Bug 433461 – Please stabilize =dev-python/bpython-0.11

• Status: RESOLVED FIXED
• Product: Gentoo Linux

[Description Mike Nerone 2012-08-30 21:09:28 UTC]

Requesting stabilization of dev-python/bpython-0.11. In tree for nearly five months with no bugs reported.

Thanks!

[Comment 1 Andrey Grozin 2012-09-21 14:35:40 UTC]

Please stabilize.

[Comment 2 Agostino Sarubbo 2012-09-23 12:30:31 UTC]

amd64 stable

[Comment 3 mike 2012-09-25 00:01:13 UTC]

Created attachment 324894 [details]
zope-interface build log

x86 not ok (repoman ok, compile fails because of dependency: net-zope/zope-interface-3.8.0)

[Comment 4 mike 2012-09-25 00:08:18 UTC]

(In reply to comment #3)
> Created attachment 324894 [details]
> zope-interface build log
> 
> x86 not ok (repoman ok, compile fails because of dependency:
> net-zope/zope-interface-3.8.0)

See Bug 420273 for reference.

[Comment 5 Andrey Grozin 2012-09-26 03:18:59 UTC]

(In reply to comment #3)
> Created attachment 324894 [details]
> zope-interface build log
> 
> x86 not ok (repoman ok, compile fails because of dependency:
> net-zope/zope-interface-3.8.0)
I don't understand. zope-interface is not a dependency of bpython (or its dependencies). I have no zope-interface installed, and bpython builds and works fine.

[Comment 6 Mike Nerone 2012-09-26 09:25:48 UTC]

(In reply to comment #5)
> (In reply to comment #3)
> I don't understand. zope-interface is not a dependency of bpython (or its
> dependencies). I have no zope-interface installed, and bpython builds and
> works fine.

Indeed. Not that a test failure in a different, stable (!) package would constitute a regression in bpython.

http://blogs.gentoo.org/ago/2012/08/22/when-you-should-block-a-stabilization/

[Comment 7 mike 2012-09-26 22:15:50 UTC]

Well yes, seems like this problem was overlooked in previous versions. The dependency to zope-interface originates from FEATURES="test" which means it wasn't compiled with this feature when testing it (which should be done regarding to http://www.gentoo.org/proj/en/base/x86/arch-testers-faq.xml#steptest). This is no regression since the test-bug was present all the time. The comment about the failed compilation on 2012-09-25 00:01:13 UTC was just made for completeness, dev's have to decide if this build goes stable.

[Comment 8 Arfrever Frehtes Taifersar Arahesis 2012-09-26 23:35:19 UTC]

(In reply to comment #7)
> The dependency to zope-interface originates from FEATURES="test"

dev-python/bpython has no dependency on net-zope/zope-interface, even with FEATURES="test".

[Comment 9 mike 2012-09-27 00:22:16 UTC]

(In reply to comment #8)
> (In reply to comment #7)
> > The dependency to zope-interface originates from FEATURES="test"
> 
> dev-python/bpython has no dependency on net-zope/zope-interface, even with
> FEATURES="test".

Did you try "emerge -p dev-python/bpython" with the test-Feature set?
I rechecked it and after all it gives:

FEATURES="ccache collision-protect":

[ebuild  N     ] dev-python/pygments-1.5  USE="-doc -test" 
[ebuild  N    ~] dev-python/bpython-0.11  USE="-gtk -urwid"

FEATURES="ccache test collision-protect":

[ebuild  N     ] dev-python/pyopenssl-0.12  USE="-doc" 
[ebuild  N     ] net-zope/zope-fixers-1.0 
[ebuild  N     ] net-zope/zope-interface-3.8.0 
[ebuild  N     ] dev-python/twisted-11.0.0  USE="crypt -gtk -serial" 
[ebuild  N     ] dev-python/nose-1.1.2  USE="test -coverage -doc -examples" 
[ebuild  N     ] dev-python/pygments-1.5  USE="test -doc" 
[ebuild  N    ~] dev-python/bpython-0.11  USE="-gtk -urwid"

[Comment 10 Arfrever Frehtes Taifersar Arahesis 2012-09-27 00:35:39 UTC]

(In reply to comment #9)

It confirms what I said. Only test suite of Pygments needs nose, whose test suite optionally needs Twisted Core, which needs zope.interface.

(Some tests from test suite of nose are skipped when Twisted Core is not installed.)

[Comment 11 mike 2012-09-27 00:54:46 UTC]

Yes, it's an (indirect) dependency and tests fail after all. But how I wrote: "The comment about the failed compilation on 2012-09-25 00:01:13 UTC was just made for completeness."

[Comment 12 Arfrever Frehtes Taifersar Arahesis 2012-09-27 01:01:03 UTC]

(In reply to comment #11)

Firstly run: FEATURES="-test" emerge -o bpython
Next run: FEATURES="test" emerge -a bpython
You do not need nose and its dependencies for testing of bpython.

[Comment 13 mike 2012-09-27 01:29:23 UTC]

(In reply to comment #12)
Thanks, I'm familiar with those options.

[Comment 14 Myckel Habets 2012-11-04 08:47:49 UTC]

Builds and runs fine on x86 for me. Please mark stable for x86.

I do however noticed that the wrapper script for bpython-urwid is created, even when USE="-urwid". Executing this script fails then, as expected. Most likely the same happens with the GTK+ wrapper script, but I didn't test that.

[Comment 15 Andrey Grozin 2012-11-15 11:27:17 UTC]

(In reply to comment #14)
> I do however noticed that the wrapper script for bpython-urwid is created,
> even when USE="-urwid". Executing this script fails then, as expected. Most
> likely the same happens with the GTK+ wrapper script, but I didn't test that.
I've fixed this, thanks. By the way, with USE=-gtk everything was fine, no fix needed.
I've also fixed bpython.disktop, so that it now passes validation without warnings.
Please stabilize.

[Comment 16 Andrey Grozin 2012-11-15 11:29:03 UTC]

(In reply to comment #15)
> I've also fixed bpython.disktop, so that it now passes validation without
> warnings.
Argh. bpython.desktop, of course.

[Comment 17 Agostino Sarubbo 2012-11-20 20:47:55 UTC]

x86 stable. Last arch, closing