Re: [Webtransport] Confirming Consensus on WebTransport protocols

On Tue, Jan 12, 2021 at 7:02 PM Martin Thomson <mt@lowentropy.net> wrote:

> On Wed, Jan 13, 2021, at 04:25, David Schinazi wrote:
> > The consensus was option 1A.
>
> Great!
>
> > The consensus was option 2A.
>
> I can live with that.  Can I request a brief summary of the reasons?
>

I asked a question like this during the interim, and I think it'd be worth
clearly documenting the reasons, as I think it'd clarify our direction a
bit.  The one that I clearly recall is that we'd have to re-invent many of
the features that HTTP/3 provides, but I'm not clear why that's the case.

>
> > Additionally, we discussed pooling and decided that we would
> > not allow pooling in WebTransport over HTTP/3 for now, due
> > to the complexities of pooling.
>
> I don't think that this group gets to say that.  If we are using HTTP,
> then we are using HTTP.  We might decide that in browsers we want to keep
> the connections separate for $reasons, but I am opposed to saying that.
>

I suspect you're right.

I'd like to call out that there are two things which could be referred to
as 'pooling'.  One is using a single connection for HTTP/3 and WebTransport
at the same time.  The other is support for multiple WebTransport sessions
inside a single connection.  This latter type of 'pooling' is part of what
makes Http3Transport more complex than QuicTransport.  Was there ever an
explicit question about whether multiple sessions on a single connection
are necessary or even useful?  In retrospect, we were picking between two
transports as well as two different feature sets and it might be worth
splitting them apart.

> Unless of course I'm missing context in the same way I missed the meeting.
>
> --
> Webtransport mailing list
> Webtransport@ietf.org
> https://www.ietf.org/mailman/listinfo/webtransport
>