W3C home > Mailing lists > Public > w3c-wai-er-ig@w3.org > March 2004

Re: Chris's test cases - first checkpoint

From: Chris Ridpath <chris.ridpath@utoronto.ca>
Date: Thu, 25 Mar 2004 16:57:48 -0500
Message-ID: <321901c412b4$36543cc0$b040968e@WILDDOG>
To: "Lisa Seeman" <lisa@ubaccess.com>
Cc: <w3c-wai-er-ig@w3.org>

Lisa,

Thanks for your comments and questions. My replies...

> Re: All IMG elements must have an ALT attribute.
> ID Number 1
>
[1]
I can't recall any exceptions to this rule. ALT can be null or spaces but it
has to be there.

> Re IMG element cannot have ALT attribute value of null ("") if WIDTH
> attribute value and HEIGHT attribute value are both greater than 25.
> ID Number 4
> In most case - but always?
>
[2]
Not always. This is a potential problem. It's perfectly OK to have large
images with null ALT but in some cases it's wrong. This may be a useful
check for programs, or people, that insert empty placeholder ALT text.

There's an important distinction between these 2 checks. The first check is
for a known problem. The second check is a potential problem and requires
human intervention to figure out it there's really something amiss.

Ideally, all the checks would work perfectly and there would be no potential
problems. Currently, more than half of the checks are for potential
problems. Human intervention is still required but at least you're pointed
in the right direction.

> Maybe we should have more maybes?
>
The checks are better if they can detect with certainty the problem. If we
add more "maybes" then it means the checks will get less use. People will
just ignore them.

> This description I really did not get :
> IMG element must contain a LONGDESC attribute.
> ID Number 8
>
[3]
This is a check for a potential problem. Likely the image does not require a
LONGDESC but you have to check for it.

I'm thinking that this check would be better if there was a size
restriction. If we keep asking people if their spacer gifs and bullets
require a longdesc then the check will be ignored.

> ID 34:
> If FRAMESET element contains 3 or more FRAME elements then FRAMESET
> element must contain a LONGDESC attribute that is a valid URI.
> -isn't that only if the relationship is complex - otherwise a title on
> each frame will do
>
[4]
How do we determine if the relationship between frames is complex? I've
guessed that this occurs when there are 3 or more frames. Can anyone provide
a clearer definition of frame complexity? Or, when does a frameset require a
longdesc?

Chris

[1]
http://tile-cridpath.atrc.utoronto.ca/acheck/servlet/ShowCheck?check=1&lang=eng
[2]
http://tile-cridpath.atrc.utoronto.ca/acheck/servlet/ShowCheck?check=4&lang=eng
[3]
http://tile-cridpath.atrc.utoronto.ca/acheck/servlet/ShowCheck?check=8&lang=eng
[4]
http://tile-cridpath.atrc.utoronto.ca/acheck/servlet/ShowCheck?check=34&lang=eng


----- Original Message ----- 
From: "Lisa Seeman" <lisa@ubaccess.com>
To: "'Chris Ridpath'" <chris.ridpath@utoronto.ca>
Cc: <w3c-wai-er-ig@w3.org>
Sent: Thursday, March 25, 2004 1:27 PM
Subject: Chris's test cases - first checkpoint


> Good work Chris
>
> Initial dilemmas.
>
> Re: All IMG elements must have an ALT attribute.
> In WCAG 1 I seem to remember an exception to this rule - were the alt
> tag had text that was all repeated in the text.
> Wendy - do you remember this one?
>
> Re IMG element cannot have ALT attribute value of null ("") if WIDTH
> attribute value and HEIGHT attribute value are both greater than 25.
> ID Number 4
> In most case - but always?
>
> I think that is a general comment of mine, that I am not sure that there
> are no exceptions to the rules
> Maybe we should have more maybes?
> Same comment for assuming that a large picture is important (it could
> still be background and have a div with absolute positioning super
> imposed on top. (ID Number 5 )
>
> This description I really did not get :
> IMG element must contain a LONGDESC attribute.
> ID Number 8
>
> Same idea for id:8
> Any text in image is also in ALT text. - as a pass criteria
> Often, if there is a lot of text, and it is important it belongs in the
> long desc
> If there is a lot of text but it is repeated in the true text we do not
> need it at all. If there is a lot of text but it is vastly irrelevant, I
> would say forget it (for example a site on dyslexia may have the word
> "dyslexia" misspelled in list of different ways)
>
> ID 34:
> If FRAMESET element contains 3 or more FRAME elements then FRAMESET
> element must contain a LONGDESC attribute that is a valid URI.
> -isn't that only if the relationship is complex - otherwise a title on
> each frame will do
>
> ID 90:
> Does a script that just changes the shading need a no scripts?
> Etc
>
> Objects- is there such a thing as an object that is just visual - with
> out semantic content? If so then will a blank title do?
> Let's take an animation/object of an dot getting bigger and smaller.
> Does the text equivalent need to be larger then the words "dot getting
> big and small". do we need it at all? What if a description of the
> object follows in the text. Does an object described above need both a
> title and embedded text?
>
>
> Applet - Will embedded text do for an applet or do they _need_ an alt as
> well - And the same kind of issues as with object
>
>
> This is a great exercise in defining what we think.
>
> All the best
> Lisa Seeman
>
> Visit us at the UB Access website
> UB Access - Moving internet accessibility
>
>
>
Received on Thursday, 25 March 2004 16:59:13 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Thursday, 9 June 2005 12:10:42 GMT