RE: the text in images issue

Kynn (et al.),

I am quite relieved that I am not in out-and-out disagreement with you here!

Your list (a) - (e) is very good, but I do want to qualify a few things for
anyone following this thread.

(b)  Good ALT text is quite sufficient (i.e., better than "generally
considered") to remove the barrier to access for people who are blind (and
using a screen reader with synthesized speech or dynamic Braille display).

(c)  Good ALT text lessens the barrier to access for many people with low
vision, but they will still find it difficult to access information in the
document.

Somewhere between (c) and (d) is the user interface.  People with low vision
have the day-today experience and extremely reasonable expectation and that
textual elements are scaleable.  They are use to one or more of the
following techniques: (1) getting a bigger monitor, (2) having the OS
default to a high contrast / large font size scheme, (3) setting an
acceptable default font (and colors) within their browser, (4) choosing the
"larger" setting from within their browser, and (5) using a customized style
sheet to format.  Graphical text, of course, defeats ALL of these common
accommodation strategies.

(d)  There are other work-around (for the bad pages that include graphical
text) in addition to disabling images.  Mouse-over pop-ups is sufficient for
some folks with low vision for example.  (Others will not benefit from this
technique.)  Screen magnification (even of graphical text) is adequate for
_some_ for most of the pages they encounter.  Alternative browsers are used
by some as an accommodation.  I am not aware of any studies, but I would
guess that the technique of disabling images is, for most people with low
vision, a "cure" that is not worth the cost.  (Just as the "cure" of using
screen magnification is often declined.)

Somewhere between (d) and (e) the point needs to be made that the existence
of these work-arounds and alternatives do NOT totally eliminate the barrier
caused by textual graphics.  That's why this is a P2 issue and not P1 or P3.
The use of the terms "satisfactory" or "sufficient" can be quite subjective.

(e)  I don't agree that your final point is the "clear solution", although
it might be a fair long term approach.  I would suggest that a more
reasonable conclusion is that content providers should use CSS when they
want to put text on a graphic!  SVG is likely to be well supported (and
therefore a good choice) long before we have some other (and completely new)
way to differentiate text-heavy images from other graphics.  I agree that
_right now_ SVG is not much of a solution, but that will change (hopefully,
rather quickly).  [I was quite disappointed to find that an SVG viewer was
NOT incorporated into Navigator 6.]

With regard to your other message on this immediate sub-thread, I agree that
permutations of:
	<a href="page02.html"><img src="rarrow.gif" alt="next"/></a>
	<br/><a href="page02.html">next</a>
are pretty terrible because of the redundant rendering of "next".

Len's example endorses your CSS / ALT technique suggestion:
	<a href="page02.html"><strong class="buttontext">Next</strong><br>
	<img class="buttonimg" src="twirl.gif" alt="" width="75" height="50"
border="0"></a>

I would guess that <Q>border="0"</Q> is there to accommodate the
non-conformant behavior of browsers.  (This could be another argument for
transitional doctype on Double-A pages.)

Anne (and perhaps others) won't like it, but I still think that:
	<a href="page02.html" title="Page two of three."><img
src="rarrow.gif" alt="Next!"></a>
is the way to go (unless one is specially setting up a web page to simulate
an aug. com. board).

If an icon image is so weak (or subject to interpretation) that it requires
text to illustrate its function -- then either drop the image (and just use
the text) or improve the graphic!  Unless, of course (and as the case with
Anne) the purpose of the page is to teach literacy skills.

Cheers,
Bruce

> -----Original Message-----
> From: Kynn Bartlett [mailto:kynn-edapta@idyllmtn.com]
> Sent: Friday, December 15, 2000 12:54 PM
> To: Bailey, Bruce; 'w3c-wai-gl@w3.org'
> Subject: RE: the text in images issue [was: errata...]
> 
> 
> I think our main problem is the following:
> 
> (a) Textual graphics are a barrier to people who require the ability
>      to magnify text.
> 
> (b) Alt text is generally considered sufficient to remove the barrier
>      to access for people who can't see an image.
> 
> (c) Using the alt text for those images -would- be sufficient to
>      allow access, BUT:
> 
> (d) To access the alt text, the user needs to disable ALL images on
>      a page, which is less than optimal for web users who need to
>      magnify text but who can otherwise see graphics.
> 
> (e) The clear solution is that there needs to be a way to designate
>      certain images as being those that should be -replaced- if
>      resized, while others as not being replaced.
> 
> There does not seem to be a way to do this in HTML or CSS at present.
> This whole argument is basically caused by a huge gaping hole in
> the markup languages used to create web pages.
> 
> So, if we were going to reinvent HTML, how would we solve the problem?
> Perhaps an attribute related to resizing, which would then be
> recognized by browsers?  Maybe a CSS property?  Something else?
> I'm not sure what the solution is, but I think there may be a way
> to overcome this problem.  (Using SVG is not, right now, a 
> solution...)

Received on Friday, 15 December 2000 15:55:03 UTC