Re: GRDDL PR or CR

Harry Halpin wrote:

> As for GRDDL Spec, I see no reason why it would have a PR phase, as we
> dropped the IETF dependency.
> 
> I don't see any reason, especially as implementers have been keeping up,
> that the PR phase would also be needed for GRDDL test cases or, say, the
> GRDDL Primer.
> 

Hi,

I think you mean CR phase not PR phase above

CR = Candidate Rec = Call for implementations
PR = Proposed Rec = Call for AC review

The final stages of the process are

LC = Last Call Working Draft
   At this stage the WG believes it has finished the design.
   Editorial changes may still be required.

Assuming the LC review raises no issues requiring substantive change (as 
I believe is likely with GRDDL), then

The Working Group proposes either CR or PR

After a successful CR the WG proposes PR

The entry requirements for PR are
+ a finished document that has been through LC review, and the comments 
have been addressed (not necessarily positively)
+ evidence of adequate implementation experience

The evidence usually consists of implementation reports, including a 
list of tests that have been pasted.

Once the document moves to PR it no longer is a WG document.
The director then controls the final stage of PR to Rec
The director liases with the AC during the AC review.
The director may consult with the WG concerning particular AC comments.

It is likely that we can gather such evidence in time for the schedule.
But ... if we want to rec track test cases, and Chime has argued the 
case convincingly (at least to me), then it seems a little out of order 
to ask for comments on the tests after having already totally frozen the 
spec. It also seems a little out of order to be asking an AC review on 
the one hand, when, we are asking the public to review a document, for 
which, at least formally, the public review may raise a torpedo, that 
requires a significant redesign.

Jeremy



-- 
Hewlett-Packard Limited
registered Office: Cain Road, Bracknell, Berks RG12 1HN
Registered No: 690597 England

Received on Wednesday, 28 March 2007 10:06:04 UTC