed, May 31 2006 2:20AM
Subject: Re: JAWS and relative positioning
← Previous message | Next message →
Hi Steven, Stephane,
> The sad thing is that I'm not too much in favour of
> telling the visitors straight away "hey, this site uses advanced CSS, and
> since you're using a relatively new version of JAWS you should press Ins+V
> otherwise this site will be unusable".
I have to agree. Though obviously a user agent must develop as technologies and "standards"
improve, it' s such a pity that older user agents user are often left out of new "developments",
or forced into (often unobtainable) expensive upgrades to keep up.
Also, its not good to fall into the trap of being very "JAWScentric", you must ask yourself, 'How does IBM Homereader deal with the same issue, or Window Eyes?".
If it is a case that advanced CSS features are causing these kind of problems, I just
would not use them, especially if the sites you develop are to explicitly serve clients with disabilities.
Best Regards
Josh
Joshue O Connor
Senior Web Accessibility Consultant
Centre for Inclusive Technology (CFIT)
National Council for the Blind of Ireland
Website: http://www.cfit.ie
E-Mail: = EMAIL ADDRESS REMOVED =
Tel: 01 8821980
Stephane Deschamps wrote:
> <quote who="Steven Faulkner">
>
>>Hi St
From: Stephane Deschamps
Date: Wed, May 31 2006 6:50AM
Subject: Re: JAWS and relative positioning
← Previous message | Next message →
<quote who="Joshue O Connor">
> Also, its not good to fall into the trap of being very "JAWScentric", you
> must ask yourself, 'How does IBM Homereader deal with the same issue, or
> Window Eyes?".
Yes, but in France JAWS is very big, so I have to be a bit "JAWScentric" ;)
> If it is a case that advanced CSS features are causing these kind of
> problems, I just
> would not use them, especially if the sites you develop are to explicitly
> serve clients with disabilities.
Hm. Actually the site addresses quality web development, accessibility and
design, and showcases a conference on those topics (a French atmedia, if
you like). So one of our criteria for the HTML/CSS combo was: use the HTML
you've got as much as you can without adding too much design-oriented
clutter in the code.
I guess one can't win on both levels. That's sad, and all the more sad
than JAWS 5 is OK with that. It's a bug/feature situation :(
I'll leave the code as it is for the moment. Anybody from Freedom
Scientific on this list?
--
Stephane Deschamps
Paris Web 2006 :
http://www.parisweb2006.org/
(qualit
From: Derek Featherstone
Date: Wed, May 31 2006 7:20AM
Subject: Re: JAWS and relative positioning
← Previous message | Next message →
On 5/31/06, Stephane Deschamps wrote:
>So one of our criteria for the HTML/CSS combo was: use the HTML
>you've got as much as you can without adding too much design-oriented
>clutter in the code.
>
>I guess one can't win on both levels. That's sad, and all the more sad
>than JAWS 5 is OK with that. It's a bug/feature situation :(
Hi Stephane,
If you don't mind, I'd like to use your site as a test case. Clearly
there is *something* not right, but I'd like to use our resources (the
Web Standards Project - Accessibility Task Force) to help derive a very
simple "test case" that demonstrates the problem. I use relatively
positioned elements all the time and haven't come across the problem
that your site demonstrates.
We are currently contacting numerous AT vendors to see if they want to
collaborate with us to test their product against a "test suite" of
cases that we prepare and it seems that yours is a good one.
Would you be able/williing to allow us to manipulate the code on a copy
of your site to help determine the problem?
Cheers,
Derek.
--
Derek Featherstone = EMAIL ADDRESS REMOVED =
tel: 613-599-9784 1-866-932-4878 (toll-free in North America)
Web Development: http://www.furtherahead.com
Personal: http://www.boxofchocolates.ca
Web Standards: http://www.webstandards.org
From: Joshue O Connor
Date: Wed, May 31 2006 7:30AM
Subject: Re: JAWS and relative positioning
← Previous message | Next message →
> Yes, but in France JAWS is very big, so I have to be a bit "JAWScentric"
In Ireland it is the same, with the majority of our blind users using JAWS.
So I am also, for better or worse, JAWScentric.
> So one of our criteria for the HTML/CSS combo was: use the HTML
>> you've got as much as you can without adding too much design-oriented
>> clutter in the code.
Well then you should be OK avec/sans CSS. If the HTML is clean, then it should be
handled well by most screen readers.
Thanks for mentioning the bug, and good luck with your project.
Josh
Stephane Deschamps wrote:
> <quote who="Joshue O Connor">
>
>>Also, its not good to fall into the trap of being very "JAWScentric", you
>>must ask yourself, 'How does IBM Homereader deal with the same issue, or
>>Window Eyes?".
>
>
> Yes, but in France JAWS is very big, so I have to be a bit "JAWScentric" ;)
>
>
>>If it is a case that advanced CSS features are causing these kind of
>>problems, I just
>>would not use them, especially if the sites you develop are to explicitly
>>serve clients with disabilities.
>
>
> Hm. Actually the site addresses quality web development, accessibility and
> design, and showcases a conference on those topics (a French atmedia, if
> you like). So one of our criteria for the HTML/CSS combo was: use the HTML
> you've got as much as you can without adding too much design-oriented
> clutter in the code.
>
> I guess one can't win on both levels. That's sad, and all the more sad
> than JAWS 5 is OK with that. It's a bug/feature situation :(
>
> I'll leave the code as it is for the moment. Anybody from Freedom
> Scientific on this list?
>
From: Jim Thatcher
Date: Wed, May 31 2006 8:10AM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
Very interesting! The versions of Window-Eyes (5.5), Hal (6.5), and HPR
(3.04) that I have all find all the links. Hal's highlighting doesn't work
correctly but HPR's does.
Jim
Accessibility Consulting: http://jimthatcher.com/
512-306-0931
From: Jim Thatcher
Date: Wed, May 31 2006 8:20AM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
Hi Steven,
> This info may be felpful [sic] in understanding why the issue occurs in
> later versions of JAWS.
Were you able to change the behavior of JAWS using this functionality?
Jim
Accessibility Consulting: http://jimthatcher.com/
512-306-0931
From: Andrew Kirkpatrick
Date: Wed, May 31 2006 9:10AM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
This issue is caused, at least in part by this line in the CSS:
div#access, div#sidebar, p.backtotop { display:none; }
JAWS is honoring that in the newer versions and not earlier. This is in line with my test results at http://webaccessibility.info/lab/displaytest.html
This is a good change in JAWS. If you set the stylesheet processing back to the old way, you'll get the same undesireable results we used to deal with. Currently, if you set something to display:none, it is not displayed at all - visually or aurally. There is something else going on that I don't have the time to dig into, but I'm wondering why I can still see and tab to these anchors even with the display:none style - is the container set to display:none and the contents are otherwise defined and there is an issue as a result?
AWK
>
From: Jared Smith
Date: Wed, May 31 2006 9:20AM
Subject: Re: JAWS and relative positioning
← Previous message | Next message →
Stephane Deschamps wrote:
> Anybody from Freedom Scientific on this list?
Good question. There is one on this list. But I rarely see activity from
any of the major screen reader development companies on ANY of the
accessibility lists. But that's an entirely different issue.
This is certainly an interesting case. And from my testing it could
certainly disadvantage JAWS users accessing more advanced CSS enabled
sites unless it is either fixed or screen reader users are educated about
these new feature behaviors. Any guesses as to which one would be most
effective?
I'll work on putting together some test cases and see if other CSS
practices invoke this behavior.
Jared
From: Stephane Deschamps
Date: Wed, May 31 2006 9:30AM
Subject: Re: JAWS and relative positioning
← Previous message | Next message →
<quote who="Derek Featherstone">
> Hi Stephane,
>
> If you don't mind, I'd like to use your site as a test case. Clearly
> there is *something* not right, but I'd like to use our resources (the
> Web Standards Project - Accessibility Task Force) to help derive a very
> simple "test case" that demonstrates the problem. I use relatively
> positioned elements all the time and haven't come across the problem
> that your site demonstrates.
>
> We are currently contacting numerous AT vendors to see if they want to
> collaborate with us to test their product against a "test suite" of
> cases that we prepare and it seems that yours is a good one.
>
> Would you be able/williing to allow us to manipulate the code on a copy
> of your site to help determine the problem?
Oh yes, by all means do. Contact me offlist if you want plain-vanilla
sources etc.
--
Stephane Deschamps
Paris Web 2006 :
http://www.parisweb2006.org/
(qualit
From: Stephane Deschamps
Date: Wed, May 31 2006 9:40AM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
<quote who="Andrew Kirkpatrick">
> This issue is caused, at least in part by this line in the CSS:
> div#access, div#sidebar, p.backtotop { display:none; }
>
> JAWS is honoring that in the newer versions and not earlier. This is in
> line with my test results at
> http://webaccessibility.info/lab/displaytest.html
>
> This is a good change in JAWS. [...]
Hi Andrew,
Thanks for taking the time for this, but actually the display:none is in
the @media print section of the CSS, so the problem *may* be connected
because the problematic links are indeed in these three containers, but
then it would be a true hair-splitter.
--
Stephane Deschamps
Paris Web 2006 :
http://www.parisweb2006.org/
(qualit
From: Andrew Kirkpatrick
Date: Wed, May 31 2006 9:50AM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
Yep, missed that. Sounds like a JAWS CSS parsing issue. I moved the line out of the @media section and it removes the sidebar visually as well as for JAWS,
Seems fixed in JAWS 7.1.
You might want to try moving reapplying the display property below the @media section, just to see if that solves the issue. A hack, but it might do it.
AWK
>
From: Steven Faulkner
Date: Wed, May 31 2006 5:00PM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
Hi Jim, I haven't tested it.
I just remembered that I had read this info about this new feature in
the JAWS documentation, and thought it may be of some relevance.
with regards
Steven Faulkner
Web Accessibility Consultant
vision australia - information & library service
454 Glenferrie Road
Kooyong Victoria 3144
Phone: (613) 9864 9281
Fax: (613) 9864 9210
Email: = EMAIL ADDRESS REMOVED =
www.accessibleinfo.org.au | www.wat-c.org
Download the Web Accessibility Toolbar
[http://www.visionaustralia.org.au/ais/toolbar/]
Vision Australia was formed through the merger of the Royal Blind
Society
NSW, the Royal Victorian Institute for the Blind, Vision Australia
Foundation and the National Information & Library Service.
ABN: 67 108 391 831 ACN: 108 391 831
>
From: Patrick Burke
Date: Wed, May 31 2006 5:30PM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
Hello everyone,
Yes, the Jaws Verbosity setting for "Style Sheet
Processing" does produce different results, as
the documentation describes. (I'm using Jaws 7.0,
not 7.1). On St
From: Alastair Campbell
Date: Thu, Jun 01 2006 4:50AM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
Stephane Deschamps wrote:
> the display:none is in
> the @media print section of the CSS, so the problem *may* be connected
> because the problematic links are indeed in these three
> containers
I've had an issue with Jaws before where it used the print styles when
they were included in the main style sheet. It didn't respect the @media
{} within a regular style sheet, however, I thought that had been fixed.
The relative positioning one would be new to this version, but not
unexpected due to IE (I'm making several assumptions about the original
question!).
On a previous design of our company site the main content area was
relatively positioned down by about 50 pixels. In IE, you couldn't
select the text in that area with the mouse. With Browsealoud (and IE)
it wouldn't read out that section of text.
It probably has the same root cause and fix as this:
http://www.positioniseverything.net/explorer/unscrollable.html
I believe Jaws is using the DOM from the browser now (depending on
mode), so I would guess that issues like this will become more common
and be different depending on the browser (IE6, IE7 or Firefox).
Given that, I think Derek's suggestion for a set of test cases is going
to be vital.
Kind regards,
-Alastair
--
Alastair Campbell | Director of User Experience
Nomensa Email Disclaimer:
http://www.nomensa.com/email-disclaimer.html
From: Stephane Deschamps
Date: Thu, Jun 01 2006 7:50AM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
<quote who="Alastair Campbell">
> Stephane Deschamps wrote:
>> the display:none is in
>> the @media print section of the CSS, so the problem *may* be connected
>> because the problematic links are indeed in these three
>> containers
>
> I've had an issue with Jaws before where it used the print styles when
> they were included in the main style sheet. It didn't respect the @media
> {} within a regular style sheet, however, I thought that had been fixed.
>
> The relative positioning one would be new to this version, but not
> unexpected due to IE (I'm making several assumptions about the original
> question!).
... and the winner is: Alastair and Andrew, ex-aequo.
Andrew found the display:none and I though it had nothing to do (print
being print and screen being screen). But last night I pushed the @media
print to the top of the stylesheet, and the problem is solved: the links
are listed as should be.
So as a conclusion: the bug is not relative positioning, it's the @media
print instruction that's neglected. To avoid it when several @media
declarations exist, make the print the first one, then insert the others
(screen, projection, etc).
(I can't show you the result, it's on a test server, but it's going to be
online within a day or two).
--
Stephane Deschamps
Paris Web 2006 :
http://www.parisweb2006.org/
(qualit
From: Alastair Campbell
Date: Thu, Jun 01 2006 8:20AM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
> I pushed the @media
> print to the top of the stylesheet, and the problem is
> solved: the links are listed as should be.
But are they shown correctly on screen now?
There isn't anything technically wrong with what you did (from your
description), Jaws shouldn't use the print media.
I've found it easiest to include print style sheets separately, as the
last style sheet to ensure correct inheritance. Then you are just
tweaking the main style sheet for print, rather than starting from
scratch.
Kind regards,
-Alastair
--
Alastair Campbell | Director of User Experience
Nomensa Email Disclaimer:
http://www.nomensa.com/email-disclaimer.html
From: Andrew Kirkpatrick
Date: Thu, Jun 01 2006 9:10AM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
Glad you got it to work. There is definitely some funkiness with the
way that JAWS deals with @media rules. The attached example is fun to
look at. If you take the styles out of the screen @media rules you get
very different results than if you leave it as is. Very weird.
This is apparantly all fixed in the JAWS 7.1 beta, fortunately.
AWK
>
From: St
Date: Thu, Jun 01 2006 3:10PM
Subject: Re: JAWS and relative positioning
← Previous message | Next message →
<quote who='Alastair Campbell' when='01/06/2006 16:10'>
>> I pushed the @media
>> print to the top of the stylesheet, and the problem is
>> solved: the links are listed as should be.
>
> But are they shown correctly on screen now?
Yes, look here: http://www.parisweb2006.org/
I've just reordered the CSS. Funny, eh?
> There isn't anything technically wrong with what you did (from your
> description), Jaws shouldn't use the print media.
>
> I've found it easiest to include print style sheets separately, as the
> last style sheet to ensure correct inheritance. Then you are just
> tweaking the main style sheet for print, rather than starting from
> scratch.
Yup, a lesson harshly learnt...
--
St
From: smithj7
Date: Mon, Jun 05 2006 4:30PM
Subject: RE: JAWS and relative positioning
← Previous message | Next message →
Where I work we have a person that is a beta tester for Jaws. (He also
is a friend of mine a former blind programer and in charge of our
Management Information System section.) Also because I work for Florida
Blind Services and Freedom Scientific is a Florida company we actually
know many people from Freedom Scientific. Do you want me to contact
someone for this issue? I can ask my friend to join to discuss this
issue. It appears that you have somewhat solved the issue.
From: St
Date: Thu, Jun 08 2006 12:30AM
Subject: Re: JAWS and relative positioning
← Previous message | No next message
<quote who='smithj7' when='06/06/2006 00:20'>
> Where I work we have a person that is a beta tester for Jaws. (He also
> is a friend of mine a former blind programer and in charge of our
> Management Information System section.) Also because I work for Florida
> Blind Services and Freedom Scientific is a Florida company we actually
> know many people from Freedom Scientific. Do you want me to contact
> someone for this issue? I can ask my friend to join to discuss this
> issue. It appears that you have somewhat solved the issue.
Yes, the 'somewhat' is actually a bit awkward. Jaws understands CSS, but
partially only. So for example the @media directive is understood
wrongly, so if my @media print declaration is *below* the screen part in
my CSS, and for example the navbar has a display:none for print, then it
won't be read in Jaws, it seems.
I had to move the print declarations at the top of the stylesheet,
before the screen declaration, and then it worked.
Someone on this list said that in beta 7.1 it seeems to have been
solved, IIRC.
--
St