| 1 |
On Wed, 24 Sep 2014 17:42:14 +0000 |
| 2 |
bugzilla-daemon@g.o wrote: |
| 3 |
|
| 4 |
> DO NOT REPLY TO THIS EMAIL. Also, do not reply via email to the person |
| 5 |
> whose email is mentioned below. To comment on this bug, please visit: |
| 6 |
> |
| 7 |
> https://bugs.gentoo.org/show_bug.cgi?id=523634 |
| 8 |
> |
| 9 |
> --- Comment #1 from Zac Medico <zmedico@g.o> --- |
| 10 |
> Also, note that the pkg_preinst code runs the test from ${D}, in |
| 11 |
> order to support binary packages. However, with the new setup.py |
| 12 |
> install locations, it's going to be a little more complicated to |
| 13 |
> locate the python files that go with the default python interpreter. |
| 14 |
> |
| 15 |
> We could simply remove the preinst sanity check, and just assume that |
| 16 |
> the default python and bash interpreters are compatible. This |
| 17 |
> assumption holds in the vast majority of cases. Issues generally tend |
| 18 |
> to arise only if users use the emerge --nodeps option (which is not |
| 19 |
> recommended). |
| 20 |
> |
| 21 |
|
| 22 |
OK, I've commented out the pkg_preinst sanity check for now. |
| 23 |
|
| 24 |
I've also updated the 9999 ebuild to create the deprecation symlinks. |
| 25 |
setup.py has been updated as well. |
| 26 |
|
| 27 |
I have a 2.2.14_rc1 tarball uploaded and I've attached the 2.2.14_rc1 |
| 28 |
ebuild here as well. Although it is the same code installed as 9999 |
| 29 |
currently. |
| 30 |
|
| 31 |
If our testing doesn't come up with any more regressions, I'll release |
| 32 |
the rc1 for some additional testing. |
| 33 |
|
| 34 |
-- |
| 35 |
Brian Dolbec <dolsen> |
Archives