Open adrelanos opened 5 years ago
What do you think about adding --ignore-fail-on-non-empty to all invocations of rmdir?
Usually one is running rmdir
by assuming that the directory is empty. I'm not sure if adding --ignore-fail-on-non-empty
to all the rmdir
calls within grml-debootstrap might not hide any underlying issues.
Is there any concrete example / steps to reproduce where a rmdir --ignore-fail-on-non-empty ...
usage would actually solve anything that's worth changing our current code?
This happens if $DEBOOTSTRAP
failed (which failed due to my configuration mistake). That erorr gets noticed by grml-debootstrap and it errors out. The failing rmdir happens as another follow up issue during bailout
which I found unnecessary.
For example if check4progs "${DEBOOTSTRAP}" || bailout 1
fails, bailout
would call cleanup
and error out and also wonder why rmdir failed on top of that.
I see, thanks for clarifying, I'll try to look into that.
debootstrap failed because of a mistake of mine (on Debian buster). As a follow up issue, rmdir failed because the directory was not existing yet.
What do you think about adding
--ignore-fail-on-non-empty
to all invocations ofrmdir
?