On 5/13/05, raid517 <raid517@fairadsl.co.uk> wrote:
I agree, it is silly that the kernel is trying to do this. It should be
a user space function and not a kernel space one. You shouldn't have to
fight to repair bootsplash every time you change/compile your kernel.
However I just recently tried splashy from the Alpha repositories - and
I found it wasn't quite ready for prime time on my system. It caused
init to go crazy on reboot with a zillion error messages, finally
stopping with an error saying that something was attempting to 'respawn
too fast' - and that it was 'quitting for 5 minutes.' The only way to
fix this was to chroot into my Debian unstable install and remove
Splashy. Also it seemed to start very late in the booting process -
which I think is simply due to the fact that it doesn't have intrid support.
So until these issues are sorted out, I would rather continue to use
bootsplash - if that is I can get it working. Which requires that I
first somehow overcome this paradox.
Sounds like the really early, dark ages, of splashy. Right now is pretty stable.
Jump to #splashy @ irc.freenode.net. There are a lot of people there
who can help.
Splashy is very non-intrusive now. It won't ask for anything at
install time or remove/purge. And it should work correctly the first
time you reboot.
However, I wouldn't be surprise if it fails for a few people out
there. We are still sorting through those issues: my BIOS doesn't
support foo. vesafb doesn't work at res bar, or with the motherboard
blah. Etc... We try to open a bug for e/a person that reports
something like this in Alioth's project repo. We have squashed plenty
already. And the stuff missing is minimal. That's why we decided to
start releasing some test packages.
However, if you feel find about patching your kernel for e/a
bootsplash or kernel release, that's fine. At least you know there is
an alternative out there.