1 |
On Saturday 24 April 2010 01:08:46 meino.cramer@×××.de wrote: |
2 |
> Hi, |
3 |
> |
4 |
> before getting into too much trouble better I aask: |
5 |
> While updateing I got the following message: |
6 |
> |
7 |
> ('ebuild', '/', 'x11-base/xorg-server-1.8.0', 'merge') conflicts |
8 |
> with <x11-base/xorg-server-1.7.99 required by ('installed', '/', |
9 |
> 'x11-drivers/nvidia-drivers-195.36.15', 'nomerge') |
10 |
> |
11 |
> |
12 |
> !!! The following update(s) have been skipped due to unsatisfied |
13 |
> dependencies !!! triggered by backtracking: |
14 |
> x11-apps/xinit:0 |
15 |
> |
16 |
> |
17 |
> The first one I understand what I wants to say, but: Is the conflict |
18 |
> based on the limition of the nvidia-driver not to run with |
19 |
> xorg-server-1.8.0 even when recompiled after the new xorg-server |
20 |
> is reinstalled or does the latter help to circumvent the problem? |
21 |
|
22 |
The former: |
23 |
|
24 |
http://bugs.gentoo.org/show_bug.cgi?id=315141#c3 |
25 |
|
26 |
> |
27 |
> What the second message wants to tell me is far beyond my knowledge ... :) |
28 |
|
29 |
It's portage trying to tell you how those incompatible versions are in the |
30 |
list to be emerged. A backtrack is simply that - start with what you have, |
31 |
find out what pulled it in, and what pulled that in, till you come to the end |
32 |
(usually something in your world file). |
33 |
|
34 |
Without digging into ebuilds, it looks like xinit pulls in xorg-x11 which |
35 |
pulls in xorg-server which conflicts with nvidia-drivers,. |
36 |
|
37 |
This is a classic case of rule #1 of program output: never expose the |
38 |
underlying implementation in your output. That info is completely useless to |
39 |
most users and needs re-thinking. Decent programming practice says that output |
40 |
shout only be given if the user asks for it like that, with say a --debug |
41 |
option for example. |
42 |
|
43 |
-- |
44 |
alan dot mckinnon at gmail dot com |