Re: Late, but I have reservations.

Gannon Dick wrote:
> 
> 
>> What's different about these two situations? What
>> mischief is enabled by  the RDFa case that isn't already enabled by the
>> <link rel> case?
> There is nothing different but as I said ...
> ... I can not see any ****audience benefits****

Okay, so we agree there is nothing different, that is a good start :)

As for the benefits, I think I mentioned them in a previous email: as 
Bob DuCharme has pointed out, it is interesting to add metadata in the 
head of the document of content management systems. The IPTC initially 
expressed this interest. In terms of expressivity, this is no more 
powerful than <link rel>, but in terms of self-containment, RDFa is more 
advantageous.

So there are clearly benefits.

> annotation=injection in case you wondered about my use of that term.
> In addition aggregation can be a problem for those lucky enough to be
> making a living off their commercial endorsements.

This still doesn't make sense to me: who are the parties involved, and 
who is doing "injection?"

> I'm not saying that you can't reference RDF in the <head> now (like
> the link), and I'm not saying that RDFa is evil everywhere.  In the
> <body> of a document, used to improve the depth of your writing it is
> much too good an idea to pass up.  OTOH, when you write code in the
> <head> of HTML, who is the audience?

As mentioned above, the audience might be a content management system.

>  You don't need RDFa to forge a
> link between two of your documents if you want to do that, but RDFa
> in the <head> might enable a third party to aggregate information
> about you without ever looking at what you authored.

This doesn't make sense to me. A *third* party? What does it mean to 
aggregate information about me without looking at what I've authored?

You need to define the problem much more precisely, as right now it 
isn't clear at all. Plus, since you agree that this is no more powerful 
than <link rel> in the head, that implies that this is also no more of a 
threat than <link rel>.

> I do not think this is a good thing, but more to the point, does the W3C have to
> recommend a superfluous (in the <head> element)standardization?

It's not superfluous. Some folks have expressed legitimate use cases as 
detailed above.

Gannon, at this point, I have to recommend that we reject your comment 
because:

1) you agree that RDFa is no more expressive than <link rel>, which 
implies that it cannot be any more of a threat than <link rel> (note 
that I still don't understand the threat, but at least we know it's no 
worse than what we currently have.)

2) you may not see the point of RDFa in the head, but others do, as 
described above.

RDFa task force folks: do you agree with the above statement and thus 
resolution?

Gannon: I hope this statement clears up confusion, and that you can 
"live with" the answer. Let us know either way.

-Ben

Received on Tuesday, 5 August 2008 03:55:48 UTC