excalibur-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Carsten Ziegeler" <cziege...@s-und-n.de>
Subject RE: [RT] About proxies and pooling
Date Fri, 25 Jun 2004 07:57:18 GMT
Hamilton Verissimo de Oliveira wrote:
> Well, dude. I can only speak for the proxy subject, I abstain 
> from Pooled Components discussion :-)
> I've been thinking about having no proxies at all. Proxy is 
> one of the symptoms of Too much magic(tm) and despite the 
> benefits (prevents users from casting, overriding finalize 
> can handle proper release) the perfomance, complexity and no 
> way to proxy a concrete implementation (without
> interfaces) counts against it.
Might be, but I think performance is in the case of Fortress
increased by using proxies. We could remove a serious bottle

> OTOH I'd like to add some Interceptor capabilities to 
> Fortress 2.0, then we have no choice: Proxy is the only way.

> So, instead of writing it now on stone, can we work on a 
> simple simple container - with no magic at all - and some 
> extended containers which may add this or that capabilities? 
> Better yet, leverage to components handlers this or that capability.
> Now it's up to you: WDYT ? :-)
Ok, actually this is - more or less - the way it is currently. You
can use the non-proxy approach or the proxy one and either get
this or that functionality.
Now, my personal opinion is that this is too much complexity :) If
we would go for "proxy always" it's simpler. Now, for the too much
magic argument, I'm not sure. This is internal magic that just
improves some things but it's hidden from the user.
Okay, anyways, what do others think?


To unsubscribe, e-mail: dev-unsubscribe@excalibur.apache.org
For additional commands, e-mail: dev-help@excalibur.apache.org
Apache Excalibur Project -- URL: http://excalibur.apache.org/

View raw message