Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Thu, 19 Jan 2023 09:05:32 +0100
From:   Stefan Tibus <stefan.tibus@gmx.de>
To:     Andreas Dilger <adilger@dilger.ca>
Cc:     Dave Kleikamp <dave.kleikamp@oracle.com>,
        Christoph Hellwig <hch@infradead.org>,
        linux-fsdevel@vger.kernel.org,
        jfs-discussion@lists.sourceforge.net,
        Harald Arnesen <harald@skogtun.org>,
        linux-kernel@vger.kernel.org
Subject: Re: [Jfs-discussion] Should we orphan JFS?
Message-ID: <20230119080532.crn7wzo4jz5x5ng3@tibus.st>
Mail-Followup-To: Andreas Dilger <adilger@dilger.ca>,
        Dave Kleikamp <dave.kleikamp@oracle.com>,
        Christoph Hellwig <hch@infradead.org>,
        linux-fsdevel@vger.kernel.org, jfs-discussion@lists.sourceforge.net,
        Harald Arnesen <harald@skogtun.org>, linux-kernel@vger.kernel.org
References: <f99e5221-4493-dba3-3e80-e85ada6b3545@oracle.com>
 <393B8E4A-8C9A-4941-9AFF-FAC9C0D0B2DA@dilger.ca>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
In-Reply-To: <393B8E4A-8C9A-4941-9AFF-FAC9C0D0B2DA@dilger.ca>
User-Agent: NeoMutt/20170609 (1.8.3)
UI-OutboundReport: notjunk:1;M01:P0:7Kk14hf0FvI=;umrs9NNFVJ/HMONmf0Ix4XSWy3L
 Jhk6zpuUJe+gIeDWV/CPRtixr0CAmGdjGTGo7PXCHsed17J7ZPCn3JM6Af2qUl4tAhENg/HEW
 qpWgVjiRL6Jq75SiZMEMNOattbmE1UBRkvZCUPaEdMRT7NZ68/eiFhGJCuWPVPSjkv/DRx2bP
 ZTR0XU8M1l/U2TifawSmFQPA42rGIhjp4cM73RRTS+Wn/M/9zA17wla0e6jIrAkdKx1JiYziN
 7mRkr66GB+TTLnxtsD7LbLGoijydHKgDqSzcKiov9ZUNiWVhYyWqFbbmWmxN7kINDDzKnIkZA
 ebn4stPvU9Swc4wCaC/LE1Fzb+fgZLCOAvKe3eWzijN5esGBoROGQHwuy+HQ1aK/BSyA1LbF0
 s4Q+uQ1hYr08AQpmYdU1rTmj6RmFdX39K1xFdQm4ck/Pfc1GTaQ6M+6VAYcYqD7OPNEWG14FM
 /RYW5AIdglBQpUt7+xWfiW6/5Y0Xe9KCfM5H6TRoEQX2nTPIU2XCoW2xalGAPTIK/cQ6z9M+E
 zWIlGONU1x1+gnUm8NAKp7FF4Gv96TrBbXeiZT93znnIwet5iJDgUV/0oJHdo7VHXGobs4UUu
 fRAfQgPG2afZ8rWbUdDYMlwQICp//Cg6EQfoyRLXUvx7Ls2/LKrfcRK1Z1qWetBOwZ2XZYtFF
 J32yEREqyVpqbbbqGmgmbrgoOLzWLUZxO8tSP6s83VWuELB0lyqyu3QnE6Hbc5HvLZ/Iw6Qqo
 qKuxCKBVbCKPwz5AXv5ExjO8VJSSxGFzwMH5Lohp44zjItDg+a3UjXTK6wL97H0EzFW49s9qh
 /EcJmQC60FysBrAtLQHwvBbIqgkLDmhNvIITeb4hZTjodwc1homNzTgAKx7Z/ReiAurHC7zaF
 SuwBV0Y8lntek2jYPnGC6YL/9T86OdRDNAuCMh9yRfeczCfiZJe4+/eF8XD5yFC3i1H1rpMUX
 Me7eoLzkCDtX4Fj8lt3w3UMK8L8=
Precedence: bulk

Hi all,

While I am mostly an ordinary user running Linux on my own machines at
home, I must say that I have been a happy user of JFS from quite early
on on all my Linux installations, for which I use the Debian distro. I
am also using it on external HDDs and SSDs. In the past I have also been
administrator for a few workgroup servers at my university for about 10
years and there we have transitioned from EXT2 and EXT3 to JFS on LVM at
some point. Only recently I have started using BTRFS because of its
additional features on my newest PC. However, I would not make that
transition on older PCs with less resources. And it is some hassle to
convert all existing filesystems to something else.

I cannot provide hard facts like performance or so for the decision to
use JFS. My first contact with journaling file systems had been on a few
AIX (3.x/4.x) machines and later on with JFS on OS/2. So having started
off based on the code of JFS for OS/2 certainly contributed to the
initial level of trust when giving JFS on Linux a try versus EXT4 and it
didn't let me down.

=46rom my perspective it would be sad seeing it removed while other much
older filesystems (or other features) are retained. But I also know that
in the end it depends on the capability, availability and willingness of
developers to maintain it. And, frankly speaking, I really do not know
how much effort it is to keep the code compatible to new kernel
versions.

So this is my vote against orphaning JFS. I still think it is a good
filesystem and certainly useful on systems with less resources where one
would probably not use BTRFS, ZFS or so. But whatever the final decision
will be, I would like to thank you all for contributing to JFS and
keeping it running over the past >20 years.

Best regards
Stefan


On Sat, Jan 14, 2023 at 05:09:10AM -0700, Andreas Dilger wrote:
> On Jan 13, 2023, at 08:15, Dave Kleikamp <dave.kleikamp@oracle.com> wrot=
e:
> >
> > =EF=BB=BFOn 1/13/23 7:08AM, Harald Arnesen wrote:
> >> Christoph Hellwig [13/01/2023 06.42]:
> >>> Hi all,
> >>>
> >>> A while ago we've deprecated reiserfs and scheduled it for removal.
> >>> Looking into the hairy metapage code in JFS I wonder if we should do
> >>> the same.  While JFS isn't anywhere as complicated as reiserfs, it's
> >>> also way less used and never made it to be the default file system
> >>> in any major distribution.  It's also looking pretty horrible in
> >>> xfstests, and with all the ongoing folio work and hopeful eventual
> >>> phaseout of buffer head based I/O path it's going to be a bit of a d=
rag.
> >>> (Which also can be said for many other file system, most of them bei=
ng
> >>> a bit simpler, though).
> >> The Norwegian ISP/TV provider used to have IPTV-boxes which had JFS o=
n the hard disk that was used to record TV programmes.
> >> However, I don't think these boxes are used anymore.
> >
> > I know at one time it was one of the recommended filesystems for MythT=
V. I don't know of any other major users of JFS. I don't know if there is =
anyone familiar with the MythTV community that could weigh in.
> >
> > Obviously, I haven't put much effort into JFS in a long time and I wou=
ld not miss it if it were to be removed.
>
> I've used MythTV for many years but haven't seen any particular recommen=
dations for JFS there. Mainly ext4 and XFS are the common filesystems to f=
ollow the main distros (Ubuntu in particular).
>
> Cheers, Andreas
>
> _______________________________________________
> Jfs-discussion mailing list
> Jfs-discussion@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/jfs-discussion