Re: Request to publish HTML+RDFa (draft 3) as FPWD

On Tue, Sep 22, 2009 at 1:20 AM, Julian Reschke <julian.reschke@gmx.de> wrote:
> Jonas Sicking wrote:
>>>>
>>>> By the HTML WG's Design Principles, error handling needs to be defined.
>>>> It's not good enough to say that it's an error, so don't do that.
>>>
>>> The design principles also say "pave the cowpaths", yet HTML5 defines an
>>> entirely new syntax.
>>>
>>> So citing the principles doesn't seem to be productive here.
>>
>> Isn't paving cowpaths *exactly* what we're doing here? HTML5 describes
>> the syntax used for the billions of HTML documents browsed by people
>> every day better than any other specification draft I've ever seen.
>> And I've worked decently closely with the HTML parser in Gecko for
>> many years now.
>>
>> For the first time ever there is actually a document which we can
>> check our implementation against and fix bugs, rather than having to
>> guess what most pages expect.
>>
>> If that's not paving a cowpath then I guess we have a different
>> understanding of that principle.
>>
>> Additionally, I don't understand your logic that if we in some part of
>> the spec had decided not to follow a principle (which I maintain isn't
>> the case here), then we should throw out an entirely different
>> principle in an entirely different context?
>
> I didn't intend to say that.
>
> When I said "The design principles also say "pave the cowpaths", yet HTML5
> defines an entirely new syntax." I referred to Microdata, which is in direct
> competition in RDFa, and which clearly violates another design principle.
>
> So, statements with respect to "blocking" publication of a WF for RDFa in
> HTML because it is in conflict with one design principle do not compute for
> me. Unless we apply the same rule for Microdata.

Ah, sorry about the misunderstanding.

If I understand Henris email correctly though, he's not opposing the
publication because of violating the "define error handling"
principle, but because it doesn't define processing of RDFa in HTML
documents at all.

I realize that people seem to have different opinion on if processing
of RDFa is defined (though it's a unclear to me if people are arguing
that it's Manu's draft or if it's the original XHTML+RDFa
recommendation that defines processing), but Henri's interpretation
seems to be that it is not defined. As is mine.

Or I'm misunderstanding Henri entirely, which of course also is a possibility.

/ Jonas

Received on Tuesday, 22 September 2009 08:27:12 UTC