1088910 - Default http/2 off on gecko 34 after EARLY_BETA

Status: RESOLVED FIXED

(Core :: Networking: HTTP, defect)

Description

[Patrick McManus [:mcmanus]]

http/2 as draft-14 has been riding the trains quite successfully - we've had good testing with twitter, jetty, google, and others. That testing has largely validated the implementation, it has also fixed both client and server side bugs and even a couple of spec bugs.

unfortunately the IETF isn't ready to say http/2 is done yet. A clear ETA is fuzzy.

Sending draft-14 out into the wild on the release channel in that state might be ok, because its versioned explicitly as a draft so it wouldn't create a protocol we couldn't update, its not appropriate to think that servers are going to treat it with the appropriate level of production respect that it would need on the release channel... 

the level of server side support appropriate for the pre-release channels is different than release.

So I'm going to hold it back another 6 weeks with the consent of the release managers.

The diff attached to this bug will leave draft -14 support in the 34 release, but it will disable it after EARLY_BETA (maximizing whatever testing we can get from our pre-release audience). It leaves support enabled on 35 and 36.

Comment 1

[Patrick McManus [:mcmanus]]

Attachment: 0001-bug-1088910-default-http-2-off-on-gecko-34-after-EAR.patch

Comment 2

[Patrick McManus [:mcmanus]]

Comment on attachment 8511341 [details] [diff] [review]
0001-bug-1088910-default-http-2-off-on-gecko-34-after-EAR.patch

Approval Request Comment

see comment 0 and general bug description

[Feature/regressing bug #]:
[User impact if declined]:
[Describe test coverage new/current, TBPL]:
[Risks and why]: 
[String/UUID change made/needed]:

Comment 3

[Lawrence Mandel [:lmandel] (use needinfo)]

Comment on attachment 8511341 [details] [diff] [review]
0001-bug-1088910-default-http-2-off-on-gecko-34-after-EAR.patch

Beta+

Do you want to land the same change on m-c and Aurora so that we will only ship this feature once we deliberately remove the flag?

Comment 4

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-beta/rev/74298f48759a

Comment 5

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-b2g34_v2_1/rev/74298f48759a

Comment 6

[Patrick McManus [:mcmanus]]

(In reply to Lawrence Mandel [:lmandel] (use needinfo) from comment #3)
> Comment on attachment 8511341 [details] [diff] [review]
> 0001-bug-1088910-default-http-2-off-on-gecko-34-after-EAR.patch
> 
> Beta+
> 
> Do you want to land the same change on m-c and Aurora so that we will only
> ship this feature once we deliberately remove the flag?

I've been talking with some standards people and I expect this feature will be binary compatible with the end result of the standards process - so let's let it ride by default at this point. If events warrant we can make a change but I don't think that will be necessary.

Comment 7

[Martin Thomson [:mt:]]

(In reply to Patrick McManus [:mcmanus] from comment #6)
> I've been talking with some standards people and I expect this feature will
> be binary compatible with the end result of the standards process

I can confirm that.  (Editor hat on)

Comment 8

[Lawrence Mandel [:lmandel] (use needinfo)]

OK. The feature will ship in 35.

Comment 9

[Florian Bender]

Nice! How do we relnote this the right way?