On Mon, 2016-04-25 at 09:32 -0400, Nicholas D Steeves wrote: > On 25 April 2016 at 07:51, Ritesh Raj Sarraf <r...@debian.org> wrote: > > > > I meant btrfs is going to remain unstable for some time. There'll always be > > mixed results reported. > > > > We should allow, for the brave, to easily try if they want to. > > > > Something like: > > > > rrs@learner:~$ cat /etc/initramfs-tools/hooks/reiserfsprogs > > Given that it's in /etc/initramfs-tools vs /usr/share/initramfs-tools, > that means it was manually copied from /usr/share/doc/package-name, > right?
No. Not in this case. rrs@learner:~$ dpkg -S /etc/initramfs-tools/hooks/reiserfsprogs reiserfsprogs: /etc/initramfs-tools/hooks/reiserfsprogs 2016-04-26 / 15:36:11 ♒♒♒ ☺ > Yeah, I'd be ok with that, because it requires manual > enabling, but I'm still very uncomfortable that a utility that causes > dataloss as often as it works is part of a base Debian install... > I'm not sure if I'd see it this way. Because by that stand, I should patch my "rm" command. I've many a times done `rm -rf .` > What do you think of diverting btrfs-convert to btrfs-convert.real, > and providing a shell script that echos something like?: > I like your thought. This will double poke the user before he knows what he's doing. > Btrfs-convert is known to cause dataloss. > Please do not convert filesystems filled with data you value. > Proceed (Yes/No)? > -- > Yes. This sounds good. > This way the binary is still shipped, but it's unquestionably clear > that this is a dangerous thing to do. Also can/should the wrapper > script go into the initramfs? I'm not sure if initramfs has a policy. I believe initramfs has minimal shell, so the script will need to be posix only. -- Ritesh Raj Sarraf | http://people.debian.org/~rrs Debian - The Universal Operating System
signature.asc
Description: This is a digitally signed message part