Re: [GIT PULL] bcachefs changes for 6.17

["Malte Schröder" <malte.schroeder@tnxip.de>]

On 11.08.25 20:13, Carl E. Thompson wrote:
> I seriously hope none of the kernel developers are foolish enough to be fooled (yet again) by this I'm-a-reasonable-guy-we-can-talk-this-out act. You've been there and done that.
>
> Kent's perplexing behavior almost makes me want to put on a tinfoil hat. Is it simply mental illness or is it something more? Is he being egged on by backers who *want* to destabilize the leadership of Linux for whatever reason? It's hard to see how any individual could be this far out there without help.
>
> And I'll point out what's obvious to people who have followed this closely but may not be to people who read an occasional email thread like this one: A very large portion of what Kent says including in this email is just factually wrong. Either he is an unashamed and extremely prolific liar or he is very sick.
>
> Carl

Frankly for me as a user who does probably not know the hole picture you
seem to just be spewing paranoid hate into these threads which I do not
quite understand. Yes, Kent can be off-putting, but really, that is
something I often observe from other people as well, the tone on LKML
tends to be pretty harsh.

My involvement in bcachefs is that of an early adopter and hence, a
tester. I find working with him productive. I give him bug reports,
observations and data and he fixes those. If I did something stupid, he
will point that out, and very directly so.

So what you call a "I'm-a-reasonable-guy-we-can-talk-this-out act" is
actually how he can and does behave, at least when interacting on IRC
with his users, testers and co-developers.

So maybe if we can dial the personal attacks down a few notches
(including against btrfs, though I have my reasons why I jumped from
it to bcachefs like two years ago) and have a calm discussion it might
be possible to build new bridges?


/Malte

>
>> On 2025-08-11 7:26 AM PDT Kent Overstreet <kent.overstreet@linux.dev> wrote:
>>
>>  
>> On Mon, Aug 11, 2025 at 12:51:11PM +0300, Konstantin Shelekhin wrote:
>>>>  Yes, this is accurate. I've been getting entirely too many emails from Linus about
>>>> how pissed off everyone is, completely absent of details - or anything engineering
>>>> related, for that matter.
>>> That's because this is not an engineering problem, it's a communication problem. You just piss
>>> people off for no good reason. Then people get tired of dealing with you and now we're here,
>>> with Linus thinking about `git rm -rf fs/bcachesfs`. Will your users be happy? Probably not.
>>> Will your sponsors be happy? Probably not either. Then why are you keep doing this?
>>>
>>> If you really want to change the way things work go see a therapist. A competent enough doctor
>>> probably can fix all that in a couple of months.
>> Konstantin, please tell me what you're basing this on.
>>
>> The claims I've been hearing have simply lacked any kind of specifics;
>> if there's people I'd pissed off for no reason, I would've been happy to
>> apologize, but I'm not aware of the incidences you're claiming - not
>> within a year or more; I have made real efforts to tone things down.
>>
>> On the other hand, for the only incidences I can remotely refer to in
>> the past year and a half, there has been:
>>
>> - the mm developer who started outright swearing at me on IRC in a
>>   discussion about assertions
>> - the block layer developer who went on a four email rant where he,
>>   charitably, misread the spec or the patchset or both; all this over a
>>   patch to simply bring a warning in line with the actual NVME and SCSI
>>   specs.
>> - and reference to an incident at LSF, but the only noteworthy event
>>   that I can recall at the last LSF (a year and a half ago) was where a
>>   filesystem developer chased a Rust developer out of the community.
>>
>> So: what am I supposed to make of all this?
>>
>> To an outsider, I don't think any of this looks like a reasonable or
>> measured response, or professional behaviour. The problems with toxic
>> behaviour have been around long before I was prominent, and they're
>> still in evidence.
>>
>> It is not reasonable or professional to jump from professional criticism
>> of code and work to personal attacks: it is our job to be critical of
>> our own and each other's code, and while that may bring up strong
>> feelings when we feel our work is attacked, that does not mean that it
>> is appropriate to lash out.
>>
>> We have to separate the professional criticism from the personal.
>>
>> It's also not reasonable or professional to always escelate tensions,
>> always look for the upper hand, and never de-escalate.
>>
>> As a reminder, this all stems from a single patch, purely internal to
>> fs/bcachefs/, that was a critical, data integrity hotfix.
>>
>> There has been a real pattern of hyper reactive, dramatic responses to
>> bugfixes in the bcachefs pull requests, all the way up to full blown
>> repeated threats of removing it from the kernel, and it's been toxic.
>>
>> And it's happening again, complete with full blown rants right off the
>> bat in the private maintainer thread about not trusting my work (and I
>> have provided data and comparisons with btrfs specifically to rebut
>> that), all the way to "everyone hates you and you need therapy". That is
>> not reasonable or constructive.
>>
>> This specific thread was in response to Linus saying that bcachefs was
>> imminently going to be git rm -rf'd, "or else", again with zero details
>> on that or else or anything that would make it actionable.
>>
>> Look, I'm always happy to sit down, have a beer, talk things out, and
>> listen.
>>
>> If there's people I have legitimately pissed off (and I do not include
>> anyone who starts swearing at me in a technical discussion) - let me
>> know, I'll listen. I'm not unapproachable, I'm not going to bite your
>> head off.
>>
>> I've mended fences with people in the past; there were people I thought
>> I'd be odds with forever, but all it really takes is just talking. Say
>> what it is that you feel has affected, be willing to listen and turn,
>> and it gets better.