W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > July to September 2004

Re: Layout versus data tables proposal for null summary attribute

From: Jesper Tverskov <jesper.tverskov@mail.tele.dk>
Date: Mon, 30 Aug 2004 09:52:07 +0200
To: <w3c-wai-ig@w3.org>, "'Michael Cooper'" <michaelc@watchfire.com>, "'Wendy Chisholm'" <wendy@w3.org>
Message-ID: <000001c48e66$40aa5690$440bc650@tversdatg7y7vv>

At the moment I use Visual Studio C#.Net for all round coding and often
I use HTML-Kit for more specific HTML coding. I then copy and paste the
code made in HTML-Kit into Visual Studio Net.

Visual Studio .Net by default only sets in attributes one by one, and
there is no automatic attribute="". Good coders will set up macros that
put in all the basic attributes needed including alt="" and summary="".
The null value of the mentioned attributes not yet meaning anything. The
same is true for most other developers' tools. Exactly how coders do it,
and how often alt="", etc., is actually used in an automatic way, I have
no statistics.

HTML-Kit sets in alt="" and summary="" automatically, and it is my
impression that very many high class developers' tools do the same. And
developers' tools ought to do it, because it is really a nice help.

Especially, developers' tools should always put alt="" and summary=""
into images and tables to remind the developers of using them, or it
should be easy to make macros doing the same thing.

If alt="" didn't mean "decoration", etc., and summary="" didn't mean
layout table, it would be easy for testing tools to tell web page
authors that they have forgotten to fill in those attributes having an
empty or null value.

*** Attribute="" is the most obvious choice for suggesting to use an
attribute, or for an attribute where the exact value is still missing.

This is usability for web page authors that could benefit usability and
accessibility for users. We are making such testing less easy to
implement by letting attribute="" mean anything else than a value not
yet in place or an attribute not yet deleted.
I honestly believe that conventions like alt="" is a result of
accessibility too often being too far removed from making web pages at
street level.

I am not proposing that there should be no way of telling that an image
is decoration or a table is for layout. But we should choose methods
where testing most likely is to be relied upon. We should choose a
method where we can be almost sure that the web page author has made a
deliberate choice and has not just forgotten to fill in a value.

Best regards,
Jesper Tverskov

Received on Monday, 30 August 2004 07:52:04 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:29 UTC