Re: Bug report and ideas

>   Show the warning during validation, NOT during editing. Or show it in
the
>   status bar, so the user can continue his/her work without interruption

Why, so one has to go back and fix it later?  The point is that the alt text
should be added during authoring, not later.

>   Anyway that message is plain stupid. I used to use gif images to create
>   round corners for tables and other merely cosmetic artifacts, and to
enforce
>   table layout as well; why the heck should a 3x3 pixel round corner have
>   an alt text? is it supposed to?

I'd put a space, or the word "space," in the alt text in this situation (a
graphic that is only present to provide graphic structure or styling that
would be of no interest to visually impaired users or to users of Lynx or
AvantGo with the graphics shut off); and for bullets, etc. I use a period.

<img src="round_corner.gif" alt="space" /> (or =" ")
<img src="rollover_bullet.gif" alt="." />

It's a useful compromise between the intentions of the WAI and the
commercial requirements of current website design trends: try to make
everything easily readable in Lynx and AvantGo (and therefore voice
browsers), and beautiful in Mozilla, without screwing it up for IE 4/5,
Netscape 3/4, or Opera 5.

>   Be practical: people don't use <table>'s to show tabular data, people
don't
>   use gif images to express deep concepts or add content to a page

They shouldn't use GIFs to add any kind of deep content, no; gifs should be
used to display graphical content (especially drawings), and jpegs for
photographs. Anything that requires really complex explanations (like a flow
chart) should be presented as a separate link anyway.  When it comes to
graphic styling, logos and buttons have obvious alt texts (<img
src="mycompanyslogo.png" alt="My Company, Inc." />; <img
src="homepagebutton.png" alt="Home Page" />) and those should be supplied
for visually impaired users.

> Yes, people do. People also use them for stylistic effects (why? If you
only
> care about a handful of users with recent browsers they have CSS support.
If
> you care about the handful that don't then it is likely the users aren't
> interested enough in layout support to either upgrade or justify the work
you
> do for them).

Support for CSS-based layout is very poor in Netscape 4; it is much better
in IE 5.5 and Mozilla.  Oddly enough, most of the folks I deal with use
Netscape 4 as their browser, unless they have AOL (groan).  It's also hard
sometimes to distinguish between a "semantic" usage for a table and a mere
"presentational" usage, as tables by their very nature are presentational
beasts: take the following as an example:

<table summary="Layout table: The first cell contains a navigation bar of
W3C technologies, the second contains news, and the third another
navigation bar of W3C pages." border="0" width="99%" cellspacing="0"
cellpadding="10">

Each of those two navigation bars could be done as floating boxes with a
proper implementation of CSS.  And someday I'm sure the W3C will do it -
when browsers with proper support of the relevant pieces of CSS2 are widely
used.  Until then, folks will continue to use the <table> element for a lot
of the presentation they wish to do.

>   > I'd even suggest that, with a check box, it should still
>   > generate a warning, explaining why alt text was needed, and
>   > defaulted to not accepting the change.
> I would like to see an option for null alt text as a checkbox on the
dialog
> in Amaya.

Amaya is a testbed for W3C standards and technologies: basically a way to
prove that "it can be done."  One thing worth proving is that adding alt
text for every image won't be a hardship. Why should W3C make it possible
for authors using Amaya as an authoring tool to violate that standard?

Just my 2c.

Patrick Rourke (no relevant affiliation)
ptrourke@mediaone.net

Received on Tuesday, 20 March 2001 08:40:31 UTC