>> Have you considered having an official "OCCI Compatibility Test",
>> which could be an impartial arbiter of whether or not a given
>> implementation is spec compliant?
>
> Sure, that's a good idea but who's to say our implementation will be any
> better than any other? Bakeoffs tend to work well too.

In the absence of an official compatibility test, what must an
implementor do in order to be sure his implementation is spec
compliant?

Don't get me wrong - testing is important, just not today while we're trying to get a[n overdue] draft out the door. Correct functioning against a reference implementation would be a good start.

I'm not familiar with how a bakeoff works - can you please explain?

We put implementors in a [virtual] room together and point their work at each other, seeing where the wheels fall off and fixing it (and/or the spec itself). With any luck we'll be able to do some of this before the spec is finalised.

Sam