Re: [RDFa] ISSUE-3 @class and @role for rdf:type

Mark,

You are tempting me to reopen the issue, since you are mostly voting for
the status quo, but my chair responsibilities force me to refuse to fall
for your smooth ways :) Let me re-frame my reasoning, because I believe
this will be important for the principle of how we resolve other issues.

@href anywhere was *not* resolved in May or in the Primer or in the
talks. Adding @href everywhere does not create a moving target, it adds
a new way to express triples that may otherwise be difficult to achieve.
All existing RDFa in the Primer, in our examples, still work. All
triples generated by Operator are still good (though some new ones may
be missed if the spec was implemented very strictly.)

If we take away @class, we break almost everything, including most of
our successes to date. That's a big risk to take, and one we agreed we
couldn't take anymore.

So we're not just ratifying existing stuff only, but we must heavily
lean towards not breaking big items, as we agreed on 5/31, with every
core member of the task force on the call, minus Elias, who later
expressed significant "moving target" worries in a separate discussion.

I know, you must hate me right now, but that is the unfortunate destiny
 of most W3C chairs, as I understand it :)

-Ben

Mark Birbeck wrote:
> Hi Ben,
> 
> I'm not with you...are we trying to resolve all syntax issues in the
> next two weeks, or just ratifying what we have? We could have avoided
> a lot of discussion on '@href everywhere' if we were merely ratifying
> what we already had, but people in the group seemed to think it was
> actually important enough to discuss. (And I spent a long time trying
> to find a compromise.)
> 
> As before, I'm not saying I'm totally against sticking with @class,
> even though I think @isA is slightly easier to understand. I'm only
> saying that we don't need to be bounced into a conclusion simply on
> the basis that it shouldn't be discussed. (Particularly when there
> have been consistent objections from Steven all the way through.)
> 
> I suggest we put it to a vote, and move on. :)
> 
> If I could vote for the various positions, I would opt for:
> 
>  +1 for a new attribute
>  + 0.9 for using @class as we have it now :)
>  -1 for using @role.
> 
> Regards,
> 
> Mark
> 
> On 28/06/07, Ben Adida <ben@adida.net> wrote:
>> Mark Birbeck wrote:
>> > Hi Ben,
>> >
>> > I'm not sure I agree. Sure we've done talks, but the main source of
>> > information for people, going forward, will be the primer and our
>> > various specs.
>>
>> Mark,
>>
>> I need to be a bit of a stickler here, because we all talked about this
>> in our May 31st conference call after all the May talks, and we agreed
>> that we couldn't make RDFa a moving target:
>>
>> http://www.w3.org/2007/05/31-rdfa-minutes.html
>>
>> Removing the meaning of @class would make RDFa very much a moving
>> target, and, no matter how technically beautiful an alternate solution
>> is, we would lose a significant chunk of folks who are beginning to
>> depend on the syntax.
>>
>> In other words, I think our "last chance" to change this particular,
>> central RDFa issue has passed.
>>
>> -Ben
>>
>>
> 
>

Received on Thursday, 28 June 2007 17:14:35 UTC