Currently, tests for Python 3 fail: https://bugs.launchpad.net/heat-cfntools/+bug/1403214

Spec URL: https://raw.githubusercontent.com/hroncok/SPECS/master/python-mox3.spec
SRPM URL: https://churchyard.fedorapeople.org/SRPMS/python-mox3-0.7.0-1.fc21.src.rpm

Please reflow %description. Also please add some information about what the package does, except being a fork.

Consider changing nosetests to nostests-%{python2_version} for clarity.

Use %license for COPYING.txt

Consider adding the following snippet:
"""
# Use the same directory of the main package for subpackage licence and docs
%global _docdir_fmt %{name}
"""

Package doesn't build: the patch also changes .pyc file. Wouldn't it be better to simply skip the test using nosetest option instead of patching it out?

I think python2-devel is preferred to python-devel.

Otherwise looks OK, fedora-review has nothing interesting to say.

Thanks.

As it seems, I might not need this package after all (meybe). So I'm changing this to NotReady and will eventually either close it or continue by fixing it.

I won't need this. Sorry for wasting your time. In case you would ever need a package review, feel free to assign me directly and I'll do it.

Alan,

will you be resubmitting the review request and becoming the maintainer? I can do the review then.

I'm trying to clarify upstream status of mox3, seems to be neglect-ware:
http://lists.openstack.org/pipermail/openstack-dev/2015-March/060054.html

Any news here? python-mox3 is (becoming) a test dependency for OpenStack Horizon

Getting back to this, mox3 now has a proper upstream, it was adopted by Oslo project: https://review.openstack.org/190330

Alan, I'd propose to close this one and to re-submit as a new package. I'd be willing to do either review or propose this as a new package, since horizon requires it now for unit tests.

@mrunge why not just continue this review, afaik nothing prevents taking it over if initial submitter abandoned it?
I've uploaded initial SRPM from here + updates to https://github.com/apevec/python-mox3

Spec URL: https://raw.githubusercontent.com/apevec/python-mox3/d4490b80e635e99b44628e5434833b6eb5904639/python-mox3.spec
SRPM URL: https://apevec.fedorapeople.org/openstack/python-mox3-0.9.0-1.fc24.src.rpm
Description: Mock object framework for Python
Fedora Account System Username: apevec

Spec URL: https://raw.githubusercontent.com/apevec/python-mox3/da699f1eb745b1ce7a880ac35ca3e2cce90196ea/python-mox3.spec
SRPM URL: https://apevec.fedorapeople.org/openstack/python-mox3-0.9.0-1.fc24.src.rpm
Description: Mock object framework for Python
Fedora Account System Username: apevec

(In reply to Alan Pevec from comment #11)
> @mrunge why not just continue this review, afaik nothing prevents taking it
> over if initial submitter abandoned it?
Not 100% sure about this, but I think that the scripts check whether the bug-reporter name matches the submitter name on the csv request. So I'm afraid you'll have to open a new bug. Mark this one as duplicate then. You can assign the new bug to me.

> I've uploaded initial SRPM from here + updates to
> https://github.com/apevec/python-mox3
Most of the stuff from comment #c3 still applies.

Also:
- you removed the removal of egg-info in %prep. It is customary to do remove it in %prep to be sure that no stale info ends up in the binary package.

> scripts check whether the bug-reporter

So we're slaves to the scripts?!
No wonder we're turning away potential contributors with such artificial obstacles :(

> - you removed the removal of egg-info in %prep. It is customary to do remove
> it in %prep to be sure that no stale info ends up in the binary package.

That was a misguided custom, python guidelines has been clarified in https://fedorahosted.org/fpc/ticket/488 (relevant here is point 5.)

Egg metadata from upstream should be preserved, in some cases ( e.g. packages using PBR.) you can't reconstruct it completely when building from a sdist tarball.

(In reply to Alan Pevec from comment #14)
> > scripts check whether the bug-reporter
> 
> So we're slaves to the scripts?!
> No wonder we're turning away potential contributors with such artificial
> obstacles :(
I think everybody agrees that this is an unnecessary obstacle. At last Flock somebody (it might have been sgallagh, I don't remember for sure) said that they added that check and it could be removed. But it's there atm.

Like I said, I'll review the new ticket, so it's just a question of filling out https://bugzilla.redhat.com/enter_bug.cgi?product=Fedora&format=fedora-review and copying comment #12. I think that should hardly matter compared to the amount of work required for a package.

> > - you removed the removal of egg-info in %prep. It is customary to do remove
> > it in %prep to be sure that no stale info ends up in the binary package.
> 
> That was a misguided custom, python guidelines has been clarified in
> https://fedorahosted.org/fpc/ticket/488 (relevant here is point 5.)
> 
> Egg metadata from upstream should be preserved, in some cases ( e.g.
> packages using PBR.) you can't reconstruct it completely when building from
> a sdist tarball.
OK.

(In reply to Alan Pevec from comment #14)
> > scripts check whether the bug-reporter
> 
> So we're slaves to the scripts?!
> No wonder we're turning away potential contributors with such artificial
> obstacles :(
> 

As far as I know, the check is currently there, at least it has been.

You currently have TWO offers for review, and ONE offer to start a new review. IMHO this shouldn't be a real issue, right? ;-)

All of comment 3 feedback addressed in https://raw.githubusercontent.com/apevec/python-mox3/3f04cfc7e68fe9f77889252727675eaa6377bbc7/python-mox3.spec except:

> Package doesn't build: the patch also changes .pyc file. Wouldn't it be better to simply skip the test using nosetest option instead of patching it out?

Even after fixing the patch, python3 tests are failing but skipping tests for supposedly python3 fork doesn't feel right either.

Matthias, feel free to take over.

> python3 tests are failing but skipping tests
> for supposedly python3 fork doesn't feel right either.

BTW upstream did just that https://bugs.launchpad.net/python-mox3/+bug/1403214/comments/12 
This was merged for 0.8.0 release so it's unclear why 0.9.0 fails with python3-3.4.2-6.fc22

Ooops, this somehow slipped through. I will have a second look.

Alan, I agree with you, py3 tests failing for something trying to fix exact py3 issues doesn't look right.

Spec URL: https://social.fedorapeople.org/python-mox3.spec
SRPM URL: https://social.fedorapeople.org/python-mox3-0.9.0-3.fc24.src.rpm

tests won't pass without bug #1259286

See comments 13-16, if you want to take over, you'll need to open a new BZ b/c certain script checks BZ reporter...

Will open new bug for the review.