[Freeipa-users] Building previous release rpms are failing

[Lukas Slebodnik]

On (07/08/14 07:39), Curtis L. Knight wrote:
>On Tue, Aug 5, 2014 at 11:26 PM, Rob Crittenden <rcritten at redhat.com> wrote:
>
>> Curtis L. Knight wrote:
>> > On Tue, Aug 5, 2014 at 7:21 AM, Martin Kosek <mkosek at redhat.com
>> > <mailto:mkosek at redhat.com>> wrote:
>> >
>> >     On 08/05/2014 12:32 PM, Martin Kosek wrote:
>> >     > On 08/05/2014 12:05 PM, Curtis L. Knight wrote:
>> >     ...
>> >     >> #./make-lint $(LINT_OPTIONS)
>> >     >>
>> >     >> run 'make rpms' again to get beyond lint errors shown below
>> >     >>
>> >     >>
>> >     >> cd install; if [ ! -e Makefile ]; then ../autogen.sh --prefix=/usr
>> >     >> --sysconfdir=/etc --localstatedir=/var --libdir=/usr/lib; fi
>> >     >> ./make-lint
>> >     >> Traceback (most recent call last):
>> >     >>   File "./make-lint", line 272, in <module>
>> >     >>     sys.exit(main())
>> >     >>   File "./make-lint", line 243, in main
>> >     >>     linter.check(files)
>> >     >>   File "/usr/lib/python2.7/site-packages/pylint/lint.py", line
>> >     626, in check
>> >     >>     self.check_astroid_module(astroid, walker, rawcheckers,
>> >     tokencheckers)
>> >     >>   File "/usr/lib/python2.7/site-packages/pylint/lint.py", line
>> >     712, in
>> >     >> check_astroid_module
>> >     >>     walker.walk(astroid)
>> >     >>   File "/usr/lib/python2.7/site-packages/pylint/utils.py", line
>> >     715, in walk
>> >     >>     self.walk(child)
>> >     >>   File "/usr/lib/python2.7/site-packages/pylint/utils.py", line
>> >     715, in walk
>> >     >>     self.walk(child)
>> >     >>   File "/usr/lib/python2.7/site-packages/pylint/utils.py", line
>> >     712, in walk
>> >     >>     cb(astroid)
>> >     >>   File
>> >     "/usr/lib/python2.7/site-packages/pylint/checkers/newstyle.py",
>> >     >> line 135, in visit_function
>> >     >>     args=(call.args[0].name, ))
>> >     >> AttributeError: 'Getattr' object has no attribute 'name'
>> >     >> make: *** [lint] Error 1
>> >     >
>> >     > This is new, I created upstream ticket to timely fix it:
>> >     > https://fedorahosted.org/freeipa/ticket/4475
>> >
>> >     Ticket 4475 is now fixed, thanks to Jan Cholasta. ipa-3-3 branch
>> >     should now
>> >     build OK again.
>> >
>> >     Martin
>> >
>> >
>> > Hey Martin,
>> >
>> > Tested ipa-3-3 and generated rpms from that branch. Many thanks for the
>> > resolution.
>> >
>> > Just a note, but I verified that ipa-3-2 and ipa-3-1 are in need of the
>> > same ipa-3-3 dependency patch. Both also complained that make-lint
>> > needed pylint installed which it already was. With the lint failure and
>> > rhino patch, ipa-3-2 did generate rpms. With the lint failure and rhino
>> > patch, ipa-3-1 did not generate rpms and gave the following logs.
>>
>> I guess it becomes a bit fuzzy, especially with these versions. We don't
>> usually offer any guarantees that older releases will build against more
>> modern distros, but both 3.1.5 and 3.2.0 crossed that line, with Fedora
>> builds in two releases (F18/19 and F19/20 respectively).
>>
>> Do you have a requirement to use these older releases or are you just
>> offering this data point in case anyone else runs into this?
>>
>> regards
>>
>> rob
>>
>
>Hello Rob,
>
>Yes this is additional information and is not any requirement for me. I was
>not sure which branches were being maintained for F20. My interest was to
>see if I could help the freeipa developers build rpms easily from git with
>Docker images/containers. That is just about finished. My next thought was
>about using a Docker containers to test code from a git working directory
>quickly. That workflow could be a) to build rpms from a git commit, install
>the generated rpms or b) push changed code into an existing freeipa
>installation (probably not recommended but maybe necessary for testing). I
>did read a couple of places that it seems to take less time and or RAM to
>build code within Docker then other methods. Overall there does not seem to
What kind of other methods did you try?
I doubt it is more efficient then mock [1]

>be enough people that are doing it yet for a lot of data points. Does any
>of that sound beneficial to the team?
>

LS

[1] http://fedoraproject.org/wiki/Projects/Mock