Re: [widgets] Conformance Checker assertions spec

On Sep 30, 2009, at 12:10 , Marcos Caceres wrote:
> Robin Berjon wrote:
>> I like this plan. My primary concern is that we communicate it  
>> clearly
>> to third parties.
>
> Agreed. However, CR is a call for implementation - it is not a sign  
> of complete stability.

I know, but that does not prevent us from providing good information.

>> Also, do we have some visibility on which implementations (and how  
>> many)
>> stand a chance of passing the TS?
>
> I have not run it through Opera's internal imp yet. But don't  
> foresee any issues passing.
> If we can get Wookie to pass too, then we have the two  
> implementations needed to proceed to PR.
>
> It should not affect third parties too much if we time it perfectly:  
> That is, republish LC once both implementations are passing every  
> test in the test suite. Then, during the first week LC period we  
> slap together the implementation report and have it ready to submit  
> to the W3C for a speedy, yet process-proof, transition to PR.

I'd like to know what Access thinks here. I'd also like to get input  
from known implementers from outside the WG (e.g. Samsung).

Your plan only fails to affect implementers if a) everyone is more or  
less tracking the WG, and b) the draft doesn't change too much.

I can go with the plan of a short LC taking place after we've fixed  
bugs in the ED, but only if those are relatively containable bugs  
(i.e. we're not rewriting the whole processing model) and they're well- 
advertised, or the concern a separate product class. Someone who  
thought we'd be in CR for 6 months shouldn't come back in a few  
months' time to find that we've tricked them.

Also, does going to LC again re-open an exclusionary period? I think  
we can't go to Rec (and probably not to PR) until the exclusionary  
period is over. Dom?

-- 
Robin Berjon - http://berjon.com/

Received on Wednesday, 30 September 2009 13:33:11 UTC