Re: Summary of Thursday's IRC conversation about @summary

At 17:27  +0100 5/06/09, Philip Taylor wrote:
>As it happens, I'm using that data now. 
>http://philip.html5.org/data/table-summary-values-dotbot.html shows 
>the <table summary> values from their set of pages. 
>(http://philip.html5.org/data/table-summary-values.html has the 
>older data from dmoz.org.)

wow.  fascinating.  depressing.

aside: anyone happen to know what software makes summaries like 
"pid<number>" (there are many many lines like pid2290201)?

I was going to ask how many summaries were used only once (on the 
grounds that a unique summary was plausibly custom-made for that 
page) until I saw all these!

(anyone want to hazard an estimate how many of these would be useful 
for someone needing a summary?)

At 12:22  -0700 5/06/09, John Foliot wrote:
>It is also important to note that the PF-WG specifically wrote:
>
>*	We reject the argument that summary should be removed from the
>HTML
>*	specification because it is not implemented on most web sites. We
>note
>*	that accessibility is poorly supported on most web sites. The
>wider
>*	web is not an example of good practice.

Hm.  I think we've already noted that "failed to establish a cowpath" 
is not a design principle (though I think we can all take it as a 
cause for concern and ask "why?", and if the problem is the 
specification, see whether we can do better.)

I think the concern here is that it has become hopelessly polluted - 
that nobody who needs what a proper summary would provide, would ever 
dream of looking in the summary attribute. (And no-one who writes UAs 
to do that would do it, either).

Which is why I am happy to see facts, even if they strike as 
depressing on first glance.
-- 
David Singer
Multimedia Standards, Apple Inc.