Mailing-List: contact cocoon-dev-help@xml.apache.org; run by ezmlm
Precedence: bulk
Reply-To: cocoon-dev@xml.apache.org
From: maciejka@tiger.com.pl
To: cocoon-dev@xml.apache.org
Date: Thu, 5 Jul 2001 14:51:43 +0200
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
Subject: Re: AW: CachingCIncludeTransformer
Reply-to: maciejka@tiger.com.pl
Message-ID: <3B447EFF.29249.B76C9F@localhost>
Priority: normal
In-reply-to: <GMEBIBHGAOFGJCDPJANDGEHDCEAA.cziegeler@sundn.de>
References: <3B447012.27123.7D1BED@localhost>

On 5 Jul 2001, at 14:17, Carsten Ziegeler wrote:

> Hi,
> 
> are you sure that the CachingCIncludeTransformer is really working?
> As far as I see the validity object gets all its information when
> the first response is generated (the timestamps are added).
> This hack works.
> 
> But when the second response comes in, the new validity object
> for this response is empty. It is compared with the one from the
> first run and they are not equal.

New validity object is not compared with the old one. Old one contain two lists:
source urls and timestamps. When it comes to checking validity of first 
response we know that generation/transformation steps before 
CIncludeTransformer are valid (otherwise we would have had discarded cached 
pipeline already) so source url list of new validity will be the same as of old 
one. Only timestamps have to be recalculated and compared. 

> The second validity object gets the timestamps after the
> response is generated (or during the generation).
> 
> So as far as I understood your algorithm, I should cache the
> correct result due to the hack but this is never retrieved from
> the cache.
> 
> Or did I oversee something?
> 


Maciek Kaminski
maciejka@tiger.com.pl

---------------------------------------------------------------------
To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org
For additional commands, email: cocoon-dev-help@xml.apache.org