[Pkg-zfsonlinux-devel] any news/reply regarding ZFS in NEW?

[Robert Millan]

On 28/02/2014 10:20, Dimitri John Ledkov wrote:
> Hello,
> 
> On 28 February 2014 09:30, Turbo Fredriksson <turbo at debian.org> wrote:
>> I'm basically Ccing half the world in this (only half sorry about that :) and I don't know who half
>> of you are :), but there have been very little information on what's happening with ZoL in Debian
>> GNU/Linux.
>>
>> Aron (and in some part Carlos) seems to have gone a-wall and the list have been VERY quiet. It seems
>> like it's only Aron and me that is actually Debian GNU/Linux Developers (unless other things have
>> happened outside the list that I'm not aware of - Carlos was/is a maintainer if I don't
>> misremembering and Darik is in the wait queue?). And no actually status information/reason from the
>> FTP maintainers about why it have been stuck in incoming for so long (accepted into incoming Sun, 07
>> Jul 2013 16:00:06 - that's more than six months ago!). Have it been rejected? Is it held up for some
>> reason? What can I/we do to help move it along?

Hi,

The proposed package is poorly integrated with existing ZFS packages (e.g. zfsutils for native
kFreeBSD support).

First and foremost, there's a namespace grab which is likely to result in trouble, as I explained
last November (and got no answer):

	https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=686447#117

There are also a number of implementation-independant add-ons which would be good practice to
coordinate in some way with the other ZFS maintainers. I explained this in November too, and
again got no answer:

	https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=686447#112

And annoyingly, there's also been complaints that ZoL developers broke partman-zfs by committing
porting updates that break existing support on kFreeBSD:

	https://lists.debian.org/debian-bsd/2014/02/msg00037.html

I'm happy to see partman-zfs support more platforms, and I don't mind myself if those platforms
are not yet part of Debian when support is merged. But I would at least find it reasonable that
porting changes include an effort to avoid breaking existing production environments. We do this
all the time when porting to kFreeBSD. I think it should work both ways. That I know of, nobody
has spent the time to fix this particular mess yet :-(

-- 
Robert Millan