Bcachefs Changes Rejected Reportedly Due To CoC, Kernel Future "Uncertain"

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • mrg666
    Senior Member
    • Mar 2023
    • 1090

    Originally posted by browseria View Post

    But that doesn't address the issue. The issue is the kernel is a system, the FS system is dependent upon the MM and Block systems to be correct and error free. You can't just isolate it to just what Kent is doing. The other parts of the kernel have to be up to par as well. And they are not. Which is what started the issue in the first place.
    MM and block systems are just fine. There are fixes and improvements all the time as in all software but Linux kernel is solid and reliable. The problem is isolated to Kent's attitude 100%. Either deal with Linus and COC professionally or get out. That is what I suggested. Do not spread the problem to smear all of the Linux project. This is on Kent and only on Kent.

    Comment

    • browseria
      Senior Member
      • Apr 2018
      • 155

      Originally posted by Nocifer View Post
      If you want to mess with critical parts of the kernel that you deem "subpar", you first have to 1) prove that they're subpar to your colleagues, and 2) play by the rules, i.e. you have to submit your patches early enough so that said colleagues have the chance to review them and voice their own opinion on the validity of your assumptions.

      This is how a meritocracy is supposed to work: experts debating on an equal basis while respecting one another.

      Kent has repeatedly and intentionally failed on both of these accounts, and on top of that he's now also started treating his colleagues with disdain and even resorting to calling them names whenever they don't immediately give him exactly what he wants.

      That is what started this issue in the first place.
      What you are describing is not what happened in _this_ thread. In _this_ case, I believe Kent believes he made a good argument as to why the status quo was subpar. He got back NOTHING (or very little public) in response. He followed the rules, and his PR was rejected. In discussing the rejection, things got a little heated and he said some things he shouldn't have. So did the other side. They made up and then the CoC came down like a hammer.

      Here's the clue to who is at fault: the first rule when you have a losing argument is to attack the messenger and not the message. All I see in this thread is people attacking Kent for this or that perceived grievance, but I don't see anyone talking about the substance of his argument. Why is that?

      Comment

      • cb88
        Senior Member
        • Jan 2009
        • 1347

        Linus needs to be able to say shut the fuck up again.

        Comment

        • cb88
          Senior Member
          • Jan 2009
          • 1347

          Originally posted by nadir View Post

          Not the first, but maybe the 3795th time.

          He has made exactly zero positive contributions to the space and is widely known for it. Kicking him out would be an improvement, but then he'd make *another* account, just as he did before.
          What are on on about, he literally wrote the wildly used bcache.... which is VERY widely used. It's probably saved the world tens of millions of dollars across all the instances of its use at least easy.

          Comment

          • skeevy420
            Senior Member
            • May 2017
            • 8670

            Originally posted by TheMightyBuzzard View Post

            You should be, yes. The CoC idiots are going to side with a technically incorrect assertion because the correct one hurt the person who was wrong's feelings.
            No one's feelings were hurt. Kent said things against the rules like "CoC idiots".

            I just realized that you have a Bird Name...that explains a lot.

            Comment

            • TheMightyBuzzard
              Senior Member
              • Sep 2021
              • 416

              Originally posted by npwx View Post

              While I don't think he's "the smartest guy in the room" or that this is even relevant for being a good programmer, you can witness live how that produces the "best results".
              IOW, this is complete nonsense.
              Obviously you're never going to have that problem. Him not being the smartest person in that particular room was kind of the point. But it being a common problem for you is absolutely something required to be a good programmer. If you aren't at least tying for it regularly, you are an average or crappy programmer relative to that room.

              Comment

              • cb88
                Senior Member
                • Jan 2009
                • 1347

                Originally posted by mrg666 View Post
                The problem is isolated to Kent's attitude 100%.
                No code or social problem at this scale exists in isolation. That is just a false argument you are trying to use to further your opinion...

                Comment

                • browseria
                  Senior Member
                  • Apr 2018
                  • 155

                  Originally posted by mrg666 View Post

                  MM and block systems are just fine. There are fixes and improvements all the time as in all software but Linux kernel is solid and reliable. The problem is isolated to Kent's attitude 100%. Either deal with Linus and COC professionally or get out. That is what I suggested. Do not spread the problem to smear all of the Linux project. This is on Kent and only on Kent.
                  MM and block systems are NOT fine. They are software, and all software is imperfect. Does Kent have an attitude? Yes. And so does Michal and Shuah. So what? How does the attitude affect whether or not there is a hole in the MM api? Pretending there aren't bugs anywhere in the kernel, well, that is pretty much indefensible. It's how you go about resolving that is what is at stake here. So why all the personal attacks? Where is the discussion of the problem?

                  Comment

                  • TheMightyBuzzard
                    Senior Member
                    • Sep 2021
                    • 416

                    Originally posted by skeevy420 View Post

                    No one's feelings were hurt. Kent said things against the rules like "CoC idiots".

                    I just realized that you have a Bird Name...that explains a lot.
                    If nobody's feelings were hurt, it's a counterproductive rule even among a bunch of fragile whiners.

                    Comment

                    • Old Grouch
                      Senior Member
                      • Apr 2020
                      • 696

                      Dressing sentiments up in nice words doesn't change the sentiments.

                      A 'public apology' would be paying lip-service to the CoC, which diminishes all concerned.

                      Most people do not understand the passion in discussions/debates/arguments between technical experts in their fields. The key thing is being able to accept that you are wrong when it is demonstrated to you when you are wrong. The name-calling is being passionate about what you believe in, but your technical arguments will be tested by fire, and you always have to be able to justify your position: there is no deference to past accomplishments. Kernel developers work hard to get to the point where they can argue at a peer-to-peer level.
                      Referring your peers to previous discussions without giving links is plain rude. It is actually dismissing your peer by valuing their time less than your own. That is being egotistical.

                      Unfortunately, very smart people have a tendency to believe that administrative rules do not apply to them (administrivia), so for some kernel developers, following the rules is simply not within their experience. It's hard to change that behaviour, but that is probably Kent's major failing.

                      If people have come to an agreement now, I don't think it is helpful for a CoC Committee to insist upon a public apology. Don't hold grudges, but reward good behaviour.

                      If Kent's goal is to have BcacheFS a well-used fileysstem, then I hope he is intelligent enough to realise that that goal is likely to be compromised if he does not fit in better with Linus and the rest of the kernel developer community. As an intellectual exercise, he can plot out a strategy that has the greatest chance of BcacheFS being widely adopted. I would be surprised if his current behaviours are the best approach, and I would be even more surprised if he can't see this. Then again, even smart people get tunnel vision sometimes.

                      I hope BcacheFS makes it. If it doesn't, then I hope future FS developers will reflect upon why it didn't, and modify their behaviour accordingly.

                      Comment

                      Working...
                      X