Re: FW: CHANGE PROPOSAL: Table Summary

Laura Carlson, Thu, 10 Dec 2009 06:42:32 -0600:
> Hi Josh
> 
> You wrote:
> 
>> Is this kind of 'sorting' mechanism necessary at all? And if so should
>> it be separate disambiguated from @summary. If a sorting mechanism is
>> needed I think this would be best. A sorting mechanism could be useful
>> for dynamically generated tables I guess.
> 
> Excelent question, Josh.
> 
> If in fact a sorting mechanism is needed, a separate change proposal
> for it would  seem to be in order.

It is unclear to me what you and Josh are saying. Doesn't @orientation 
also need a separate change proposal then? And aria-sort is part of 
ARIA, so it would take a change proposal to *not* get aria-sort. Ian 
said that information about sort order should be linked to something 
that actually changes the sort order. I agree that something like that 
probably needs a separate change proposal. However, no one has actually 
proposed such a feature. The definesorder="" attribute that I suggested 
is also only about information. (However, it gives slightly less 
information than aria-sort, and thus would be less likely to get out of 
touch with the data of the table.)

A compelling usecase for @summary was that it could be used to inform 
specifically about the sort order or sort column. Obviously, such info 
could be relevant to get in your preferred localization. And since 
@summary will inherit the language of the <table> element, it seems 
wider accessible to use e.g. @definesorder for that, so that the user 
agent can inform you about it in your own language.

@summary could still have a role to play w.r.t. to giving information 
about the sorting and/or the defining column/row. For example, even 
aria-sort is very unspecific about the kind of sorting.

May be a good thing to have in the spec about @summary is that authors 
should try to create it in a "programmatic" way. Don't be "creative", 
but use it to present short facts about the table, derived from the 
table itself.
-- 
leif halvard silli

Received on Thursday, 10 December 2009 13:23:10 UTC