On 12/02/11 10:52, Shawn Walker wrote:
On 11/28/11 16:13, Brock Pytlik wrote:
On 10/03/11 13:54, Brock Pytlik wrote:
Webrev:
https://cr.opensolaris.org/action/browse/pkg/bpytlik/18975-v1
Bug:
18975 resolve should be able to restrict which packages are used as
fall back
I've also changed our gate to use the new feature which gives a nice
speed boost when doing make packages.
Thanks for taking a look,
Brock
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
There was discussion on this bug, but no review of the code. Any chance
someone can take a look?
src/man/pkgdepend.1:
line 179: s/be use/be used/
line 181: The wording here seems a bit awkward at the end, could you
try simplifying it a bit?
I've taken a shot at rewording it.
src/modules/publish/dependencies.py:
line 1319: missing system_patterns and use_system from docstring;
added
also curious as to whether system_patterns could imply use_system.
You gave me a good idea which was that use_system was totally redundant
given system_patterns, so I've removed use_system entirely.
src/pkgdep.py:
general: should we s/fmri/FMRI/ in error strings? It is an acronym.
There seem to be plenty of places in the code that use fmri and plenty
which use FMRI, so I'll just leave these as is.
Just for reference:
ggrep -RI 'FMRI' modules/ | grep '"' | wc -l
111
ggrep -RI 'fmris' modules/ | grep '"' | wc -l
100
lines 603-611: These don't match current gate tip, I'll assume this
just needs another resync.
Yep.
src/tests/cli/t_pkgdep_resolve.py:
lines 718, 724: s/self. output/self.output/
That's bizarre. I'm surprised the testing didn't blow up w/ things like
that.
src/tests/pkg5unittest.py:
The downside of this change is that it makes it harder to find
broken tests. I had originally purposefully not set these attributes
as it made it failure obvious when a test checked for output or errout
before running a command. Just an FYI; not an objection.
I'm trying to remember why I added this in. I think...
The problem I was running into was that when a command failed, I was
getting a second traceback because one of these wasn't defined, which
was hiding the real issue. That said, I can't figure out how to make
that happen given the existing code. I'll remove this change for now
since it's not needed for my changes, and file a separate bug if I come
across the problem again.
A new webrev will be out shortly, once I finish testing after syncing
with tip.
Brock
-Shawn
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
