RE: [picocontainer-dev] Nano2 (baby steps - work in progress) 'config' based entries in component tree.
"Michael Rimov" <[email protected]>
| Newsgroups | gmane.comp.java.picocontainer.devel |
|---|---|
| Organization | Centerline Computers, Inc. |
| Message-ID | <010301c7981c$134eb520$39ec1f60$@com> |
> I prefer config (in addition to component) as it reminds us that
> Inversion of Control is three things - dependency, lifecycle and
> configuration.
Agreed.
That said, however,
I would say that PicoContainer and the current building methods only truly
handles Dependency Injection. (Yes, we have lifecycle, but I don't think
anybody has had a real revelation of how to do it as elegantly as we have
done CDI)
I personally have problems with putting configuration into the build script.
For example: I don't want admins mucking with the composition of a
complicated program just to change a database connection password!
To me, I explicitly put all configuration in a separate location and apply
that configuration after that object is constructed. (Closest thing in Pico
is BeanSetter CA)
But the real point here is that I let ComponentAdapters handle the
configuration, not the composition script, and I think we're walking down
the wrong path by including it.
-Mike
---------------------------------------------------------------------
To unsubscribe from this list please visit:
http://xircles.codehaus.org/manage_email