Re: [picocontainer-dev] package renaming
Erik Putrycz <[email protected]>
| Newsgroups | gmane.comp.java.picocontainer.devel |
|---|---|
| Organization | National Research Council Canada |
| Message-ID | <[email protected]> |
Paul Hammant wrote: >>> The old DPC tried too much. On one hand we tried to set it up, how we >>> thought Pico should be used and on the other hand we tried to make >>> it so >>> flexible that you could change almost anything. So my advice is: >>> Split it >>> up! >>> Create an abstract PC implementation that provides the flexibility and >>> derive from it the specialized, but now inflexible container and >>> name it >>> properly. Now every can have a look why the container is setup in this >>> special way and create easily an own one, if the one we deliver (or the >>> ones we deliver) does not match any of the needs. >> >> +1 > > OK, I've deleted the contentious containers. I always intended to, > after making NanoContainer extent MutablePicoContainer (and reuse > methods). > > Here is the list of implementors of MutablePicoContainer :- > > DefaultPicoContainer implements MutablePicoContainer > DefaultNanoContainer implements NanoConatiner (which extends > MutablePicoContainer) and delegates to a DefaultPicoContainer instance > CommonsLoggingTracingContainerDecorator > Log4JTracingContainerDecorator > > That's it, there are no others :-) > > If we go back to a composed DefaultPicoContainer (that's what the > first alphas of Pico we're back in 2003), then I'm cool with that. Now sure how that looked like, but if there was less inheritance involved then I think it is better. > MutablePicoContainer becomes PicoContainer Yes! That's cool. > But not on... > > PicoContainer becomes 'Components' That's actually quite elegant. Components components = ... MyTable t = components.getComponent(MyTable.class) Erik