W3C home > Mailing lists > Public > www-archive@w3.org > January 2004

Re: Use of Axis in data-table for categorizing cells

From: Al Gilman <Alfred.S.Gilman@IEEE.org>
Date: Sat, 10 Jan 2004 12:43:46 -0500
Message-Id: <>
To: "Sailesh Panchang" <sailesh.panchang@deque.com>
Cc: Lisa Seeman <seeman@netvision.net.il>, www-archive@w3.org

[distribution note:  I copied www-archive so that you will have a publicly
accessible record of this message to cite if you wish to.  But I haven't
copied any distribution list so as not to overfill inboxes.  -Al]

At 05:25 PM 2004-01-09, you wrote:
>Well help me understand this correctly:
> >Whatever is enclosed
> >>within quotes following axis= is interpreted as a single concept  even if
> >>two words are separated by a comma.
>Yes I certainly mis represented what I meant.  The value of an 
>axis  attribute is a single set of one or more concepts with a comma 
>separator. And that set in its entirety  is associated with the data cell 
>that references it using headers="the id value". Right?


It is only in processing queries to filter the contents that the
separate items in the list are treated separately.

>So for instance a data cellwill have both "toxic waste" and "into 
>watershed" associated with it if headers="a2" where"
><th id="a2", axis="toxic waste, into watershed">Waste's properties</th>
>There is no way for one data cell to have only "toxic waste" associated to 
>it and another data cell to have  "into watershed" associated with it if 
>they both refer to the above cell a2. Right?

Right.  There is no way to have some cells with one property and other cells
with another via headers=id association and the consequent heredity from the
'axis' value for a header cell.

Now I suspect that you have an application where you would like to have
an irregular pattern of category-membership-properties in the superficially
regular array of data cells.  Is this true?  Are you looking for an HTML
coding technique where one cell can have the "toxic waste" and another
have the "into watershed" property in patterns more irregular than

  row X column ==> cell properties


In cases like this, the print practice is usually to use some styling
effect, a variation in background color, cell boundary, or a small icon,
to signal the changes from cell to cell.  And provide a key somewhere on
the page.

class="toxicWaste intoAtmosphere" -- on one cell
class="toxicWaste intoWatershed" -- on another

Note that the HTML syntax lets you do this with 'axis' attributes on the
TD cells.  You don't have to inherit an 'axis' value through a header.
However, unless you have the community of browser and AT tools committed
to processing the 'axis' attribute in special ways already, it makes more
sense to use 'class' which style writers will understand as grist for
their selectors.  Not that selectors couldn't select on values of 'axis,'
mind you, but there is no practical (meaningfully enforced) difference
between a category invoked with 'axis' as opposed to a category invoked by

Unless IE fails to cope with multiple tokens in the 'class' attribute
(is there a bug like that?) I would suggest this is the superior approach.

One way to code this 'accessibly' is with class tokens on the data cells and
a machine interpretable key somewhere in the page or associated with the page
in a way that the assistive technology will recognize as associated with the
page in a way it cares about.

If you wish to pursue a solution to that problem I suggest that you
get in touch with Lisa Seeman and work on the RDF techniques.  RDF is
one way to provide the machine-interpretable key for cell properties
that are marked with 'class' tokens.

Compare with the request for something that works like labeledBy from the
Java Access API.


>This does not come out clearly  from reading the HTML specs  and an 
>example with an axis-value containing a comma separator    would have 
>brought out the  point more clearly.

True, and I would want WAI to furnish or cite such an example if I thought
that AXIS were worth fighting to save.

>Or was I the only one  who was confused?

No way.

Our experience is that for each one list member with the courage to ask
a "dumb question" there are ten who are lurking out there who don't
know the answer.  The only dumb question is the one you were too afraid
to ask.  Keep up the good work.


>Thanks for the clarifications.
>Sailesh Panchang
>Senior Accessibility Engineer
>Deque Systems,11180  Sunrise Valley Drive,
>4th Floor, Reston VA 20191
>Tel: 703-225-0380 Extension 105
>E-mail: <mailto:sailesh.panchang@deque.com>sailesh.panchang@deque.com
>Fax: 703-225-0387
>* Look up <<http://www.deque.com>http://www.deque.com> *
Received on Saturday, 10 January 2004 16:33:59 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:42:36 UTC