E-mail List Archives

Search E-mail List Archives for

Number of posts in this thread: 6 (In chronological order)

From: John Foliot - WATS.ca
Date: Wed, Mar 09 2005 8:21AM
Subject: Title Attribute (was Re: IMG with a caption?)
No previous message | Next message →

webmaster29 wrote:
> Thanks Joelle! Great article. As for the ALT, it also shows the short
> text for the sighted readers when the pointer is over an image. I know
> it works on IE, but on Firefox ALT does not show anything.
>
> So I add TITLE. For example <img src="someimage.png" ALT="some image
> short description" TITLE="some image short description"> Then the
> description appears both on IE and Firefox. I do not know if
> it is right
> or wrong way to do it (any idea?), but at least the text
> appears on both
> IE and Firefox.
>
> brgds A.
>


Sorry for the delay in responding, but this should not go as is..

There are a few issues here, which should be clarified.

First, using the Title attribute exclusively to produce a "tool tip" is both
a misuse of the Title attribute, and a dangerous way of thinking, as it
leaves me with the impression that it is being done to convey information
that would not be present any other way - and that's even "wronger".

The W3C states:

"This attribute offers advisory information about the element for which it
is set.

Values of the title attribute may be rendered by user agents in a variety of
ways. For instance, visual browsers frequently [but are not obligated to...
- JF] display the title as a "tool tip" (a short message that appears when
the pointing device pauses over an object). Audio user agents may speak the
title information in a similar context. For example, setting the attribute
on a link allows user agents (visual and non-visual) to tell users about the
nature of the linked resource."
(http://www.w3.org/TR/html4/struct/global.html#adef-title)

Meaning that while it can be a good idea to add the Title attribute, we
shouldn't "rely" on it to do "anything", except provide advisory information
to the user agent (leaving the "method" of providing that advise to the user
agent...). Also, relying on "the pointing device" exclusively (see above)
goes against universal accessibility guidelines, which suggest we should be
building device independent content (right?).

Second item is "how" screen readers handle the Title attribute. This is
really dependant on how the end user has configured their screen reader; and
from my experience most daily screen reader users *do* tweak their tool to
their personal preferences - be it speed of delivery (voice), verbosity
settings, grammar settings, etc. For example (depending on which screen
reader being used), in the verbosity settings, users can choose to have
*just* the ALT attribute read, *just* the Title attribute read (when
present), or both/neither. In the grammar settings, they can choose to have
all, some or none of certain grammatical elements read, for example the
square brackets - [] - can be announced or ignored; it depends on the user.

Finally, I am troubled by the line, "...but at least the text appears on
both IE and Firefox." List members should be reminded that there *are* more
than 2 browsers on the market today, and developing web content for specific
browser behaviour is a slippery slope and should be strongly discouraged.

Cheers!

JF
--
John Foliot = EMAIL ADDRESS REMOVED =
Web Accessibility Specialist / Co-founder of WATS.ca
Web Accessibility Testing and Services
http://www.wats.ca 1.866.932.4878 (North America)

From: Webmaster
Date: Fri, Mar 11 2005 10:58AM
Subject: Re: Title Attribute (was Re: IMG with a caption?)
← Previous message | Next message →

JF, thank you for your informative reply.

So I can not use tool tips if I build a BUI (browser user interface)
application. It is the news for me.

I notice now that there are no "tool tips" in Google Gmail, but there
are tool tips in Yahoo webmail interface.

In general my opinion is that the browsers are to support some universal
basic standards. Even TITLE attribute IE and Firefox display
differently. Worse than that the length of the displayed text is different.

In an ideal world the Safari, IE, Firefox, Opera people seat together
once in a while and agree on some basic standard features. It would
prevent appearance of such code just to figure out what browser user has:

if (document.all){
....
}else if (window.sidebar){
....
}else if( window.opera && window.print ) {
....
}

In the real world, however, I have the impression that the producers of
current browsers more think of market cornering than about users and
web-developers.

I do not think that, if there is, for example, the standard way to
display TITLE or a clear function to indicate what browser user has,
say, whatbroswer();, that it would hinder innovation.

We will see how it goes further this summer when IE7 comes out.

brgds A.

foliot wrote:

>>
>
> Sorry for the delay in responding, but this should not go as is..
>
> There are a few issues here, which should be clarified.
>
> First, using the Title attribute exclusively to produce a "tool tip" is both
> a misuse of the Title attribute, and a dangerous way of thinking, as it
> leaves me with the impression that it is being done to convey information
> that would not be present any other way - and that's even "wronger".
>
> The W3C states:
>
> "This attribute offers advisory information about the element for which it
> is set.
>
> Values of the title attribute may be rendered by user agents in a variety of
> ways. For instance, visual browsers frequently [but are not obligated to...
> - JF] display the title as a "tool tip" (a short message that appears when
> the pointing device pauses over an object). Audio user agents may speak the
> title information in a similar context. For example, setting the attribute
> on a link allows user agents (visual and non-visual) to tell users about the
> nature of the linked resource."
> (http://www.w3.org/TR/html4/struct/global.html#adef-title)
>
> Meaning that while it can be a good idea to add the Title attribute, we
> shouldn't "rely" on it to do "anything", except provide advisory information
> to the user agent (leaving the "method" of providing that advise to the user
> agent...). Also, relying on "the pointing device" exclusively (see above)
> goes against universal accessibility guidelines, which suggest we should be
> building device independent content (right?).
>
> Second item is "how" screen readers handle the Title attribute. This is
> really dependant on how the end user has configured their screen reader; and
> from my experience most daily screen reader users *do* tweak their tool to
> their personal preferences - be it speed of delivery (voice), verbosity
> settings, grammar settings, etc. For example (depending on which screen
> reader being used), in the verbosity settings, users can choose to have
> *just* the ALT attribute read, *just* the Title attribute read (when
> present), or both/neither. In the grammar settings, they can choose to have
> all, some or none of certain grammatical elements read, for example the
> square brackets - [] - can be announced or ignored; it depends on the user.
>
> Finally, I am troubled by the line, "...but at least the text appears on
> both IE and Firefox." List members should be reminded that there *are* more
> than 2 browsers on the market today, and developing web content for specific
> browser behaviour is a slippery slope and should be strongly discouraged.
>
> Cheers!
>
> JF
> --
> John Foliot = EMAIL ADDRESS REMOVED =
> Web Accessibility Specialist / Co-founder of WATS.ca
> Web Accessibility Testing and Services
> http://www.wats.ca 1.866.932.4878 (North America)


> webmaster29 wrote:
>
>>Thanks Joelle! Great article. As for the ALT, it also shows the short
>>text for the sighted readers when the pointer is over an image. I know
>>it works on IE, but on Firefox ALT does not show anything.
>>
>>So I add TITLE. For example <img src="someimage.png" ALT="some image
>>short description" TITLE="some image short description"> Then the
>>description appears both on IE and Firefox. I do not know if
>>it is right
>>or wrong way to do it (any idea?), but at least the text
>>appears on both
>>IE and Firefox.
>>
>>brgds A.
>
>
>

From: John Foliot - WATS.ca
Date: Fri, Mar 11 2005 11:22AM
Subject: Re: Title Attribute (was Re: IMG with a caption?)
← Previous message | Next message →

webmaster29 wrote:

> So I can not use tool tips if I build a BUI (browser user interface)
> application. It is the news for me.

That is not *exactly* what I said, but close enough. The real issue is not
to *DEPEND* on the tool tip function to provide information that is not
available in another fashion. Subtle, but important difference.

> In general my opinion is that the browsers are to support
> some universal
> basic standards. Even TITLE attribute IE and Firefox display
> differently. Worse than that the length of the displayed text
> is different.

Right, but that assumes that all browsers (user agents) are installed and
running on a standard tower or laptop configuration. But what of other
devices? Or Adaptive Technology browsers? I constantly go on about the
fact that universal accessibility is so much more than just blind users, but
what of them? If they can't "see" the tool tip, they had better be able to
still get to the important information, no?

>
> In an ideal world the Safari, IE, Firefox, Opera people seat together
> once in a while and agree on some basic standard features.

Well, the W3C has set the standards, but left the implementation to the
developers to best deliver the intent to their focus users. Title attribute
is an important and useful attribute, but *mandating* that it do one thing
over another is wrong... Mandating IBM HomePageReader to show a tool tip is
just plain foolish (although it can, and I believe does, as they use IE's
rendering engine). However, since most blind users I know don't use a mouse
anyway , it's kind of a mute point don't you think?

> It would
> prevent appearance of such code just to figure out what
> browser user has:
>
> if (document.all){
> ...
> }else if (window.sidebar){
> ...
> }else if( window.opera && window.print ) {
> ...
> }

But why are you doing such hacks in the first place. Why do we, as
developers, need to hijack the client's user agent? Let the client
configure their agent as they require... Stop trying to tell the end user
you know better than them how to set up and run their machines.

>
> I do not think that, if there is, for example, the standard way to
> display TITLE or a clear function to indicate what browser user has,
> say, whatbroswer();, that it would hinder innovation.

Well... A) you have not yet made a clear case as to why the title attribute
"must" display a tool tip. B) there are other cross browser solutions which
could deliver this type of "visual functionality" that could still address
accessibility... Roughly speaking here I'm thinking of a with display
properties set via CSS that could create a show and hide scenario based upon
the :hover selector... (but don't have the time to hack it out here).

>
> We will see how it goes further this summer when IE7 comes out.

Relying on a specific user agent to deliver functionality that is not
available to other user agents is contrary to the accessible design
principle. (Besides, I hope IE expends more time/effort on security than it
does on "pretty", or "whiz-bang" tricks).

JF

--
John Foliot = EMAIL ADDRESS REMOVED =
Web Accessibility Specialist / Co-founder of WATS.ca
Web Accessibility Testing and Services
http://www.wats.ca 1.866.932.4878 (North America)

From: Mary Martinson Grossnickle
Date: Fri, Mar 11 2005 5:31PM
Subject: Re: Title Attribute (was Re: IMG with a caption?)
← Previous message | Next message →

JF wrote:
> > First, using the Title attribute exclusively to produce a "tool
> tip" is both
> > a misuse of the Title attribute, and a dangerous way of thinking, as it
> > leaves me with the impression that it is being done to convey
> information
> > that would not be present any other way - and that's even "wronger".

Thank you for making this point, which is something I've been trying to
convey to a client who is using exclusively because they claim that
Netscape does not read alt text. Could it be that Netscape _does_ read alt
text, but the screen reader they're using for testing is set to only read
?

I do understand your point below:
> List members should be reminded that there *are* more
> than 2 browsers on the market today, and developing web content for
specific
> browser behaviour is a slippery slope and should be strongly discouraged.

However, since we encourage testing with several browsers, we need to be
aware of how different browsers act. So I'm wondering about whether Netscape
reads alt text.
Thanks,
Mary

From: Patrick H. Lauke
Date: Fri, Mar 11 2005 5:42PM
Subject: Re: Title Attribute (was Re: IMG with a caption?)
← Previous message | Next message →

maryg wrote:
> Thank you for making this point, which is something I've been trying to
> convey to a client who is using exclusively because they claim that
> Netscape does not read alt text. Could it be that Netscape _does_ read alt
> text, but the screen reader they're using for testing is set to only read
> ?

At the risk of making a far too generalised statement, I was under the
impression that the majority of current screenreaders on Windows are
only tuned to integrate with IE. Any other combination would probably
yield unreliable results.

This is eerily similar to a recent discussion over at Accessifyforum
http://www.accessifyforum.com/viewtopic.php?t=2592

--
Patrick H. Lauke
_____________________________________________________
re

From: John Foliot - WATS.ca
Date: Sat, Mar 12 2005 10:43AM
Subject: Re: Title Attribute (was Re: IMG with a caption?)
← Previous message | No next message

maryg wrote:
>
> Thank you for making this point, which is something I've been trying
> to convey to a client who is using exclusively because
> they claim that
> Netscape does not read alt text. Could it be that Netscape _does_
> read alt text, but the screen reader they're using for testing is set
> to only read
> ?

Clarifications:
A) there are two "titles",
* the title element ()
* and the title attribute (<img, src="file.gif" alt="alt text here"
title="title information here" />)

B) Netscape (?) does not "read" anything... It renders HTML content into a
GUI environment.

Now the question is: does the Netscape browser (and you need to specify
which version) render a tool tip using only the ALT attribute? In current
builds (based upon the Gecko rendering agent), no. However, the title
attribute will *currently* produce this result in the Gecko based browsers
(Netscape 7, Firefox, Safari, etc.). This may change, at any time, in any
build. (and/or maybe just for one OS... Netscape/Gecko is available for
Windows, Mac, Linux...)

C) Screen readers come in many flavors, and can be configured to suit the
end users preferences. Some screen readers interact with other applications
(JAWS, WindowEyes), others are stand-alone applications which read HTML
documents (IBM HomePageReader). "Expecting" a specific behaviour from any
of these tools is a pipe-dream - see comments above regarding *one* browser.

>
> However, since we encourage testing with several browsers, we
> need to be
> aware of how different browsers act. So I'm wondering about
> whether Netscape
> reads alt text.

Why? If a browser is set to not display images (voluntarily or by default),
then the W3C says that the ALT text should be rendered. Walk away from
"Tool Tips" as a method of providing any supplementary information that is
critical... That behavior is not *Mandated* anywhere, it is at the sole
discretion of the browser/user agent developer. (And thankfully, this will
never change)

Yes, Netscape will provide the ALT text if you have configured it to not
display images; no, it will not produce a "Tool Tip" via the ALT attribute.
So what? Who cares, and more importantly, why do they care?

I will repeat it again... Tool Tips as a means of conveying critical
information is a misuse of the Title Attribute, and a dangerous methodology
- PERIOD.

What about the ABC browser? The FooBar Browser? And so on and so on. What
of non-visual users? (when was the last time you saw a blind user with a
computer mouse?) What of text only users? (Lynx does not produce "Tool
Tips" ever).

Yes, you want to test in multiple browsers to ensure that the visual
rendering is approximately the same, but please always remember that it is
humanly impossible to set up a testing lab with every conceivable user agent
configuration possible (well, unless you have Bill Gate's money, and even
then, there is a practical limitation to how many machines you can feed and
manage...). Use your HTML the way the W3C tells you to, and leave the
"interpretation" to the end users.

JF --
John Foliot = EMAIL ADDRESS REMOVED =
Web Accessibility Specialist / Co-founder of WATS.ca
Web Accessibility Testing and Services
http://www.wats.ca 1.866.932.4878 (North America)