[Grok-dev] Re: final sprint report - please review my branch!
faassen at startifact.com
Tue Apr 8 18:01:40 EDT 2008
Leonardo Rochael Almeida wrote:
> Would it make any difference one way or the other if we used
> "classProvides()" instead of "implements()" for this? Conceptually, I
> think it makes more sense to declare that a class provides a certain
> interface if we want to grok it. "implements()" declares things about
> what the future instance provides instead of declaring things about
> the class.
A good point. The only trouble is that classProvides() is a bit more
verbose and obscure, but I'm sure we can fix this.
> In this case we could even use the the component architecture itself
> to look up the grokkers for a class (would that even make sense?).
I think that's also a very interesting idea! You've obviously thought
about this more than I have.
Grokkers could be a form of adapters that adapt the classes they need to
register... I think this one needs experimentation to see whether it's
conceivable. If you're interested, you could make a branch of Martian to
try to implement this one. classProvides() for the base classes, and
then looking up grokkers by using adapters with the component architecture.
More information about the Grok-dev