Linus Torvalds Comments On Bcachefs Prospects For Linux 6.6

[blackiwid]

From reading this thread the things I am picking up on are:
Victim complex. (Whether warranted or not in the context of their life and social connections.)
Behavioral problems/disorder. (Becoming easily irate)

Well that is a moral question conservative men believe more in strong fix hierarchies and procedures while more liberal people are more freedom orientated.

I would also give a utilitarian view on this, A I gone more on his side with the reservation that I don't know all details and could be wrong, so from a actor with imperfect information, but he tries to get this code in not since yesterday that is not his first attempt so do you have to beg the masters for 1 year to get it in? I also stated that it matters if the contribution is better.

Like Processes get ignored or for certain groups there are huge exceptions like the Graphics driver company can basically push trillions of lines of codes 1 day before it get released more or less.

So your rules have to be checked under a utilitarian framework, and yes there is maybe a danger of some apparently big egos getting hurt and some internal fraction but it could also happen a external fracture, there is no guarantee that linux tree will always stay the Upstream, if rejected or long term delayed code becomes to valuable for many folks to have, one of the forks could become successful, there is already linux-libre it just get's not much traction because of incompatibility to most modern hardware.

I am just saying that you get not more people into kernel development if you make it extra hard to contribute to the kernel. Would it not be the kernel most other opensource projects would be thankful for this great contribution or if not there would be a fork the next day.

Now the argument is that this project is so big and different from every other opensource project, fine but if we have here problems should they not be fixed instead of talking about what way Linux talks to people. Somehow they never can get along, Linus with the kernel devs (ohh he said something mean) RMS and Linus, Torvalds and Lennart Poettering, maybe this is deeper than this new guy being not formal enough.

I think iterating this needs to have 1 knee-jerk action (Hauruckaktion), to make it happen and then things can be calibrated afterwards. Adding it part by part seems very inefficient to me.

The code seems to have good quality I don't hear "this is trash" like linus had no problem to call other peoples code. That's my 2 cents.

[oiaohm]

It's insane. I don't understand why they don't use a bug tracker! I even don't understand why they don't use something similar to GitLab

Linux kernel is a multi national development and this causes issue. Like it or not email system with signed messages is quite good at working around countries filtering their internet where items like GitLab get blocked lot more simply.

We don't have a good development tracking model like bug trackers that is highly decentralized that works well. There are different open source projects that stay with mailing list because they are very fault tolerant including to faults caused by governments.

Please don't take this as me saying there should not be a better solution. Just there are advantages to the old school email transport system that should not be ignored.

https://docs.kernel.org/admin-guide/reporting-issues.html

The Linux does have a bugzilla bug tracker. Yes one of reasons it not more popular is how often it end up blocked by one country or another firewall for no good reason.

Yes for all the development there is a not well solved problem here.

[blackiwid]

Imagine if all the 5k or how many active contributors sent all their work to Linus and went "I'm too special to follow the established procedure"

You moving the goalpost, the claim was not that linus is anoyed and overworked and that's why he should not do it, the claim was that if he overrules the decisions his "generals" did then they would feel overgone and there could come a divide, not that this makes a president and linux would be to overworked.

[blackiwid]

That's a weird way to look at it. It makes it sound like the rules are there just to annoy people, but with good enough stuff you may bribe the gatekeeper. In my view, the only excuse for such a "fast path" would be that something is already horribly broken, and you need to fix it fast (i.e. when the chance of it becoming worse than it already is, is negligible).

That assumes that every rule is a necessary step to produce / secure good code quality, while it could just be safeguards like in air planes where everything is secured by having every system double and ideally not the identical in implementation but in function.

Again this is practice that happens already, Linus accepts regularly from graphic cards drivers huge amounts of code in very late points of the development, he knows that a good graphic stack is for desktop Linux / gaming Linux important and so he pays the price of taking a risk for it, so that Linux can stay more relevant.

[christianb]

It looks like Kent won't be resubmitting his patches any time soon, which is a shame.

But my life is different now; I have new and better people in my life,
and I have to be thinking about them, and if merging bcachefs means I
have to spend a lot more time in interactions like this then it's going
to make me a shitty person to be around; and I don't want to do that to
myself and I definitely don't want to do that to the people I care
about.

I'm going to go offline for awhile and think about what I want to do
next.

Still, reading through all the messages, I see him far more at fault for this than the rest of the developers.

[oiaohm]

Again this is practice that happens already, Linus accepts regularly from graphic cards drivers huge amounts of code in very late points of the development, he knows that a good graphic stack is for desktop Linux / gaming Linux important and so he pays the price of taking a risk for it, so that Linux can stay more relevant.

This is where it pays to look at the Next branch before Linus accept the requests in those cases. It insanely rare that Linus is pulling anything into his branch that has not been sent to Linux-next first. Remember Linus does not perform the merges into linux-next there is a different person who has to be asked for the merge into Linux-next.

Yes huge amount of code get pulled in at times by Linus into his branch quite late in the development cycle that no human could have reviewed that fast but since it been though linux-next fist its had automated tool review.

There is a problem this time bcachefs was not submitted to Linux-next and I cannot find where an attempted merge request to vfs ,maintainer and so on.

[Mr.Elendig]

You moving the goalpost, the claim was not that linus is anoyed and overworked and that's why he should not do it, the claim was that if he overrules the decisions his "generals" did then they would feel overgone and there could come a divide, not that this makes a president and linux would be to overworked.

Thank you for selective quoting...

[oiaohm]

It looks like Kent won't be resubmitting his patches any time soon, which is a shame.

Still, reading through all the messages, I see him far more at fault for this than the rest of the developers.

There is one problem here the historic bad language with nasty streak personality Linus Torvalds is the same I have overworked myself disaster. The same thing Kent is worried about doing people he is close to. Overloading Linus risks doing that all Linux kernel developers.

Linus Torvalds had to take 3+ months off from kernel development at one point to recenter himself. If Kent has to take 3 months+ off to recenter himself and loss I want to rush the process problem that perfectly fine.

Yes might annoy us a little that bcachefs will take longer to get to mainline. This is another problem people need to learn the process to mainline into Linux kernel at times can take years to decades. Think about the RT set of patches.

Saying "patience is a virtue" is very true for Linux kernel developers.

There is a problem on our side not developing these patches is a lot of people put pressure on developers like Kent to mainline something as soon as possible when it should be yes we would like it mainlined but don't think you have to attempt it before the code is ready and please don't skip any steps in the mainline process as that will only cause a backlash problem..

[Khrundel]

This might be a stupid question but why can't Kent just submit it to linux-next for a cycle or two? I know that'd be a bummer but to keep pushing and pushing for it to be mainlined for years without even trying (hard) to get it ito linux-next seems kind of dumb. I mean, IIRC the first merge attempt was in 2018.

According to later thread, there were some kind of ambiguity, someone suggested him to go ask Linus, and he have asked Linus this and Linus just ignores this question, because it looks "obvious".

Why can't he just submit? Well because it became "kafkaesque". When people are giving one excuse after another and end up with "it is better for you not to show up for another decade", you would start to question if this is a subtle hint. Maybe nobody want your filesystem in kernel at all. Last month main concern was who will maintain this code in case Kent will be hit by bus, now, after he found some comaintainer this linux-next problem reappeared out of nowhere.
And I can understand kernel team's point: they do not want 2 kinds of btrfs to support and it is easier for them to not to accept new one then deprecate old one. Maybe I would prefer to btrfs to steal bcachefs features too, but it seems they've already bitten more than could chew.

[cb88]

Lets be clear... if he was backed by "insert random corporation with 100+ employees" and followed the normal drone of getting code into the kernel... he'd already have it merged.

There are several things blocking it.
Acutal issues (which he is resolving).
Being a single individual. (arguably shouldn't be an issue but it is).
Doing things that trip up merge... because he hasn't really done this before, eg, should have merged with linux-next first.... frankly I can understand why he'd skip it, he's been maintaining this out of tree for YEARS, its not NEW code and it HAS seen lots of review, that said going for linux-next first would have lowered his radar profile.