+1 here
Once you start working with svn you will quickly realize how annoying
cvs can be, Jens ;-)
Stefan
On 3. Mai 2005, at 19:20, Jens Vagelpohl wrote:
On May 3, 2005, at 4:28 , Florent Guillaume wrote:
How about migrating CMF to subversion too? I'd say, 1.5 and HEAD.
To be honest, I don't see
I get a whole bunch of Unauthorized errors when running CMF-1_5-
branch tests with Zope 2.8. Interestingly, CMF HEAD does not appear
to suffer from the same...
Stefan
==
ERROR: test_CopyPasteSetsOwnership
Alright, this particular instance contained VerboseSecurity, which
doesn't seem to work with 2.8. Bummer.
Stefan
On 13. Jun 2005, at 13:08, yuppie wrote:
FYI: I can't reproduce this error. All tests pass for me using
latest CMF-1_5-branch and Zope 2.8.0 final.
--
Software Engineering is
Dylan,
This is how Zope works. You get one copy of every object per thread,
and - for skin objects - per CMF/Plone instance. This has always been
the case, so please excuse the lack of enthusiasm :-).
Additionally, nobody ever claimed that you could run Zope/Plone off
your average
Thank you Jens, I share your sentiments. ;-)
Stefan
On 28. Aug 2005, at 10:45, Jens Vagelpohl wrote:
So hopefully this is the last pre-Plone 2.1 release.
--
Anything that happens, happens. --Douglas Adams
___
Zope-CMF maillist -
I get a traceback now from Guard.py
exceptions.NameError: global name 'r' is not defined
[snip]
File /autotest/temp/python21-zope26-cmf14-plone20/Products/
CMFCore/WorkflowTool.py, line 119, in manage_addWorkflow
ob = factory(id)
File
Well, you are supposed to use a subclass anyway, and in that subclass
you can fix everything, even _refreshSkinData. Maybe I should make
PortalTestCase truly abstract to emphasize the point.
NotImplementedError anyone? :-)
Stefan
On 9. Sep 2005, at 11:54, yuppie wrote:
That's great,
I see two options:
a) Use a different portal name
portal_name = 'mysite'
CMFTestCase.setupCMFSite(portal_name)
b) Use getPortal() to create a new portal per test
def getPortal(self):
manage_addCMFSite(self.app, portal_name)
return getattr(self.app, portal_name)
On 5. Okt 2005, at 23:27, Paul Winkler wrote:
I'm trying to use CMFTestCase along with ZopeTestCase.Functional
to test my skin scripts. But how do I get them set up
properly? Anybody have working examples of this? I wondered
if there might be some in the Plone test suites but if so I
didn't
Exactly!
ZopeTestCase (CMFTestCase, PloneTestCase) was designed for testing
*applications*. As you say, when I write a CMF application I should
be able to assume the CMF below works. In this case a fat fixture
makes perfect sense (to me). Also, I firmly believe application
developers
Right, so at what point is the content-type-registry supposed to be
added now? Plone Site creation breaks after this checkin with:
File /Users/zope/plonehd/Products/CMFCore/utils.py, line 83, in
getToolByName
raise AttributeError, name
AttributeError: content_type_registry
Thanks,
The use of sorted makes GS only work with Python 2.4. Intentional?
http://mail.zope.org/pipermail/zope-cvs/2006-February/006797.html
Stefan
--
Anything that happens, happens. --Douglas Adams
___
Zope-CMF maillist - Zope-CMF@lists.zope.org
This turned out to be a bug in FSPythonScripts (no __file__ in script
globals). Fixed on all branches of CMF = 1.5.
Cheers,
Stefan
On 13. Feb 2006, at 16:48, Julien Anguenot wrote:
Florent Guillaume wrote:
Julien Anguenot wrote:
Tim Peters wrote:
[Julien Anguenot]
I'm having some
I am using it rather heavily. The overengineery impression comes
from the fact that it is intended to be highly pluggable (CMF-style,
i.e. by replacing tools). For example I always replace the generator
with something more robust than a simple counter. Using it is
straight forward, you
Talking about abuse...
For the UnicodeLexicon product I have created a setuphandler that
updates ZCTextIndexes in the catalog. This means it basically
contains the code that would formerly have lived in
Extensions.Install.install.
Would this be considered abuse of GenericSetup?
On 16. Aug 2006, at 18:18, yuppie wrote:
AFAICS the catalog.xml in the profile is a normal catalog setup
file. The default import step for this file should be registered
and updates the catalog if all steps are run.
Yes, the default machinery creates the UnicodeLexicons from
On 17. Aug 2006, at 15:50, yuppie wrote:
My approach would be to improve the catalog adapters. Even if you
stick to the behavior implemented in setuphandlers.py it would have
been easier and more straight forward to implement it in a subclass
of ZCatalogXMLAdapter, overriding the default
In the light of Z3 container events existing since Zope 2.9, should
CMFCatalogAware still __recurse? I don't think so as ATM this appears
to result in multiple calls to indexObject/unindexObject for
sublocations.
Stefan
--
It doesn't necessarily do it in chronological order, though.
You don't want to cache views for Files and Images. I use this for
predicate:
python: not view and object.portal_type=='File'
Stefan
On 13. Sep 2006, at 01:30, Dmitry S. Makovey wrote:
Playing with CPM I've discovered that if I put a policy like:
id: Files
Predicate:
Tres, I am afraid you seem to be the only one able to verify this.
IIRC the security fix is yours as - of course - is CMF. ;-)
I am ready to remove irrelevant combinations, but in this case I have
been reluctant up to now because of the possible security implications.
Stefan
On 24. Okt
I vote for (at least optional) backward compatibility. See my
followup here
http://www.zope.org/Collectors/Zope/2191.
Stefan
On 22. Nov 2006, at 17:40, Dorneles Treméa wrote:
I just made a comment on http://www.zope.org/Collectors/Zope/2191.
As soon as we get consensus I can proceed with
We use CMF 1.6 branch with GS 1.2 branch in Plone 2.5. This, and
itchy fingers, made me update the external.
FWIW, when I check out CMF/tags/1.6.2 I expect to get a GS tag as
well. When I check out CMF/branches/1.6 I expect to get the HEADs of
the respective development branches.
Another
CMF 1.6 is supposed to work with Zope 2.8. However, either there is
no queryDefaultViewName or it lives someplace else...
from zope.app.publisher.browser import queryDefaultViewName
ImportError: cannot import name queryDefaultViewName
Stefan
Begin forwarded message:
From: Philipp von
FWIW, this and related test failures are due to
xml.sax.saxutils.quoteattr changing behavior in Python 2.4.4.
Stefan
On 20. Dez 2006, at 10:48, Yvo Schubbe wrote:
- disabled hidden() test (no time to track down why this fails in
the nightly tests)
--
It doesn't necessarily do it in
Much fun to be had with daylight-saving, eh?
FWIW, I have double-checked the system time, timezone, and locale
settings on the machine running the nightlies, and everything appears
to be sane. I see the same failure on my Mac, so /me shrugs...
Stefan
On 28. Mär 2007, at 14:00, CMF Tests
Nope, clearFindAndRebuild uses indexObject, for the very reason not
to touch the modification time. And so should you.
Stefan
On 28. Mai 2007, at 06:57, Darryl Dixon - Winterhouse Consulting wrote:
Why is
this? It has implications for, eg, the new clearFindAndRebuild()
function
in
No, it adds more weight to my recommendation to use indexObject :-)
Stefan
On 28. Mai 2007, at 12:30, Darryl Dixon - Winterhouse Consulting wrote:
time...? Which lends even more weight to my original query - why on
earth
does reindexObject() update the modification time? I can see
+1
On 25. Okt 2007, at 12:37, Jens Vagelpohl wrote:
Can we have a show of hands, so to speak?
--
Anything that, in happening, causes something else to happen,
causes something else to happen. --Douglas Adams
___
Zope-CMF maillist -
You may want to try CMF.buildout:
svn://svn.zope.org/repos/main/CMF.buildout
Stefan
On 24.04.2009, at 11:54, Michael Haubenwallner wrote:
Just easy_installed the new Zope2.12.a4 release and Products.CMFCore
(2.1.2)
___
Zope-CMF maillist -
29 matches
Mail list logo