On 2013-06-26 9:09 AM, Robert O'Callahan wrote:
On Google's side, it is harder to find examples because I do not follow
that as closely but requestAutocomplete() is an example of an API that
Mozilla might not look at for quite some time.
If we think these use cases are (or ever will be) relevant, we need to give
feedback even if we don't plan to implement them soon. We should at least
try to make sure these APIs are something we wouldn't feel bad about
implementing, and complain if they are.
I don't think that in practice we can expect other vendors to give us
feedback on all proposed APIs, or that we give them feedback on their
proposed APIs.
For whatever it's worth, Blink has made the decision to implement Web
MIDI without receiving any feedback from us (and to the best of my
knowledge from other vendors), and has also made the decision to ship
requestAutocomplete hoping that Hixie will add it to the spec before it
hits the Chrome release channel. In the case of Web MIDI at least they
have asked us to give feedback on a few occasions, but we've been unable
to do so because of not having enough resources. But it looks like
they're also interested in making progress on the features that are only
interesting to them while making an effort to solicit feedback from
other engine vendors.
If "no answer" means "we don't care about your use cases", then we can't
let that block our progress, because there are always going to be
use-cases
we need to solve that no other vendor is currently interested in.
I agree that getting blocked because other vendors don't care about our
use cases is not a good situation. However, it is easy to push bad APIs
because no one had time to look into it and then, when they do, having
two competing API which ends up in a mess (IDB vs WebSQL for example).
My point is not that we will do that on purpose but we are opening a
door to situations like these.
I think that door has to be opened.
AFAIK Apple never really solicited feedback on WebSQL before shipping.
Also, there was no attempt to create a proper spec. (A proper spec would
have included a spec for SQLite, so if they had done that, maybe we would
have adopted it after all!) So I don't think we would repeat the WebSQL
error under our proposed policy.
So I see your point, but I think it's fair to expect other vendors to
provide feedback on our API proposals if they care about the use-cases, and
for us to do the same for them. The available bandwidth and level of
investment is higher than it used to be.
So, are you proposing that we should give feedback on all APIs even if
they're not interesting to us? I'd be fine with that, but I don't think
we can assume that it will be reciprocated.
Cheers,
Ehsan
_______________________________________________
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform
