W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > April to June 2002

RE: Accessibility Valet

From: Jukka Korpela <jukka.korpela@tieke.fi>
Date: Mon, 24 Jun 2002 09:40:09 +0300
Message-ID: <621574AE86FAD3118D1D0000E22138A95BDD48@TIEKE1>
To: w3c-wai-ig@w3.org, w3c-wai-er-ig@w3.org

Nick Kew wrote:

> - - should I still do more to
> enhance the accessibility of the reports themselves?

Yes.

But, Nick, is it really productive to have discussion on four fora, even if
they are coupled two by two? This is now on two lists, and you posted the
question to two Usenet groups as well. I would strongly recommend having _a_
discussion on one forum - you could of course invite people there on other
fora. People who read these lists but not the groups, or vice versa, will be
unaware of the contributions elsewhere, perhaps repeating the same points,
perhaps missing relevant information, perhaps failing to make a valuable
comment on something he missed.

> Jukka - I still have in mind some of your comments regarding Page
> Valet.  The new tool offers a "Listed" option based on what I think
> you were suggesting: is this something like you had in mind?

It's in that direction. It avoids the problem with the old format (which is
still the default report format) of repeating information. Once you've told
me that, say, table axes should be identified, I don't gain much from seeing
that for _every_ <th> element. (Except in special cases.)

I still think Site Valet outputs too much information, so that essential
information is missed. I think results from an accessibility check should
_begin_ with a simple, accessible statement that tells the overall result
(like failing or passing a test, together with a number of errors detected,
and with the URL of page echoed so that, for example, if I print the results
I know what the tested page was).

Displaying the "Normalized Markup" (how accessible is that term?) can be
nice at times, but done routinely, it produces far too much data.

Even the explanations are overinformation. They should be on a separate
page, so that one does not need to see (or hear) them every time. And on
that separate page, there should be an explanation of what "confidence"
really means. Your confidence on being right when reporting that the tested
page might have errors? For example, the message "Ensure that documents are
readable without stylesheets too." has confidence Low. But the sad reality
is that all too often people create pages that _rely_ on CSS effects for
essential matters. How will those people be affected by the adjective "Low"?

I have difficulties in understanding some of the messages, but maybe I need
to study the WAI guidelines in more detail. Is it really the recommendation
that we specify axes for all tables and explicitly associate data cells with
headers, even for simple tables?

I don't quite get the point of giving the message "Ensure that documents are
readable without stylesheets too." about a <style> element (apparently,
about _any_ style element) but not about <link rel="stylesheet">. Surely
using an external style sheet does not remove the potential problem!

"Summary page assessment" was obscure to me - I thought it tried to tell me
something about the summary of my page! I then realized that it's the
summary of the evaluation.

And I don't really like a checker nagging about each and every blockquote,
suggesting that I'm using it for formatting, and asking about each and every
heading (sic) "Is this really a header?". If you start listing _possible_
problems, there's no end. Or at least a checker should suppress such
messages unless they are specifically requested for.

-- 
Jukka Korpela, senior adviser
TIEKE Finnish Information Society Development Centre
http://www.tieke.fi
Phone: +358 9 4763 0397 Fax: +358 9 4763 0399 
Received on Monday, 24 June 2002 02:40:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:14:05 GMT