Bug 277947 – Stabilize dev-python/pygobject-2.18.0

• Status: RESOLVED FIXED
• Product: Gentoo Linux

[Description Dustin Polke 2009-07-15 14:53:53 UTC]

Needed for python-2.6 stabilization as pygobject-2.16.1 fails tests (see bug 266639)

Reproducible: Always

[Comment 1 Jeremy Olexa (darkside) (RETIRED) 2009-07-22 04:24:16 UTC]

Hello gnome team,
As you know this is a blocker for python-2.6 stabilization. amd64 is ready for that now, this being the last blocker. Mart in irc stated that "the gnome list" should be ready for arches during this week, if that date is slipping please consider adding arches for this bug alone if there are no compatibility issues. Thank you.

[Comment 2 Mart Raudsepp 2009-07-24 09:09:29 UTC]

Ok, I did various QA fixes to the ebuild and it should be ready for stabilization now.
Arches, please proceed

[Comment 3 Samuli Suominen (RETIRED) 2009-07-24 14:41:53 UTC]

Please "cvs up" before; I've fixed some last minute issues.

[Comment 4 Jeremy Olexa (darkside) (RETIRED) 2009-07-25 03:32:44 UTC]

Created attachment 199055 [details]
build.log

On amd64. I could use some advice on if this is a problem or not?

======================================================================
ERROR: testQueryWritableNamespaces (test_gio.TestFile)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "../tests/test_gio.py", line 389, in testQueryWritableNamespaces
    for info in infolist:
TypeError: 'NoneType' object is not iterable

----------------------------------------------------------------------

The ebuild *does not* die from this. With python-2.6.2-r1 in otherwise stable system.

[ebuild   R   ] dev-python/pygobject-2.18.0  USE="X debug doc examples libffi" 0 kB

[Comment 5 Arfrever Frehtes Taifersar Arahesis (RETIRED) 2009-07-27 20:42:27 UTC]

(In reply to comment #4)

Tests should be fixed now.

[Comment 6 Markus Meier 2009-07-30 21:19:55 UTC]

amd64/x86 stable

[Comment 7 nixnut (RETIRED) 2009-08-09 13:07:03 UTC]

ppc stable

[Comment 8 Sebastian 2009-08-16 13:30:34 UTC]

Hi all,

I *think* the eclass "alternatives" needs to be inherited. Otherwise the ebuild calls "alternatives_auto_makesym" but it's unknown. 

After upgrading to python 2.6 I reinstalled pygobject but afterward alacarte wouldn't compile. After adding the eclass mentioned above to the pygobject ebuild and once again installing pygobject alacarte was installed without any complaints.

[Comment 9 Mart Raudsepp 2009-08-17 03:02:43 UTC]

(In reply to comment #8)
> Hi all,
> 
> I *think* the eclass "alternatives" needs to be inherited. Otherwise the ebuild
> calls "alternatives_auto_makesym" but it's unknown. 

While it would probably be more correct to explicitly inherit alternatives, it is currently being inherited by python.eclass too, which the ebuild does inherit, so the alternatives_auto_makesym functions from alternatives.eclass should be available and known to my knowledge of inherit behaviour...

[Comment 10 Mart Raudsepp 2009-08-17 03:14:09 UTC]

Oh, nevermind. The python folks meanwhile broke it by removing inherit alternatives from python.eclass and this was handled in bug 281714

[Comment 11 Sebastian 2009-08-17 15:51:24 UTC]

Thanks for clearing things up!

[Comment 12 Jeroen Roovers (RETIRED) 2009-08-19 16:31:58 UTC]

Stable for HPPA.

[Comment 13 Tobias Klausmann (RETIRED) 2009-08-26 12:45:02 UTC]

Stable on alpha.

[Comment 14 Raúl Porcel (RETIRED) 2009-08-27 20:05:58 UTC]

arm/ia64/s390/sh/sparc stable

[Comment 15 Tom Gall (RETIRED) 2009-09-22 04:40:14 UTC]

stable on ppc64

[Comment 16 Gilles Dartiguelongue (RETIRED) 2009-09-22 06:22:10 UTC]

damn, arches forgot to close this again...