[Grok-dev] Re: grok_component gone after actions refactoring?

[Philipp von Weitershausen]

Martijn Faassen wrote:
> Philipp von Weitershausen wrote:
>> Yeah, that makes sense. I actually changed my mind between then and 
>> now, and grok_component isn't that non-sensical anymore :). I brought 
>> it back, with the original tests passing just fine.
>>
>> I will explain my refactorings in an email to the list soon.
> 
> Thanks, that'll be interesting. Thanks for doing all this work, it looks 
> like you made a lot of progress starting off Godefroid's work.

Yup. I will discard my previous grokcore.* branch because this work has 
made it much easier to do the split-up. I also managed to keep the 
changes to Martian minimal and backwards-compatible.

More *after* I've given my talk today :).

>> I do think, however, that grok_component should move to grok.tests 
>> since it's really meant for testing. Especially the way it's currently 
>> implemented. Would that be ok with you?
> 
> Wouldn't grok.testing be a better place for this? Or did you mean this?

We don't have grok.testing yet, but it seems make sense to me. Also the 
grok.tests.grok() function could move there.

-- 
http://worldcookery.com -- Professional Zope documentation and training