1 |
On Fri, 15 Nov 2013 22:06:45 +0000 |
2 |
"Robin H. Johnson" <robbat2@g.o> wrote: |
3 |
|
4 |
> On Wed, Nov 06, 2013 at 04:12:47PM +0100, Michał Górny wrote: |
5 |
> > Please review the following news item. I would prefer committing it |
6 |
> > as soon as I get an ACK from all the relevant parties since the |
7 |
> > issue is hitting users pretty hard. |
8 |
> |
9 |
> I don't know why nobody looked at a better automatic solution for the |
10 |
> python-exec issue, but since it's still biting me on systems, and the |
11 |
> manual solution is annoying, I found a working automatic solution. |
12 |
|
13 |
Had on mind to have some kind of "if this then that" rules in some |
14 |
pre-defined file like the profile updates; this would do specific |
15 |
emerge actions in specific situations (like this case). |
16 |
|
17 |
But as I saw it more as a hack, I didn't suggest it; it is better to |
18 |
keep the Portage tree consistent instead, than to add another file that |
19 |
grows the complexity further. Seems that you have cracked it. |
20 |
|
21 |
> Add this line to the dev-lang/python-exec ebuilds: |
22 |
> PDEPEND=">=dev-python/python-exec-10000:$SLOT" |
23 |
> |
24 |
> I wrote up the full research of it here: |
25 |
> http://robbat2.livejournal.com/240335.html |
26 |
|
27 |
Was wondering why CC missed, but I see you have contacted mgorny on IRC. |
28 |
|
29 |
> If there are no objections, I'd like to do this to the affected |
30 |
> ebuilds in a few hours. |
31 |
|
32 |
It is an improvement and it has been tested on a few systems; I don't |
33 |
think this is something that would hurt or be irreversible, it reverses |
34 |
the condition and adds a SLOT to be easier on Portage. It should work. |
35 |
|
36 |
-- |
37 |
With kind regards, |
38 |
|
39 |
Tom Wijsman (TomWij) |
40 |
Gentoo Developer |
41 |
|
42 |
E-mail address : TomWij@g.o |
43 |
GPG Public Key : 6D34E57D |
44 |
GPG Fingerprint : C165 AF18 AB4C 400B C3D2 ABF0 95B2 1FCD 6D34 E57D |