Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-24 Thread Lennart Regebro
On Sat, Apr 24, 2010 at 00:14, Tres Seaver tsea...@palladion.com wrote: Maybe the short-term fix would be to add a shim package (reviving 'zope.app.dublincore', maybe) whose only job in life is to make those permissions loadable.  Maybe the package ships with some code to help write the evolve

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-24 Thread Marius Gedminas
On Fri, Apr 23, 2010 at 06:14:23PM -0400, Tres Seaver wrote: Jim Fulton wrote: On Fri, Apr 23, 2010 at 4:18 AM, Jacob Holm j...@improva.dk wrote: Hi Michael, Tres I agree a new major version is required due to the new feature of having new permission names, but there is no reason to

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Michael Howitz
Am 19.04.2010 um 22:21 schrieb Tres Seaver: Log message for revision 45: Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`. Applications may need to fix up grants based on the old permissions. fixes lp:566724 Changed: U

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Jacob Holm
Hi Michael, Tres I agree a new major version is required due to the new feature of having new permission names, but there is no reason to break compatibility with code using the old names. IIRC the redefinePermission/ zcml-directive is there exactly to provide backwards compatibility when

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Tres Seaver
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Jacob Holm wrote: Hi Michael, Tres I agree a new major version is required due to the new feature of having new permission names, but there is no reason to break compatibility with code using the old names. IIRC the redefinePermission/

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Jacob Holm
Hi Tres Tres Seaver wrote: Assuming we put the 'redefinePermssion' directives in place on the trunk, why shouldn't we leave the version number as is? I consider the rename a bugfix, not a feature, and if we make it backwared compatible, there is no reason to bump the major version. It's a

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Michael Howitz
Am 23.04.2010 um 15:19 schrieb Tres Seaver: [...] Thanks for pointing out that directive, whose existence I had forgotten. Assuming we put the 'redefinePermssion' directives in place on the trunk, why shouldn't we leave the version number as is? I consider the rename a bugfix, not a feature,

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Fred Drake
On Fri, Apr 23, 2010 at 3:08 PM, Tres Seaver tsea...@palladion.com wrote: Done.  The 3.6.3 release is up on PyPI: I guess I should get out more; just noticed this thread. Changing the stored permission names is a big deal. If a permission gets stored with zope.dublincore 3.6.3, then an

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Tres Seaver
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Fred Drake wrote: On Fri, Apr 23, 2010 at 3:08 PM, Tres Seaver tsea...@palladion.com wrote: Done. The 3.6.3 release is up on PyPI: I guess I should get out more; just noticed this thread. Changing the stored permission names is a big deal.

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Jim Fulton
On Fri, Apr 23, 2010 at 4:18 AM, Jacob Holm j...@improva.dk wrote: Hi Michael, Tres I agree a new major version is required due to the new feature of having new permission names, but there is no reason to break compatibility with code using the old names. IIRC the redefinePermission/

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Jim Fulton
On Fri, Apr 23, 2010 at 5:58 PM, Tres Seaver tsea...@palladion.com wrote: ... I disagree, pretty strongly:  making code forever responsible for bad old data is responsible for a lot of horrors in both Zope2 and Zope3 code bases.  Releasing the packages separately allows the folks who need time

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Tres Seaver
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Jim Fulton wrote: On Fri, Apr 23, 2010 at 4:18 AM, Jacob Holm j...@improva.dk wrote: Hi Michael, Tres I agree a new major version is required due to the new feature of having new permission names, but there is no reason to break compatibility

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Tres Seaver
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Jim Fulton wrote: On Fri, Apr 23, 2010 at 5:58 PM, Tres Seaver tsea...@palladion.com wrote: ... I disagree, pretty strongly: making code forever responsible for bad old data is responsible for a lot of horrors in both Zope2 and Zope3 code bases.

Re: [Zope-dev] [Checkins] SVN: zope.dublincore/trunk/ Renamed the ``zope.app.dublincore.*`` permissions to ``zope.dublincore.*`.

2010-04-23 Thread Tres Seaver
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Tres Seaver wrote: Why should zope.dublincore be carrying BBB-water for zope.app.dublincore? We have plenty of precedent for having the zope.app. version of the package stick around *purely* for BBB: let's do that in this case. Even better: