Re: New I-D: Bootstrapping WebSockets with HTTP/3

(speaking solely as a QUIC/MASQUE/WebTransport enthusiast)

I agree with MT here. WebTransport is currently relying on Extended
CONNECT, and MASQUE is going to soon (assuming the MASQUE WG reaches
consensus on that, but everyone seemed to agree last time we discussed it -
I've been planning on writing PRs for CONNECT-UDP and CONNECT-IP but I got
distracted with other things). Over in WebTransport and MASQUE we had
somewhat forgotten about the fact that HTTP/2 SETTINGS need to be redefined
for HTTP/3, but that was an oversight. So I'm in favor of progressing
draft-hamilton-httpbis-h3-websockets as both WebTransport and MASQUE will
likely have a dependency on it. The sooner we get this in the pipeline, the
better.

David

On Mon, Aug 30, 2021 at 7:46 AM Ryan Hamilton <rch@google.com> wrote:

> On Sun, Aug 29, 2021 at 4:51 PM Martin Thomson <mt@lowentropy.net> wrote:
>
>> On Sat, Aug 28, 2021, at 04:19, Ryan Hamilton wrote:
>> > I'm not terribly familiar with the discussions in MASQUE or WebTrans
>> > about extended CONNECT. Are they intending to revise 8441 for
>> > WebSockets?
>>
>> My understanding is that they just plan to use it (though the recent
>> MASQUE CONNECT-IP update did not include that change).  I don't think that
>> your proposal is affected by this.  I would say carry on.
>>
>
> Thanks for the update! That's excellent. Ok, per Lucas's suggestion I've
> added text about stream closure and added a link to the relevant section of
> RFC 9000. I've updated the copy of the doc in github. Should I also push a
> new ID?
>
> Cheers,
>
> Ryan
>