cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Unico Hommes <un...@hippo.nl>
Subject Re: [vote] Reschedule the release
Date Fri, 14 May 2004 16:22:07 GMT


From: Sylvain Wallez <sylvain <at> apache.org>
Subject: Re: [vote] Reschedule the release 
<http://news.gmane.org/find-root.php?message_id=%3c40A4DB95.90908%40apache.org%3e>
Newsgroups: gmane.text.xml.cocoon.devel 
<http://news.gmane.org/gmane.text.xml.cocoon.devel>
Date: Fri, 14 May 2004 16:45:41 +0200

Carsten Ziegeler wrote:

>Given the current discussions about the problems in the
>persistent store, it seems not to be such a good idea
>to release today.
>
>So, I think we should simply reschedule the release. It
>seems that switching to EHCache is the most painless
>solution for now (JCSCache would be good as well but
>it seems that they still have problems and we shouldn't
>wait for them).
>
>So, what do you think about:
>- switching to EHCache now
>  
>

<SW>
+1 even if EHCache is not as "ASF-friendly" as JCS (referring to David's 
post) but just works (referring here to JCS shutdown problems)
</SW

Both EHCache and JCS are not designed to be reliable persistence stores in the excalibur store
sense. 
Both do supply disk-overflow mechanism when the maximum number of objects is exceeded (btw
I don't like).
But it they don't consider it a central concern for a caching system. I think I agree partly
with that.

JCS has the feature whereby the memory cache is written to the disk-overflow and the disk-overflow
is 
retained between server runs but it was previously broken (this was the JCS shutdown problem
mentioned above).
EHCache does not retain its cache after shutdown.

--
Unico


Mime
View raw message