[Zope3-dev] Re: zodbcode vs zope.app.module and zope.app.interface.PersistentInterface

[Ross Patterson]

Ross Patterson <me at rpatterson.net> writes:

> Ross Patterson <me at rpatterson.net> writes:
>
>> zope.app.interface.PersistentInterface gets used properly for
>> persistent modules containing classes that subclass
>> zope.interface.Interface when those modules are are managed and
>> registered through managers and registries from zodbcode.  However,
>> similar persistent modules are broken when managed and registered
>> through managers and registries from zope.app.module.  Attached is a
>> diff against zope.app.module that adds a test demonstrating the
>> failure.
>>
>> I'll create an issue in the collector if I can't fix this, but I hope
>> to fix it.
>>
>> What I'd like to know is why zope.app.module seems to have totally
>> independent manager and registry implementaitons from those found in
>> zodbcode.  It looks like Fred Drake did a lot of the work on zodbcode
>> and Setphan Richter did a lot of the work on zope.app.module, so I was
>> hoping to learn something about the rationale from one or both of
>> you.  Is there any reason not to realign the implmentations in
>> zope.app.module with those in zodbcode?
>>
>> I realize that most of this work was done long ago, but I appreciate
>> any insight.
>>
>> Ross
>
> Attached is a new verion of the patch against zope.app.module that I
> think is a bit clearer.  It demonstrates that the manager from
> zodbcode works where the manager from zope.app.module doesn't.
>
> Ross

I figured this out and demonstrated the workaround in a doc file and
test:

http://svn.zope.org/Zope3/trunk/src/zope/app/module/?rev=72384&view=rev

Ross