On Tue, May 6, 2008 at 10:14 PM, Rickard Öberg <[EMAIL PROTECTED]> wrote:
> This makes it much more explicit that it is the configuration that is > being used, and gives a good place for the refresh() method, which is > important for the Service to be able to refresh the state of the > configuration, if it has been changed externally through UI or similar. > > What do you think? Would this make more sense as a general Service > configuration mechanism? Any other options? > > The only problem with the above is the implementation, since the > interface is generic, i.e. Configuration<T> and so the implementation > must somehow figure out what T is mapped to in the injection, which is > non-trivial. But if we want it to work I'll make it work :-) Will it be generic?? I am a bit tired now, but the SCM would implement Configuration, no?? Either way, I agree this communicates intent a LOT better, and we are generally in favour of explicit intents than the "abstract away reality" in the name of "because we can" like others do. Cheers Niclas _______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
