On Mon, Mar 18, 2019 at 12:53:12PM +0100, Gianfranco Costamagna wrote:
> On Tue, 12 Mar 2019 16:09:35 +0100 Adam Borowski <kilobyte@angband.pl> wrote:
> > Control: retitle -1 ITA: btrfs-progs -- Checksumming Copy on Write Filesystem utilities
> > Control: owner -1 !
> >
> > As a long-time user ("thou shalt have no other filesystems before btrfs")
> > and an occassional contributor, I can take it.
> >
>
> Thanks Adam for taking care of it!
Yes, thank you Adam. In Debian is there anyone else besides us (and
nonpackaging Christoph Anton Mitterer), and possibly Hideki, who are
working to make trying out btrfs a more positive experience?
> Just a side note: I think I sponsored a lot of backports for Nicholas in the past,
> he is a good contributor, and he sends patches upstream and to Debian packaging (dated 2017).
>
Thank you Gianfranco :-)
> Maybe you can both work on the package, he is a DM so he might even upload by himself if you
> feel comfortable with his work (Nicholas please tell us if you are still interested in btrfs!)
>
I am still interested, and plan to continue to maintain the
stretch-backport and future buster-backport of btrfs-progs. IIRC I've
been maintaining the stable-backport since ~2016, and would appreciate
DM upload permissions for this purpose.
> In any case, two is better than one, specially when one is an upstream contributor and Debian maintainer :)
>
IIRC Adam is also an upstream contributor; I remember a while back he
sent a patch for btrfs-progs which more explicitly warns a user who is
about to enable raid5 or raid6 profile. That impressed me, and is one
of the reasons I think we're generally on the same page wrt btrfs.
P.S. I suspect the recent sparse/hole punched file bug that affects
transparent compression might also affect btrfs-convert, and when
convert support is presumably reenabled in the bullseye dev cycle it
would be worth adding some kind of warning.
> I hope to see Nicholas back in the team :)
>
Thanks! :-D On that note, Adam, let's form a ng-fs-integration team.
For now goals could be things like adding support for things like
btrfs installation on subvolume, doing upgrades in a rw snapshot
before rotating the snapshot to default boot rootfs, boot
environments, et al. If possible it would be nice to do this in a
modular enough way to add support for bcachefs--when the time comes.
Oh, and Adam prefers to support sysvinit while I prefer systemd, so
we've already got a bit of diversity ;-)
Regards,
Nicholas
Attachment:
signature.asc
Description: PGP signature