W3C home > Mailing lists > Public > w3c-wai-au@w3.org > April to June 2001

Re: WHat makes Icons accessible or not?

From: Matt May <mcmay@bestkungfu.com>
Date: Wed, 2 May 2001 10:48:04 -0700
Message-ID: <010301c0d330$19e98840$6601a8c0@sttln1.wa.home.com>
To: "Charles McCathieNevile" <charles@w3.org>, <DPawson@rnib.org.uk>
Cc: <w3c-wai-au@w3.org>, <w3c-wai-gl@w3.org>
----- Original Message -----
From: "Charles McCathieNevile" <charles@w3.org>
> CMN Since these are essentially replacing text labels, there is an issue
> they eat too much space. (I think). For people who are reading the text
> packed for reading efficiency, doubling out the spacing can be very
> frustrating. So it is important that the icons can be rendered at some
> small size, I think - 16x16 seems to be good based on the test we have
> (I the example page I produced the other icons at that size in some of the
> examples, just to see how it worked). But I agree that there will be
> who would like to have the icons large. Me for example.

Why the urge to remove the textual labels? Text, properly done, can be at
least as helpful as iconography, especially given the learning curve of
icons that require a key. Not to mention you can't do a find in your browser
for an image, and it won't search alt text. I think I have time in my busy
schedule (hahaha. :) to do a textual treatment of your example text, if
you're interested.

I think, with that said, that set 3 does the job most adequately. Color,
which is extremely helpful when scanning a document like this, is used to
good effect here, insofar as the image still communicates the message
without the color, and going down to 1-bit color displays, the backgrounds
will most likely dither to white. Then, if I'm a multimedia tool provider, I
can rely on color first, if that's available to me, to scan the page for
salient information. And if I can't rely on it, I can fall back on pattern
matching, which is slower, but that's what all of the other icons here are
relying on. (Though it's worth noting to the author the high incidence of
red-green colorblindness among males...)

It's worth noting that set 3 isn't trying to do too much, and it's clear
where the icons start and end. I agree with Emmanuelle's comment that set 2
is an excellent set of illustrative icons, but since they're used
repetitively, I think the goal here is to make the simplest possible icon.
To critique set 1, I should point out that the other two sets define the
boundaries of their icons, where set 1 does not. Doing so keeps the viewer
from having to define where the icon starts and ends, which takes time to
process (defined here as fractions of a second, but still) before they can
process the image itself. I think set 1 may have tried to communicate too
much information in its icons, since these are only needed as indicators of
relevance to a tool set, not as descriptive or explanatory images.

Received on Wednesday, 2 May 2001 13:51:12 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:39:46 UTC