Resent-Date: Wed, 29 May 2013 06:44:03 +0000
Resent-Message-Id: <E1Uha79-0004Og-0U@frink.w3.org>
Content-Type: text/plain; charset=iso-8859-1
Mime-Version: 1.0 (Mac OS X Mail 6.3 \(1503\))
From: Mark Nottingham <mnot@mnot.net>
In-Reply-To: <845DDED8-04F7-42D0-834A-9F0AD0CAD556@mnot.net>
Date: Wed, 29 May 2013 16:43:18 +1000
Cc: "ietf-http-wg@w3.org Group" <ietf-http-wg@w3.org>
Content-Transfer-Encoding: quoted-printable
Message-Id: <F849403F-019A-4B72-AB21-E75C29B38789@mnot.net>
References: <D528E812-45A0-426F-972F-3F4AC6F8DEA7@mnot.net>
 <519413CF.7010007@measurement-factory.com>
 <A73CEF6C-EE77-462C-BA7E-80147A948D31@mnot.net>
 <51907161-4A59-4225-A61D-414A322DB589@mnot.net>
 <519A4BE8.9070700@measurement-factory.com>
 <845DDED8-04F7-42D0-834A-9F0AD0CAD556@mnot.net>
To: Alex Rousskov <rousskov@measurement-factory.com>
Received-SPF: pass client-ip=216.86.168.182; envelope-from=mnot@mnot.net;
 helo=mxout-07.mxes.net
Subject: Re: #486: Requiring proxies to process warn-date
Archived-At: <http://www.w3.org/mid/F849403F-019A-4B72-AB21-E75C29B38789@mnot.net>
Resent-From: ietf-http-wg@w3.org
Resent-Sender: ietf-http-wg-request@w3.org
Precedence: list

OK, I've marked this for incorporation; see =
<http://trac.tools.ietf.org/wg/httpbis/trac/ticket/486#comment:3> for =
latest text.


On 28/05/2013, at 3:36 PM, Mark Nottingham <mnot@mnot.net> wrote:

> Anyone have strong feelings about this?=20
>=20
> Personally - it seems reasonable to me to shift the overhead for =
assuring Warning correctness to those consuming it, since practically, =
they need to anyway today (as intermediaries don't implement this at =
all, IME).
>=20
> Cheers,
>=20
>=20
>=20
> On 21/05/2013, at 2:14 AM, Alex Rousskov =
<rousskov@measurement-factory.com> wrote:
>=20
>> On 05/19/2013 07:36 PM, Mark Nottingham wrote:
>>> On 17/05/2013, at 11:40 AM, Mark Nottingham <mnot@mnot.net> wrote:
>>>=20
>>>>>> and the second to be "If a recipient receives...", also removing
>>>>>> "forwarding" later down.
>>>>>=20
>>>>> This would not be sufficient because "using" may be interpreted to
>>>>> include "forwarding". How about this:
>>>>>=20
>>>>> "A response sender MUST NOT generate warning-value with a =
warn-date
>>>>> different from the Date value in the response. A cache MUST NOT =
send a
>>>>> warning-value with a warn-date different from the Date value in =
the
>>>>> from-cache response. A recipient MUST ignore a warning-value with =
a
>>>>> warn-date different from the Date value in the response."
>>>>>=20
>>>>> Would that cover all important cases without being too restrictive =
(like
>>>>> requiring the cache not to store something when there is no harm =
in
>>>>> storing, only in serving from the cache)?
>>=20
>>>> I think so, although we should remove the first 'response'.
>>=20
>>> Looking at this again:
>>=20
>>>> A cache MUST NOT send a warning-value with a warn-date different =
from the Date value in the from-cache response.
>>=20
>>=20
>>> This has the effect of requiring caches to check warning-values in
>>> all cached responses; do we still want to require that?
>>=20
>>=20
>> Good question. I do not know what the use cases behind the original
>> MUSTs were, and whether new use cases appeared since then, so I =
cannot
>> answer this question. If there is consensus that policing Warnings by
>> intermediaries (including caches) is not needed, then yes, we can =
remove
>> the entire MUST NOT.
>>=20
>>=20
>> Cheers,
>>=20
>> Alex.
>>=20
>>=20
>=20
> --
> Mark Nottingham   http://www.mnot.net/
>=20
>=20
>=20
>=20

--
Mark Nottingham   http://www.mnot.net/