Bcachefs Fixes Pull Once Again Frustrates Linus Torvalds - Two Choices Offered

[RussianNeuroMancer]

Personally both bcachefs (2 months ago) and btrfs (10 months ago) have lost data for me and thats on a simple 1 disk setup both required a format of the fs to get it working again, scrubs and "repairs" did bugger all.

Sounds like defective hardware to be honest. I running btrfs on everything (from SBC to storage servers with x86 tablets in between) for like ten years and the overall amount of reboots/power-loss is in hundreds - not a single time I lost a data stored on btrfs.

[mobadboy]

You are are a small, jealous, insignificant corporate shill.

wow he knows me

[Developer12]

They are nice to each other again. Please simply read the mailing list.

They sure as hell weren't as of the time of writing. And it sure as hell doesn't change the fact that bcachefs "fixes" have broken builds for multiple other groups of people *multiple* times after being mainlined. Frankly every single batch of fixes seems to break builds for multiple somebodies in different areas.

[PuckPoltergeist]

Perhaps you didn't understand me, or maybe you are just trolling me, but I'll answer:

I was saying that your so-called "social problems" in reality are minor issues, which you and some others are just overblowing by misrepresentations.

That has nothing to do with any "tin foil hat". Essentially, you are doing a "red herring" fallacy.

If it was just me (and some others), I would agree with you. But as it's also people Kent is working with, I don't think it's minor. But in the end, I'm with Josef

[PuckPoltergeist]

And those "social problems" are the best ones, especially when an elephant is made out of an ant, supported by lies, misrepresentations and fact-twisting.

Lemmy quote:

Pls remove the tin foil hat

[PuckPoltergeist]

They have addressed the issue but you have to create an entirely new btrfs partition, its impossible to fix a currently existing partition as the fix requrires a breaking change to the on disk format.

At this point, you may as well choose a better filesystem which has a known history of not having these issues, i.e. openzfs

Nice whataboutism, switching from social problems of offending and accusing to technical discussion

[jacob]

As someone that was around at the time and knows how FAT* works, it wasn't clever even in the day. The most I'll give it is: "it worked... most of the time". Most people that could program computers in that era could do the same thing and often did, often arguably better. We had to in order to get the most out of the limited available resources and mechanical timings. In fact, there were plenty of other software that created custom physical and logical disk layouts to minimize lengthy access times for mechanical storage that catered to the quirks of the hardware.

(*) FAT12, 16, 32 & vFAT, exFAT are all basically just extensions of the version that came before

Yes it was common to have custom disk layouts, I've been there, done that The point is, those custom layouts were optimised for particular workloads and data types. But as a general purpose filesystem meant to handle more or less anything without any assumptions, FAT was not nearly bad, especially compared to the competition (we are talking personal computers and floppies, mid-80s):

• CP/M had a far more primitive storage system
• The various "Disk Operating Systems" used respectively by the C64, Apple II and other 8-bit computers usually didn't even support hierarchical folders and were much more limited
• The original MacOS filesystem was a hack, more akin to those custom disk layouts than to a bona fide, general purpose filesystem. It also did not support hierarchical directories (the "Finder" file manager did, but it was only handled at the GUI-level, the OS itself knew nothing about it).
• The Atari ST's TOS used basically FAT. There were minor differences but for all intents and purposes it was a FAT.
• The original Amiga filesystem was simply terrible, it didn't have any block index or actual directory. That made it incredibly slow and error prone. Later, Amiga FFS was arguably much better, but IIRC that didn't come until 1990 or so.

[mdedetrich]

I really don't understand how you can read this any differently. Kent literally said they are not addressing issues. It's incredibly rude, especially since there have been so may raid56 patches, and there's still work being done for the raid stripe tree feature.

Yes, there's also design flaws, which Josef acknowledge, but it sounds like the only acceptable approach for Kent would be to throw the entire fs into the trash and help him develop bcachefs instead.

Honestly, his entire response to Josef sounds like a slap to the face and I'm not surprised that no-one wants to work with this guy.

They have addressed the issue but you have to create an entirely new btrfs partition, its impossible to fix a currently existing partition as the fix requrires a breaking change to the on disk format.

At this point, you may as well choose a better filesystem which has a known history of not having these issues, i.e. openzfs

[Mathias]

Holy shit

> I'm more than happy to work with people, but that's got to be a conversation, and one based on mutual respect.

As someone who spent a year and a half working directly with you,
You're full of shit
[...]
You keep hating on btrfs like an insecure child.
You never afforded any respect to your users, like saying sorry for your constant temper tantrums on IRC at noobs asking simple questions.
You put your ego before stability,
[...]
You've been a solo dev for over 10 years, and you'll stay a solo dev for another 10 if you don't harden up, put your big boy pants on, lay off the weed, and get a therapist.

Also: Ted Ts'o lectured him that his test infrastructure is easy to spin up and use...

[Deathcrow]

What I have said, and I repeat again:
perhaps they are not addressing issues because those issues are too hard to address, due to early design flaws, and not due to developers not wanting to.

I have never heard someone make a complaint about "$issues not being addressed" if the $issues are inherently un-addressable because of a design flaw. You'd never say it this way.

You're either gaslighting or intentionally obtuse to stan for Kent.