Excerpts from Jason L Tibbitts III's message of Mon Jul 02 07:38:40 +1000 2012:
>>>>> "DC" == Dan Callaghan
DC> I will take all three (they look straightforward :-) in exchange for
Since you appear to be familiar with sugar, is there any possibility
that you (or anyone else who is familiar with sugar stuff) could cast a
glance at https://bugzilla.redhat.com/show_bug.cgi?id=708663
poor guy has been waiting for nearly 13 months for someone to look at
his package. I can do the bulk of the review and the sponsorship if no
sponsor/sugar expert is available but I don't know how to actually do
any testing of sugar-related things.
I am by no means a Sugar expert. I didn't know anything about it until
I took these three reviews last night. :-) I am working from the Sugar
activity guidelines here:
and I installed F17 in a VM and did
yum groupinstall Sugar\ Desktop\ Environment
for testing out the packages. It also helps that the activities are in
Python which I am well versed in.
As it happens I was planning to try and track down somebody responsible
for those guidelines, because there are a few things which do not seem
right to me. I've found the OLPC list (cc'ed) which seems to be the
closest we have to a Sugar SIG. Hopefully somebody there can help out.
* The sample spec in the guidelines uses Group: Sugar/Activities, but
it's not mentioned anywhere else that I can find, and rpmlint
complains because it is non-standard. Are we supposed to just waive
that warning in the package review?
* The upstream-supplied setup.py for these three Sugar activities just
calls sugar.activity.bundlebuilder.start(), which seems to copy the
entire directory contents wholesale into /usr/share/sugar/activities
-- including stuff like setup.py, NEWS, COPYING... There are
apparently a lot of existing packages that do it too:
$ repoquery -f /usr/share/sugar/activities/\*/setup.py | wc -l
Is it really necessary to include these files in the activity
directory? COPYING, NEWS, etc are already installed by %doc, as they
should be, so it seems superfluous to have them installed again. And
setup.py shouldn't be needed for an installed package. Should this
sugar.activity.bundlebuilder be patched to be more selective in what
it copies? Or should each activity package be responsible for cleaning
out its buildroot after the bundlebuilder has run? Or is there some
reason why these files should be installed in the activity directory?
Dan Callaghan <dcallagh(a)redhat.com>