[dnssd] Review comments on hybrid-00

Markus Stenberg <markus.stenberg@iki.fi> Thu, 07 May 2015 09:32 UTC

From: Markus Stenberg <markus.stenberg@iki.fi>
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
Date: Thu, 07 May 2015 12:31:30 +0300
Message-Id: <AF9D96B5-0537-4EAD-AE3B-186B89CCC0F5@iki.fi>
To: dnssd@ietf.org
Mime-Version: 1.0 (Mac OS X Mail 8.2 \(2098\))
Archived-At: <http://mailarchive.ietf.org/arch/msg/dnssd/EZYMxxxwLBv7-_LkrW1aU_acOF0>
Cc: Markus Stenberg <markus.stenberg@iki.fi>
Subject: [dnssd] Review comments on hybrid-00
Precedence: list

Section 2 seems to turn from terminology to something quite different; perhaps new section header is needed?

Section 3.2:

I am not sure what’s up with A/AAAA magic handling - it seems weird to me, and I am just brutally using UTF-8 for everything. It is a MUST though, so I guess I have to live with breaking the draft compatibility (or actually care and fix it :-p).

Section 3.5+:

I would remove the LLQ reference and just make more generic description which both LLQ and push would fill.

Section 3.5:

Also, I would add logic to CF based RR handling; those can be answered immediately if in cache, or if received (that’s what our implementation does at any rate).

I am not sure I am too happy with 10 second default TTL for non-LLQ client case.

Also the whole ’three times’ sounds highly suspicious to me; why? Client probably sends later requests in any case (and there is no e.g. retry delay specified; is it just supposed to send 3 multicasts immediately? or what? hmm).

Cheers,

-Markus

[dnssd] Review comments on hybrid-00 Markus Stenberg