Mailing-List: contact dev-help@httpd.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@httpd.apache.org
Received-SPF: neutral (herse.apache.org: local policy)
Message-ID: <45B36A63.1040805@qbrick.com>
Date: Sun, 21 Jan 2007 14:28:03 +0100
From: Fredrik Widlund <fredrik.widlund@qbrick.com>
User-Agent: Thunderbird 1.5.0.9 (Windows/20061207)
MIME-Version: 1.0
To: dev@httpd.apache.org
Subject: Re: mod_cache+mod_rewrite behaviour
References: 
 <3B21A253728EA247A10A692547A271530E7548@EITO-MBX03.internal.vodafone.com>
 <45B0ABD5.1020807@qbrick.com> <eoqsqi$9cg$1@sea.gmane.org>
 <45B36860.3060705@apache.org>
In-Reply-To: <45B36860.3060705@apache.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 8bit

I wrote a patch friday, had a few minutes over to test it and it seemed 
to work fine. Of course it's a minimal patch, but I'll double check it 
tomorrow anyway, and try to convince you guys to accept it. :)

Kind regards,
Fredrik Widlund

Ruediger Pluem skrev:
> On 01/19/2007 05:49 PM, Scott MacVicar wrote:
>
>   
>> I already have a patch that we're using in production, attached are
>> patches for the 2.2 and trunk.
>>
>> I was under the impression that there wasn't any interest in a patch
>> like this hence I've not opened an issue to get this integrated into the
>> main package.
>>
>> The option is CacheStoreQueryString.
>>     
>
> The option you introduce does not really solve Fredrik's problem.
> Your patch allows responses to requests with a query string to be cached
> if the response does not contain an Expires header.
> As far as I understood him the responses get cached (they seem to contain
> an Expires header), but he wants to avoid that the query string gets
> incorporated in the cache key. E.g.
>
> /abc.html?x=a
> /abc.html?x=b
>
> result in two different cache entities. He wants them to be the same, or more
> generally he wants all adjustments that have been made to the URI / query string
> during the processing to be considered for the generation of the cache key.
> Currently the original URI / query string is used for this.
>
> BTW: Although it is not exactly the same your problem can be solved without
> the patch by either:
>
> - Making the content handler (whether in httpd or somewhere else) set an Expires
>   header in the response.
> - If this is not possible let mod_expires set one.
>
> Regards
>
> R�diger
>