On Mon, 26 May 2008 13:36:37 -0400, Matthew Saltzman wrote:
[...]
> Vesa driver? I've seen the word during boot-ups on at least
one
> machine; but since it means nothing to me, I don't even recall which
> machine. I'll be glad to try that or anything else I can; how do I do
> it?
I believe you can do this:
When you see the GRUB splash screen on boot, press a key.
Did that -- and nothing happened.
The machine was still where it had been, saying it had started
the network manager. I hit Ctrl-Alt-Delete, and it went down as far as
sending the TERM signal -- and then sat there, without doing anything I
could see, but with the machine's light still on.
I hit the reset button. It went as far as the grub menu, and I
hit a key. That left it with its F9 line highlighted. I hit e -- and
nothing happened. In fact, it didn't respond to anything on the keyboard
-- the first time I've ever seen grub do that.
I hit the reset button again, and this time it let me tell it to
edit both times. Then it froze again.
Select
the top kernel line of the options offered and type 'e', select
the kernel line and type 'e'. Append 'video=vesa' at the end of
the line. Type [Enter], type 'b'.
I hit the reset button yet again, and did all the above hastily
(but, I hope and believe, accurately) -- getting it all in ahead of
another freeze.
Then X should detect your video card as vesa, which is the most
generic
sort of interface.
It booted, but only into another login line, white on black; I
logged in -- and it gave me my prompt, along with a line saying "Can't
open display."
If the machine boots, you can boot in runlevel 3 (edit the kernel
line
in GRUB as above, but instead of "video=vesa" add "3". Then
you'll get
a virtual console to log into. Then (as root) run
system-config-display --reconfig
and see if that helps.
It tried some sort of traceback, said something cryptic about
/usr/share/system-config-display/xconf.py, and (apparently) tried to
import something from somewhere. Then it said
ImportError: No module named ibmasm
and gave me my root prompt back.
Meanwhile, back at the ranch, the scp -r that I was running quit
with the message
scp: error: unexpected filename: ..
lost connection
So there must be some stuff I didn't get; I don't know what, but
I got vast amounts -- enough to run my df up, despite deleting hand over
fist, from its original 50% to 78%. (Baobab, which sees only a 35.2 GB
drive, makes it 72.8%. I'm pretty sure this machine has a nominal 40 GB
drive; I thought the #1 machine had two -- 40 GB for Fedora and 40 GB for
XP, just because the one thing I do on it is GPS-compatible topo maps,
which are apt to be graphic-intensive and to grow hugely -- but it may be
80 and 80.)
The actual numbers on machine #2 are
[btth@hbsk ~]$ df
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
36693560 26814088 7985484 78% /
/dev/sda1 194442 18754 165649 11% /boot
tmpfs 501108 48 501060 1% /dev/shm
The corresponding top line numbers from df on #1 are
73545144 30178956 39570016
Since I haven't even tried to salvage anything from anywhere but
my user's home directory, it looks to me like I have very nearly all of
it. (I don't keep financial or otherwise critical files on any computer.)
Bottom line: I'll gladly keep this up so long as you, Patrick
O'Callaghan in the X thread, or any other clearly knowledgeable person
here thinks it worth while. Hey, we might uncover the anaconda bug! But
if the worse comes to the worst, I'm confident I can now live with the
results of wiping the hard drive and re-installing F8 from scratch.
--
Beartooth Staffwright, PhD, Neo-Redneck Linux Convert
Remember I know precious little of what I am talking about.