Rich Burridge wrote:
Now that the pkg gate is open again, I'd like to try to check
in some of the potential fixes I've got for various bugs. I'm
looking for code reviewers for some of them.
Here's a summary:
...
4134 pkg operations should report operation sizes
Bugid: http://defect.opensolaris.org/bz/show_bug.cgi?id=4134
7736 image plan should provide space requirements for operation
Bugid: http://defect.opensolaris.org/bz/show_bug.cgi?id=7736
Last message:
http://mail.opensolaris.org/pipermail/pkg-discuss/2009-May/013742.html
Last webrev: http://cr.opensolaris.org/~richb/pkg-4134-v3/
http://cr.opensolaris.org/~richb/pkg-4134-v3/src/modules/client/imageplan.py.wdiff.html
lines 201-202: replace 'size' with 'dsize'; my fault
line 225: Need _() around "Estimated transfer size"
line 227-228: drop
line 230: s/operation_size/self.get_opstats()[3]/
line 229: Need _() around "Estimated space required"
----
4178 unfound_fmris is a poor name
Bugid: http://defect.opensolaris.org/bz/show_bug.cgi?id=4178
Last message:
http://mail.opensolaris.org/pipermail/pkg-discuss/2009-May/013539.html
Last webrev: http://cr.opensolaris.org/~richb/pkg-4178-v2/
http://cr.opensolaris.org/~richb/pkg-4178-v2/src/modules/client/image.py.html:
line 3030: s/unfound/unmatched/
----
4395 Use pkg.summary for "pkg info" output
Bugid: http://defect.opensolaris.org/bz/show_bug.cgi?id=4395
8829 solaris.py should be able to add pkg.summary attributes
Bugid: http://defect.opensolaris.org/bz/show_bug.cgi?id=8829
Last message:
http://mail.opensolaris.org/pipermail/pkg-discuss/2009-May/013515.html
Last webrev: http://cr.opensolaris.org/~richb/pkg-4395-8829-v2/
http://cr.opensolaris.org/~richb/pkg-4395-8829-v2/src/modules/client/api.py.wdiff.html
line 794: break after this so you don't continue looping through set
actions needlessly?
----
5707 depot should support SIGHUP to refresh configuration
Bugid: http://defect.opensolaris.org/bz/show_bug.cgi?id=5707
Last message:
http://mail.opensolaris.org/pipermail/pkg-discuss/2009-May/013432.html
Last webrev: http://cr.opensolaris.org/~richb/pkg-5707-v2/
http://cr.opensolaris.org/~richb/pkg-5707-v2/src/modules/server/depot.py.wdiff.html
After further research, I'm wondering if lines 157-159 are necessary.
If you remove those, and you send a SIGHUP to a depot process (that is
*not* running from a terminal/console) does the configuration get
refreshed properly?
It should, because cherrypy performs an execve of the process by default
when a SIGHUP is received.
I think we only need the SIGUSR1 hook. Can you check?
----
9002 check_classifications.txt should be adjusted for new post
2008.11 classifications.
Bugid: http://defect.opensolaris.org/bz/show_bug.cgi?id=9002
Last message:
http://mail.opensolaris.org/pipermail/pkg-discuss/2009-May/013728.html
Last webrev: http://cr.opensolaris.org/~richb/pkg-9002-v2/
This one needs Dan or Stephen (or whoever) to tell me where the
static file "located somewhere centrally (like a DTD would be)
on opensolaris.org (similar to how DTDs work) or on opensolaris.com
or on sun.com." should be.
When I have that, I can adjust the webrev accordingly and ask for code
review.
Speaking of that ...
Danek and I discussed last week if there was a better option here than
having a centrally located classifications file. After discussion, it
became clear that we didn't really need one at all.
Instead, the client API will be changed to read the classification
information for all packages each time we cache the catalog data and
then cache the resulting package mapping and set of classifications.
In turn, the API will provide a method to get this set of
classifications, so no file will be needed.
Your changes to check_classifications seem fine though, I would just
drop the _resources/ directory and contents from your webrev.
Cheers,
--
Shawn Walker
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
