1 |
On Monday, March 21, 2016 01:08:19 PM Francisco Ares wrote: |
2 |
> 2016-03-21 11:52 GMT-03:00 Alan Grimes <ALONZOTG@×××××××.net>: |
3 |
> > Anyone who boots directly to X'doze without first going through a safe |
4 |
> > console login is a raging mad lunatic who needs his head examined. This |
5 |
> > is linux we are talking about. It's crap. It always has been crap, and |
6 |
> > it always will be crap. Never ever ever trust it. I leave my computer on |
7 |
> > continuously because booting it is such a risk. Every single time I load |
8 |
> > X'doze and find that my keyboard and mouse are working I ghasp with |
9 |
> > surprise. The linux developers, or the penguins as I like to call them |
10 |
> > are so smug on the sublime superiority of the open source approach that |
11 |
> > they never bother to design essential things such as fail-safe design, |
12 |
> > fallback drivers, stable apis so that it doesn't just die if it's not |
13 |
> > compiled against this specific point release. ... You know, the kind of |
14 |
> > things that any competent programmer would think about. =| |
15 |
> > |
16 |
> > Bertram Scharpf wrote: |
17 |
> > > Hi, |
18 |
> > > |
19 |
> > > since an emerge-update-world on my notebook the keyboard |
20 |
> > > does no longer respond in X. This is extremely annoying |
21 |
> > > because when I have xdm in rc-update, X is started right at |
22 |
> > > boot. I have no chance to get back to the console using |
23 |
> > > Ctrl-Alt-F1, and the device in unusable. |
24 |
> > > |
25 |
> > > Yet, this is only a problem of the boot process. At home, |
26 |
> > > when I ssh into the system, I can do an |
27 |
> > > |
28 |
> > > # /etc/init.d/xdm restart |
29 |
> > > |
30 |
> > > and from that point on the keyboard works. It is even |
31 |
> > > possible to disable xdm in rc-update and start it after the |
32 |
> > > boot process has completed. I solved the problem temporarily |
33 |
> > > this way, but the problem probably is a bug and should be |
34 |
> > > reported. |
35 |
> > > |
36 |
> > > So I have a closer look. When I diff "Xorg.0.log" and |
37 |
> > > "Xorg.0.log.old" (after removing the time stamps) I find one |
38 |
> > > line that doesn't appear in the log of the working X. |
39 |
> > > |
40 |
> > > (EE) kbd: Keyboard0: failed to set us as foreground pgrp |
41 |
> > |
42 |
> > (Inappropriate ioctl for device) |
43 |
> > |
44 |
> > > What does this mean? I estimate that "us" is the personal |
45 |
> > > pronoun and not a keyboard layout, and that the server tries |
46 |
> > > to do some chgrp on some /dev/*. I have no clue what to try |
47 |
> > > next. |
48 |
> > > |
49 |
> > > Thanks in advance. |
50 |
> > |
51 |
> > -- |
52 |
> > IQ is a measure of how stupid you feel. |
53 |
> > |
54 |
> > Powers are not rights. |
55 |
> |
56 |
> ... or you may provide checking points, like a script to run all "emerge |
57 |
> world" processes automatically, |
58 |
> |
59 |
> Open source and Linux' software begins with the premisse you know what you |
60 |
> are doing, as if you issue a "rm -fR /" you will get exactly what you have |
61 |
> asked for, a dead system, no "are you sure?" questions will ring. |
62 |
> |
63 |
> Those "craps" made me learn a lot! |
64 |
|
65 |
Me and a friend did that once to a system that needed reinstalling anyway. |
66 |
It doesn't actually wipe everything off the disk and processes in memory are |
67 |
likely to keep running. |
68 |
|
69 |
If we'd been running a shell with a lot of built-in functionality or a decent |
70 |
editor, we might have been able to restore some of the functionality :) |
71 |
|
72 |
-- |
73 |
Joost |