Re: [HTMLWG] CR Exit Criteria redux

On Sep 20, 2012, at 7:45 PM, Boris Zbarsky <bzbarsky@MIT.EDU> wrote:

> On 9/20/12 10:40 PM, Adrian Bateman wrote:
>> If you're worried about this in your implementation then don't submit it as
>> a candidate for demonstrating interop.
> 
> The other part I'm worried about is people who want to "just get the spec done" submitting their implementations before they really have enough implementation experience with them...
> 
>> This does bring up a good point though - I think we should be clear that the
>> person or organisation responsible for the feature must be the one that
>> contributes to the interop testing.
> 
> Yes, that would be a good start.

Interesting discussion. I tend to agree that a one month period is somewhat arbitrary. I think requiring some form of certification from the responsible organization would serve a similar purpose. However, it seems to me such a rule would be hard to apply in some cases:

(1) What about software that doesn't have a controlling organization that can speak for it? WebKit has no single controlling organization, but almost any implementation claim for Safari or Chrome (or many other browsers) will be based on WebKit code. If a feature is started by a RIM engineer, completed by an Apple engineer, and then appears in a Google Chrome Canary build, is it RIM, Google, or Apple that has to speak for it? Or all three? Or any of the three?

(2) What if someone wants to use a combination of separately developed software to make an implementation claim? For example, what if someone says the combo of Firefox for Windows and JAWS implements a feature? Who would endorse the interop claim in this case, Mozilla or Freedom Scientific, or both? What if neither Mozilla or Freedom Scientific are motivated, but some third party is?

Note that both of these examples are likely to be scenarios we actually face for at least some of our specs, not just hypothetical edge cases.

Boris, Adrian, others, do any of you have suggestions for a form of "endorsement by implementor" that would give clear answers to these problem cases?

Regards,
Maciej

Received on Friday, 21 September 2012 04:37:17 UTC