On 09/30/2011 09:30 PM, Amos Jeffries wrote:
> On Thu, 29 Sep 2011 17:59:13 +0200, Kinkie wrote:
>> On Thu, Sep 29, 2011 at 5:55 PM, Alex Rousskov wrote:
>>>>> We got a malformed max-stale value. We have only
>>>>> two options when it comes to that value interpretation, I think:
>>>>>
>>>>> A1. Treat it as valueless max-stale.
>>>>> A2. Ignore it completely.
>>>>>
>>>>> Since valueless max-stale results in possibly very stale content being
>>>>> served to an unsuspecting client, I think #2 is much safer (and
>>>>> compliant with HTTP).
>>>>> BTW, we have three options when it comes to forwarding the malformed
>>>>> directive:
>>>>>
>>>>> B1. Forward our own valid directive.
>>>>> B2. Forward nothing.
>>>>> B3. Forward the malformed directive.
>>>>>
>>>>> While option B3 could be the best, it requires more parsing changes.
>>>>> Option B2 is much better than B1, IMO, and requires minimal changes.
>>>>>
>>>>> We currently implement A1 and B1.
Hi Amos,
I sense misconmmunication. Please keep in mind that A* options are
concerned with internal interpretation of a malformed CC directive by
Squid while B* options are concerned with forwarding of a malformed CC
directive by Squid. We treat those two areas separately, and we do not
discuss a whole CC header.
> B2 is not a valid option AFAIK.
Why not? We can drop a CC directive if we consider it malformed, I
think. And if it is the only directive, we would drop the whole CC
header, of course.
In fact, don't we already implement B2 for all other malformed
integer-valued CC directives?
> HTTP does not offer the option A2 as you are discussing right now.
> Since we know the header we must either treat it as an option known (A1)
> or something unknown (A3).
A2 is "ignore it", which (from interpretation point of view) is
equivalent to "treat it as something unknown" because Squid ignores
unknown CC directives.
> * A3 means we can ignore it for operational purposes (like A2), the
> natural result being B3.
I do not understand the difference between A2 and A3. A2 does not imply
B3 though. A2 implies (either B2 or B3).
> A1 means we have to decide betrween B1 or B3. Treating it as a known
> directive which has been extended with some unknown extension value
> (B3). Or as definitely malformed and invalid (B1).
A1 can be combined with all three Bs. I would not recommend A1 at all
though because we may end up serving stale content to a client that does
not want it. If we have to gamble, it is better to server "too fresh"
than "too stale" content, IMO (it is also compliant!).
Thank you,
Alex.
