W3C home > Mailing lists > Public > ietf-http-wg@w3.org > October to December 2019

Re: [EXTERNAL] Re: HTTP Alternative Services Best Practices?

From: Lucas Pardue <lucaspardue.24.7@gmail.com>
Date: Tue, 17 Dec 2019 23:07:18 +0000
Message-ID: <CALGR9oY6DzjsH46AT6C4BXTx0MtH0JCsBsgrBeWRcjA4rXKGVA@mail.gmail.com>
To: Cory Nelson <Cory.Nelson@microsoft.com>
Cc: Mike Bishop <mbishop@evequefou.be>, Ryan Hamilton <rch=40google.com@dmarc.ietf.org>, QUIC WG <quic@ietf.org>, HTTP Working Group <ietf-http-wg@w3.org>
Hi Cory,

I think your questions are probably good implementation-oriented ones but
less about BCP to my mind. If there is something worth documenting, my
instinct is a separate document that compliments RFC7838.

To answer:

On Tue, 17 Dec 2019, 22:46 Cory Nelson, <Cory.Nelson@microsoft.com> wrote:

> Two questions I had while implementing Alt-Svc, and ended up looking at
> other implementations for guidance:
> Is it valid to have Alt-Svc: clear followed by, on another line, more
> Alt-Svc to replace the old services?
RFC7838 says about Alt-Svc:

"The field value consists either of a list of values, each of which
indicates one alternative service, or the keyword "clear"."

So I think it is clear that it is invalid to have both clear and an
alt-value. But what might be less clear is the expected failure mode if
this does actually occur.

> Is it valid for an authority to extend their lease by sending an Alt-Svc
> for themselves?
If you mean is it valid for a selected alternative to advertise itself and
therefore keep itself fresh, then yes, the alternative is fully
authoritative and I think that this is captured in the specification.


Received on Tuesday, 17 December 2019 23:07:28 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:15:43 UTC