W3C home > Mailing lists > Public > ietf-http-wg@w3.org > January to March 2015

Re: Editorial: ALTSVC HTTP/2 Frame.

From: Bence Béky <bnc@chromium.org>
Date: Wed, 14 Jan 2015 15:24:27 -0500
Message-ID: <CACMu3to24+Kp-FOrJHj+vVkAC-dhjwcPDFnmg1LeYPgqHZhEfg@mail.gmail.com>
To: Erik Nygren <erik@nygren.org>
Cc: Martin Thomson <martin.thomson@gmail.com>, Julian Reschke <julian.reschke@gmx.de>, HTTP <ietf-http-wg@w3.org>
Created a commit at
https://github.com/httpwg/http-extensions/pull/39/commits.  Martin, is
this what you had in mind?

On Thu, Jan 8, 2015 at 2:13 PM, Erik Nygren <erik@nygren.org> wrote:
> What Martin suggests here makes sense to me.  Good catch, Bence.
>
> On Thu, Jan 8, 2015 at 11:46 AM, Martin Thomson <martin.thomson@gmail.com>
> wrote:
>>
>> On 8 January 2015 at 08:22, Julian Reschke <julian.reschke@gmx.de> wrote:
>> > It can contain what the header field contains, and that allows
>> > specifying
>> > the authority (uri-host and port).
>> >
>> > Am I missing something?
>>
>> It would seem so.
>>
>> The header field doesn't contain origin information because that is
>> derived from the request.  The header field says: "for requests to
>> *the origin of this request*, please send here: list of service
>> endpoint tuples"; when the frame is sent on stream 0, it lacks that
>> context.
>>
>> I suggest that you define the frame as length (sadly, 255 bytes isn't
>> long enough to hold the longest conceivable origin string) plus origin
>> plus the remainder being exactly the same as the Alt-Svc header field.
>>
>
Received on Wednesday, 14 January 2015 20:25:19 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:14:42 UTC