cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stefano Mazzocchi <stef...@apache.org>
Subject Re: [C2] Cache behaviour
Date Thu, 05 Oct 2000 13:04:43 GMT
Ross Burton wrote:
> 
> Hi,
> 
> One of the problems with C2 was that no cache information was output to

You mean C1, right?

> the outside world, i.e. pages which never change and pages which change
> ever 10 seconds have the same cache parameters in HTTP: none.

Correct. The reactor designed didn't allow internal stages to have
access to the response headers, this made impossible for producers (the
owners of the caching concern) to provide caching information to the
external world but only to the internal cache using hasChanged().

This was identified as a problem and this is why the Enviornment object
in C2 provides access to both request and response to all stages (while
protecting the output stream from all but serializers).
 
> With C2 we can change this!

Correct.

> Somehow C2 needs to know the lifetime of
> the data it is processing and include it with the HTTP headers. This
> seems intimately related to the internal cache so it might be wise to
> discuss these together...

Yes, I agree.
 
> So: how is the internal cache going to cache data, how will it know when
> to refresh (are we keeping the C1 interfaces or creating new ones) and
> can this data be used in the headers, or will be sitemap need a
> <map:cache/> tag?

The internal cache will work more or less like C1's but with a few
differences:

 1) sitemap components will use the Modificable interface from Avalon to
signal their behavior and allow the cache to provide caching for dynamic
resources as well
 2) sitemap components have direct access to the HTTP headers so they
can provide fine-tuned caching capabilities without having to access any
cocoon internal

but I agree that it should be easier to integrate the two things since
they are (even if loosely) connected.

The XSP 1.1 specification will have a built-in caching behavior that
will take care of both internal and external caching... any proposal in
this sense is welcome.

About the <map:cache>, I don't think the sitemap owns this concern...
but I'm not sure... what do you guys think about this?

-- 
Stefano Mazzocchi      One must still have chaos in oneself to be
                          able to give birth to a dancing star.
<stefano@apache.org>                             Friedrich Nietzsche
--------------------------------------------------------------------
 Missed us in Orlando? Make it up with ApacheCON Europe in London!
------------------------- http://ApacheCon.Com ---------------------


Mime
View raw message