W3C home > Mailing lists > Public > public-html@w3.org > November 2008

Re: PFWG report on @headers status

From: Gez Lemon <gez.lemon@gmail.com>
Date: Tue, 25 Nov 2008 17:58:33 +0000
Message-ID: <e2a28a920811250958i6399bc10m6588de5b0d23bb9@mail.gmail.com>
To: "Robert J Burns" <rob@robburns.com>
Cc: "Laura Carlson" <laura.lee.carlson@gmail.com>, joshue.oconnor@cfit.ie, "Al Gilman" <Alfred.S.Gilman@ieee.org>, "Janina Sajka" <janina@a11y.org>, "Michael Cooper" <cooper@w3.org>, "HTML WG" <public-html@w3.org>

2008/11/25 Robert J Burns <rob@robburns.com>:
> One thing I've recommended before is that we require the headers attribute
> to only reference cells before themselves (the referenced cell has to make
> sense as a header cell for the referencing cell). In terms of document
> conformance this could be confined even tighter. In terms of processing
> conformance, UAs must only process headers attributes that reference prior
> cells in tree order. All other headers attribute values must be ignored
> (i.e., those referencing cells that follow in tree order). We could also add
> that headers cannot reference cells in the tfoot element for completeness.
> With these rules, there is no possibility of a circular referencing of
> header cells. The author may accidentally include such circular references,
> but by ignoring references that follow the referencing cell, the
> error-handling is determined completely.

I think we need to consider the purpose of headers before solving
problems that may not exist. The purpose of the attribute is to
specify the headers for a particular cell so that they can be queried
by a user or announced to provide context when the user navigates
along a particular axis. When the user queries the cell, only the
headers pointed to by the headers attribute should be announced. The
fact that the headers themselves may have headers is irrelevant; they
would only be required if the user navigates to the header cell, and
request headers for that cell. If that wasn't the case, any chained
header would result in irrelevant information being announced to the

I think it's sufficient to say that cyclic references for headers is
illegal, without defining rules to prevent something happening that
couldn't happen.

I think the confusion arises from the phrase, "This seems to some
degree to be unfounded". Concerns about cyclic references causing
problems in implementations are either founded or unfounded. I can't
think how cyclic headers could cause a problem with implementations,
but maybe someone has?



Supplement your vitamins
Received on Tuesday, 25 November 2008 17:59:10 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:39 UTC