[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#801441: jessie-pu: package bcfg2/1.3.5-1



On Mon, 2015-10-12 at 15:38 +0300, Arto Jantunen wrote:
> "Adam D. Barratt" <adam@adam-barratt.org.uk> writes:
> 
> > On 2015-10-12 13:13, Arto Jantunen wrote:
> >> "Adam D. Barratt" <adam@adam-barratt.org.uk> writes:
> >>
> >>> Control: tags -1 + moreinfo
> >>>
> >>> On Sat, 2015-10-10 at 09:43 +0300, Arto Jantunen wrote:
> >>>> I would like to update bcfg2 in stable to match the version currently in
> >>>> testing to enable it to work with Django 1.7 (bug #755645). To do this I
> >>>> would
> >>>> add the attached patch, which looks much worse than it is due to the db
> >>>> migration files being moved around.
> >>>
> >>> That is rather noisy, yes. :-(
> >>>
> >>> Is there any chance we could have an interdiff of the before-and-after
> >>> patches, to highlight the actual differences between them? In any case
> >>> we'd need a full debdiff of a package built and tested on jessie before
> >>> giving a definite ack.
> >>
> >> The patch doesn't exist in jessie, so either interdiff isn't possible or
> >> I'm misunderstanding something.
> >
> > The latter, although possibly because I wasn't clear enough.
> >
> > The patch contains, for example:
> >
> >  src/lib/Bcfg2/Reporting/migrations/0001_initial.py | 1006
> > +++++++++++---------
> >  .../migrations/0002_convert_perms_to_mode.py       |  171 ----
> >  .../Reporting/south_migrations/0001_initial.py     |  465 +++++++++
> >  .../south_migrations/0002_convert_perms_to_mode.py |  171 ++++
> >
> > What I was looking for was an indication of what the actual difference between
> > the two sets of files is, ignoring renames (e.g. are the two
> > 0002_convert_perms_to_mode.py files actually exactly the same?).
> >
> >> The debdiff is attached (for convenience
> >> built on jessie but without modifying the version number for a stable
> >> update).
> >
> > Sorry, I meant a _source_ debdiff (in practice, that's basically always what's
> > needed).
> >
> > On the subject of the binary debdiff however...
> >
> > Files in first set of .debs but not in second
> > ---------------------------------------------
> > -rw-r--r--  root/root
> > /usr/lib/python2.7/dist-packages/Bcfg2/Reporting/migrations/0002_convert_perms_to_mode.py
> > -rw-r--r--  root/root
> > /usr/lib/python2.7/dist-packages/Bcfg2/Reporting/migrations/0003_expand_hash_key.py
> > -rw-r--r--  root/root
> > /usr/lib/python2.7/dist-packages/Bcfg2/Reporting/migrations/0004_profile_can_be_null.py
> > -rw-r--r--  root/root
> > /usr/lib/python2.7/dist-packages/Bcfg2/Reporting/migrations/0005_add_selinux_entry_support.py
> > -rw-r--r--  root/root
> > /usr/lib/python2.7/dist-packages/Bcfg2/Reporting/migrations/0006_add_user_group_entry_support.py
> >
> > Shouldn't there be a corresponding set of files appearing in the second
> > package? (in /Reporting/south_migrations)
> 
> I don't know the answer to that question, my understanding of Django is
> rather limited (which is also why I didn't write the patch to do
> this). The "initial migration" file grows quite a bit, so it's entirely
> possible that it ends up containing the relevant parts of those, but
> this is only a theory.

My understanding of Django is likely less than yours, it just seems odd
to have the patch create the new files and then for them not to be
shipped.

> Source debdiff is attached,

How was that generated? It appears to be missing at least debian/changelog.

Regards,

Adam


Reply to: