4 ms·
I agree with how he's saying it. He's being honest. He's letting you know where he stands and what the rule is, without any kind of hidden message or room for i
by kadenshep 9y ago
I agree with how he's saying it. He's being honest. He's letting you know where he stands and what the rule is, without any kind of hidden message or room for interpretation. The other guy? Not being so honest.
- lightedman 9y agoHe's being stupid, using words he doesn't understand (He already failed in the usage of the word 'regression.') He's throwing a childish tantrum and THAT is the primary reason I won't touch Linux (well that and I can hardlock any Linux system by making a 4kb encrypted RAMDisk, and it's been left unfixed for years, allowing me to find dozens of nice memory issues, [especially with the buffer cache,] and most of these issues are ELEMENTARY-GRADE MISTAKES, which matches very well with Linus' elementary school attitude.)
- ryanx435 9y agoI'm willing to bet my entire life savings that Linus understands the meaning of the word regression in this context better than you.
- lightedman 9y agoHaving read the entire mailing list thread, no, he is NOT using the word correctly. I suggest you read the entire thread. Regression in the case of code indicates this problem was there before, fixed, and resurfaced. Simply breaking user space is not a regression. It is simply breaking things and nothing more unless this problem occurred once before.
- newen 9y agoAs it's used almost everywhere in coding circles, regression means a patch that made things worse than it was before. No implication of the worse state being there before.
- kadenshep 9y agoAre you able to provide a patch for the aforementioned elementary grade mistakes to the open source kernel project without introducing any kind of regression in user space?
- lightedman 9y agoThat's not my job. I report bugs. I'm not getting paid to do that so doing so in my free time to be ignored is a waste of my productivity. I reported the bare-basic problem inherent in the buffer cache which lead to my kernel hardlock exploit and was summarily ignored as non-critical. So I've done my job. It's on Linus now to eat crow and get shit done. That's obviously not happening with his attitude.
- kadenshep 9y agoBut they're elementary bugs you could have just fixed for yourself. You clearly were trying to do something and the mistakes that were made were easily spotted, I'm confused as to why you just didn't do it at the very least for yourself? >That's obviously not happening with his attitude. Seems he's making a lot of things happen with his attitude as evidenced by the thread that was linked.
- lightedman 9y ago"But they're elementary bugs you could have just fixed for yourself" I'm not selfish enough to want something to work for me and only me, thank you. That's why I submitted a bug report. And that ends up just providing another reason why Linux is a crap OS - everything is done entirely backwards as if everyone on the team learned their entire education the same way RPN is written.
- sctb 9y agoWe've asked you not to post like this before, and it seems as though you're going to continue to violate the guidelines, so we've banned the account. We're happy to unban accounts if you email us at hn@ycombinator.com and we believe you'll post civilly and substantively in the future.
- forapurpose 9y agoThere's an implicit idea around our culture that the more uncontrolled emotional content that someone communicates, especially anger, the more 'honest' they are - I think because it's intimidating, leaving less room for debate; it makes reasoning very difficult, leaving even less room for a rational approach; it's message is of being uncompromising, which also cuts off any thinking or discussion. By eliminating thinking and discussion, it's simple, and I don't mean that in a good way, but that makes it appealing to people too lazy to deal with the necessary complexity of situations and who are not committed to getting the best outcomes. But it's not honest at all and it's not hard to see how it leads to bad results. Acting out is actually dishonest and it's mostly about a hidden message. It's a failure to recognize and deal with your won frustrations which you then take out on others, usually when an opportunity presents itself to rationalize your behavior - someone does something you perceive as wrong, so they become a justifiable target for the anger that's already there. What I read in the email's message is, 'I'm out of control, I'm not able or willing to mange my emotions, so you better watch out'. That is a poor management and communication technique. IMHO, a leader or manager should hold their leadership and management work product to the same standard that they hold their subordinates' technical work product. If they want the highest standards for the kernel code, they need to have the highest standards for their communication and other conduct. A great challenge of being the boss is that there is nobody to hold you to high standards; you have to do it yourself and that's not something people are good at. As they say: Power corrupts, and absolute power corrupts absolutely. Clearly, this message could have been communicated much more effectively - while being truly honest and without hidden messages or room for interpretation - if the communicator chose to make an effort. Finally, if this form of communication is effective, do others communicate to Linus in this manner? Does he ask them to?
- AnthonBerg 9y agoI will never understand why this has been downvoted.
- forgottenpass 9y agoIt's a cliche hacker news comment that adds nothing to the conversation: - Speaks in platitudes about the topic in a way that shows a thin understanding of the context and only sounds smart to people that also aren't familiar with the context. - Armchair psychiatry used to advance a moralizing message. - All over the place while deep in a comment thread, seemingly more interested in hearing themselves talk than carrying on the conversation.