• Remember Mailing List

Re: reverted modification to membrane tool

from "Andrew Burkhalter" on Oct 31, 2006 01:01 PM
On 10/30/06, Brian Gershon <briang@...> wrote:
>
> Hey Andrew,
>
> Yes, in Plone 3.0 "named adapters" were breaking the tests.
>
> (My understanding is that named adapters are adapters that require
> parameters, and the tests weren't passing any parameters so they were giving
> some sort of unsupported Interface error)
>
> Maybe (for 2.5) this code needs to be wrapped with another "if" (was it
> giving an attribute error or something based on .name?) so that it doesn't
> break in 2.5.  I wasn't able to unit test in 2.5 since the tests were
> broken - giving us an infinite loop of unit testing fun :)


Well, if you're feeling ambitious and would like to put this code back into
membrane, you can also test in the Plone 2.5 branch.  See the exact testing
platform in the thread below.  As of this point, you should only see to 2
assertion errors showing that I believe copy and cut don't work and one
failure associated with the deleteMembers test.  The latter I'm working on
and hope to get Rob's eyes on things.

How about we make a note about this in the ReMember Issue Tracker so we can
> continue to play with this down the road after the tests are working - then
> we can make sure it works on both 2.5 and 3.0.
>
> I'll go ahead now and create an entry in the Tracker for this...


Fantastic.  Thanks Brian.

Cheers, Andrew


> On 10/30/06, Andrew Burkhalter < andrewburkhalter@...> wrote:
> >
> > First off, my platform:
> >
> > - zope 2.9.5
> > - Plone 2.5 development branch
> > - remember trunk
> > - membrane trunk
> >
> > Upon seeing the remember tests completely crash and burn at the
> > RememberProfileLayer as well as basic add new user support TTW broken, I've
> > backed out the following code from membrane/tools/membrane.py, which were
> > added in attempt to support the Plone 3.0 developer branch:
> >
> >             # avoid checking 'named' adapters
> >             if adapter_reg.name:
> >                 continue
> >
> > Once completed, we're back to the two assertion errors and one failure.
> >
> >
> > I'm not sure what's going on here, so hopefully briang has joined this
> > list and can shed some light on the subject.  I've kept the code intact
> > within a comment and added a bit of my own thoughts, but for now thought it
> > was a priority to get this working on our primary platform at this point and
> > also to save others a lot of head scratching.  Sorry if I'm missing
> > something obvious or not understanding something.
> >
> > Please post to the list if you have any thoughts, questions about this.
> >
> > Cheers, Andrew
> >
> >
>

Return to date view: threaded or flat