Rafal Krzewski wrote:
>
> Raphael Luta wrote:
> >
> > Jon Stevens wrote:
> > >
> > > I'm going to make the Turbine.java class a final class. People should NOT be
> > > subclassing it as there is absolutely NO reason to do so (or at least tell
> > > me a good reason as I can't think of one) and the Jetspeed people seem to be
> > > into this for some reason. I think we should nip this idea right in the bud.
>
> +1 to making it final. Subclassing the servlet subverts the whole
> Turbine model.
>
> > There are good reasons (just read the thread on this subject in the Jetspeed
> > mailing-list). Basically it boils down to: while we move most of our
> > components to services, we must initialize them with our own initialization
> > code (it was currently done in the Layout: very ugly !)
> > As Jeff Brekke pointed out, it can be done with a JetspeedService but that
> > would be an awful hack because such a service would have *no* function,
> > only initialization code. Subclassing Turbine in order to override the
> > init() and destroy() calls is much cleaner.
>
> Can you elaborate more what is the problem that you have with
> initializing
> your services? The framework for initialization/deinitialization of
> services
> is in the Turbine has been in place for more than three months. (If you
> are not
> familiar with it, check out current Turbine sources from CVS, build
> javadocs
> and read the Javadocs for package org.apache.turbine.services).
>
> If there is something that can be improved in this framework to suit
> your
> needs better, I'm the person who wants to know about it!
>
We have no problems with the Service infrastructure but a lot of code
sits in the CVS that *should* be Turbine services but currently are not
and are in the process of being moved as service.
During this process, we still need to initialize the components that are not
yet services and this is done by overriding the Turbine servlet init() method.
When all the components are moved, this init() should be empty and thus
we can remove the subclass.
The only alternative I see to this process is writing a new service that
would only initialize all the other components without providing any real
"service" method, but I find this solution much more a perversion of the
Turbine service model than the subclass one.
--
Rapha�l Luta - [EMAIL PROTECTED]
------------------------------------------------------------
To subscribe: [EMAIL PROTECTED]
To unsubscribe: [EMAIL PROTECTED]
Search: <http://www.mail-archive.com/turbine%40list.working-dogs.com/>
Problems?: [EMAIL PROTECTED]
