It's over.

>>105769155 (OP)
so is linus giving up on linux?

>>105769247
he's dropping bcachefs

>>105769155 (OP)
qrd on this situation? Is bcachefs finally the btrfs+zfs+xfs god tier fs option and Linus won't merge it due to some drama?

>>105769277
the bcachefs dev keeps trying to skirt development rules and arguing with linus so linus is dropping it from the kernel
https://news.itsfoss.com/linux-kernel-bcachefs-drop/

Good riddance. BTRFS is good enough, effort should be spent on BTRFS and indirectly supporting ZFS to the extent possible.

>>105769277
kent is a dick to everyone ep #3283928392...
I wonder why this dude is still allowed to contribute, lennart was banned and told to fuck off back to the nearest gas chamber when he submitted dogshit code and blamed the entire planet instead of fixing it.

It's to be expected, from what i've seen Kent doesn't work well together with the kernel folks.

I was excited for bcachefs (been running root on zfs on linux for nearly ten years), but unfortunately the dev is an idiot, and not just in terms of pissing off Linus for no reason. I thought there was no chance of btrfs becoming reliable before bcachefs became feature complete, but bcachefs still doesn't even have a scrub. I legitimately don't understand what the fuck the guy is thinking, since he's making a filesystem with bold claims of reliability without the scrub necessary to actually test and validate that claim.

It's a shame but I guess my next build will also have to deal with the pain of root on zfs, even with openzfs going to shit since it shifted to linux-first development.

If he's moving too fast for the kernel why doesn't he just do a out-of-tree module until it's stable enough for the kernel?

btrfs is the only file system that shat itself out of nowhere in all my time of usin linux never using that shit again

>>105769827
What’s wrong with zfs?
Our company had to drop btrfs due to hard-locking up when replicating out to a mirror.

My opinion is, like, go back to standard ufs for root, and zfs for data volumes.

>drama among fat ugly Linux nerds
This is why no one takes Linux seriously

>>105769891
>incremental send is locking up jeet's servers
jej

>>105769155 (OP)
how long until he shoots up a school?

>>105769891
ZFS overall is fine, but running root on zfs on linux has always been a bit painful, especially on a distro that keeps the kernel up-to-date. There's also been weirdness with grub and next build I'll be trying to go grubless. I've been saved multiple times by root on zfs on my machines so I consider it worth the pain. In one case, I lost both root drives (usb flash drives hit their write endurance limits) of my home server but migrated to three new usb drives without needing to reboot.

But zfs becoming linux first introduced two new problems. One is that linux devs just aren't as careful and there have been a series of very buggy releases. The same dev culture that led to btrfs being an unstable mess of abandoned features infested zfs. The second issue is that Linux support actually got worse. New linux kernel support is now tied to feature releases, so it actually takes longer to support new kernels than before when ZoL was only maintaining compatibility patches.

>>105769894
loonix was always a garbage os for garbage hardware and users. loonix fans = either coping while misinformed, a cult-like desire to belong to some community, and/or because it's not some other os they really hate. Nobody actually likes loonix because it's a well designed os that runs everything well.

still using ext4

>>105770018
Thanks for the nice explanation.
> lost both root drives, fixed, no reboot
That’s where I was going with plain old ufs… disposable boot drive… but I don’t mind downtime to swap them with cloned ones. You should be the resident uptime autist :)

>>105770243
Nothing wrong with that… it’s practically the new ufs.
The theory here, is just make frequent backups of critical stuff a la rsync

>>105770109
>Nobody actually likes loonix because it's a well designed os that runs everything well.
I can attest to only liking GNU+Linux because it is fun.

>>105770289
It's not only that I didn't reboot, but after the second I wasn't sure if I could reboot, so I kept it running for another 3 or 4 months until I was ready to build a replacement server. All the hardware in it was getting old anyway and I was planning the replacement with the new generation of CPUs.

My current desktop also had a bad SSD, but since I was using ZFS mirrors I was able to easily swap it out. I also manually mirror (and back up) my EFI partitions which saved me when a grub update scrambled one; the other one was still usable for booting without needing to go into a recovery environment.

>>105770312
I backup once a year

>>105769155 (OP)
ZFS wins again.

I'm content using ZFS. Works on every OS I could ever decide to use (NT, BSD, Linux, XNU, et al.), good feature set, good tools.
Don't see the point in making a new FS that's platform specific and doesn't offer anything new.

>>105769247
can you read or are you just faking it with ai or some shit?

Love ZFS
Hate BTRFS
simple as.

can someone redpill me on all these different file systems? should i just stick with ext4 if i am not running a complex system of automatic backups?

>>105770109
TRVKE. I came to this conclusion during my undergrad after taking an OS course.

Why not just have one?
Why must Linux be such special snowflakes.

tinkertroons

>>105771803
None of them have much to do with backups, and none of them are substitutes for backups. I think the biggest benefit is checksumming, any checksummed filesystem (even running dm-integrity under ext4 would be an improvement) at least tells you when something is starting to go wrong. Pooled storage is nice to work with, CoW makes snapshots effectively free, encryption being built-in and on the filesystem level instead of the block device level can be a performance win when you don't need it. You can get many of the advantages with other tools, see stratis, but there are functionality gaps that they just can't bridge (like self-healing, at least to the same degree) and it's a pain to manage so many moving parts.

>BTFOFS

>>105769247
he hasn't been the same for five or six years now. Linux and especially Linux desktop are now declining, the peak is behind us, then will come the dissolution.

https://lkml.org/lkml/2021/6/10/957

>>105772192
I'm still waiting for that mass die-off that was supposed to happen in 1/2/3/4/5 years from the vaccine.

>>105772209
no mass die off but in Europe the excess mortality is still at +10% compared to 2016-2019

>>105772238
Big goalpost move from "literally everyone who took the vaccine will drop dead in 12/24/36 months" that was being screamed from the rooftops.

>>105772252
that was a conspiracy theory from /pol/, no one took that seriously.
for clarity, here the official stats from the European union for every month since 2020
https://ec.europa.eu/eurostat/cache/website/covid/vis/DIR_CV2/?simple=true&hideIndicators=true&indicator=0&lang=en

>>105772277
>no one took that seriously
Oh, I wish I were as naive as this.

Honestly we're getting at the point where the baby boomers are starting to age into natural deaths (the oldest are early 80s now) and between that and people who got covid I'm not surprised. We're probably looking at a good three decades of increased background deaths. But I've yet to see the mass die off of everyone who got vaccinated that people were (and still are, but they've moved to 5-10 years) claimed will happen due to tentacle aliens in the vaccines or whatever they thought.

>>105769637
>lennart was banned and told to fuck off back to the nearest gas chamber when he submitted dogshit code and blamed the entire planet instead of fixing it.
Wait, what?! How did I miss this?

>>105769155 (OP)
Btrfs won. Kneel bcachefs and zfs trannies.

>>105769155 (OP)
I'm out of the loop with the filesystems. Why would you use bcachefs, over other cows, such as btrfs and zfs?

>>105772192
Microshaft will absorb us into WSL and we'll become "WSL Desktop Edition"

>>105769637
is that the systemd guy?

>>105773557
the wayland compositor on wsl is already using microsoft only extensions.

>>105769247
He'll focus on his Youtube channel

>>105769155 (OP)
reiserfs will make a comeback

>>105772685
btrfs is full of bugs and zfs is a shitty port.

>>105772277
>that was a conspiracy theory from /pol/
No, it was obvious well-poisoning faggotry.

That's not good my root fs is bcachefs

>>105774103
So, a conspiracy theory from/pol/.

>>105774052
wouldn't be better to just to fix all these bugs in btrfs, than reinventing essentially the same thing?

>>105772238
1.2 is not 10 lil guy

>>105773586
lennart is instead the systemd dude

>>105773753
I would watch Linus Torvalds' Tips.
... wait....

I can't believe we still don't have a real next generation filesystem

>>105772277
>no true scotsman
>suddenly trusts official data
Could you just say "I was wrong and I've changed my mind?"

>>105772252
>literally everyone who took the vaccine will drop dead in 12/24/36 months
can you show me the source of the quote?

Because i do have quite a lot of sources for claims that all unvaccinated will die. This whole
>i took the vax and didnt instantly drop dead, so i won!
seems like a very dishonest strawman.
If the bast thing your vax does is to not kill you immediately, it is still the better decision not to take it.
The people who didnt take it, made objectively the correct decision.

>>105776614
and they got rid of him? based

>>105777701
I think it was a temporary ban but linus was big mad at him (it was partly about breaking user-space, everyone knows linus stance on this and yet), lennart even wrote a blogpost bitching about linus sending trolls to harrass him and how linux would be better if it was led by someone else *wink* *wink*
that was a decade ago or something, I can't find the mailing list messages anymore

>>105776693
>Could you just say "I was wrong and I've changed my mind?"
4chan isn't one person you retarded faggot

>>105777751
I hate both of them, but I loved it when torvalds finally and publicly bitch slaps some idiot.
If you look at the history, and look at what became of the situation, it’s clear that linus was way too tolerant of their initial malfeasance until it became a lot harder to solve problem.
Look at the rustards that tried to invade.
I don’t like openbsd, but theo would have never put up with that.
Before python was abandoned, do you think vanrossum would have entertained shit like that?

What’s the limit on the number of languages and software development systems required to build a large progect? Typically, it’s one.

Except for microsoft windows, where they invaded themselves with that .NET shit, but that is mostly periphery.

>>105770378
>but after the second I wasn't sure if I could reboot, so I kept it running for another 3 or 4 months until I was ready to build a replacement server.
Relatable.

>>105778693
>I don’t like openbsd, but theo would have never put up with that.
old linus would have told them to pound sand with their female penises...
now that rust is there, the only path forward is to replace all the C code, I don't see a world where C coexists with an half-assed rust api, either they stick to rust and remove C or ditch rust because it's not going anywhere anymore.
I pray for the latter but I guess the former is more realistic...
>Except for microsoft windows, where they invaded themselves with that .NET shit, but that is mostly periphery.
most of the internals are still in C/C++ anways, I've heard they were trying rust but it's still mostly a monolanguage codebase for kernel space.

>>105770109
Well said. I only use linux because Windows and Mac are too cucked for me, and the BSDs. while being interesting and better designed, have less support and manpower. I don't care about freetardism, I use proprietary software when needed, and in most cases the proprietary software is much better and pleasant to use than an open-source one, because people are getting paid for their efforts.

Linux is really that mediocre middle ground mostly produced by corporations, where something is considered working when it can be shipped.

>>105769155 (OP)
Now maybe he can start a microkernel project, called say Minux, as an apology for holding back OS progress for over 30 years.

Oracle can end this shit. They could just add GPL compatibility in the CDDL license and change the ZFS one.

>>105777751
This? https://lwn.net/Articles/518945/