Re: summary attribute compromise proposal

On Thu, Aug 6, 2009 at 12:07 PM, Julian Reschke<julian.reschke@gmx.de> wrote:
> Maciej Stachowiak wrote:
>>
>> ...
>> I think you have to read it in context. Notice how those two sentences are
>> followed by a long paragraph explaining the right considerations for when
>> summary is appropriate.
>>
>> Keep in mind that in addition to the sentence that authors SHOULD use one
>> of the other techniques, the listing of the summary attribute means that
>> authors MAY use it. I think this makes sense given the guidance.
>
> "MAY do a" + "SHOULD NOT do a" does not work, using RFC2119 terminology.
>
> Either it's ok, or not. If it's ok in some cases (which it is), then "should
> not" is the wrong way to talk about it.
>
>>> ...which I think is the wrong thing to do if one believes that @summary
>>> *does* have a special purpose for screen readers, which none of the
>>> alternatives have.
>>
>> If you read the following guidance, I think readers get the right advice
>> on the whole.
>
> I appreciate the advice, but in the end the spec still says "should not",
> lists it in the "obsolete, but conforming" section, and points out it will
> produce an error.
>
> So, from that point of view, we are almost where we were seven days ago.

Bingo, Julian got it in one. There's been a lot of word usage, and
some interesting massaging of text, but if you boil it down to its
simplest, we're in the same place we were seven days ago.

>
>>> Furthermore, the spec still lists @summary under "obsolete but
>>> conforming".
>>
>> To my reading, the warning for @summary is mentioned under "obsolete but
>> conforming", as a Note, with all the other warnings, but @summary is not
>> labeled "obsolete". It is listed as a conforming attribute (with some
>> guidelines for proper use). So I would say summary is "mentioned" rather
>> than "listed" in that section.
>
> So why is it "mentioned" then?
>
> BR, Julian
>
>

Shelley

Received on Thursday, 6 August 2009 17:23:21 UTC