4 ms·
> Yes, now I am implying something: you won't bother to rewrite the utilities. > So we cannot replace any base utility, unless the toolchain to build it is in
by 013a 6y ago
> Yes, now I am implying something: you won't bother to rewrite the utilities.
> So we cannot replace any base utility, unless the toolchain to build it is in the base. Adding such a toolchain would take make build time from 40 minutes to hours. I don't see how that would happen.
So, to summarize: "We're not going to include the capability to write core utils in Rust, and also, no one is bothering to do it."
Its rather startling to me how tail-wagging-the-dog, or maybe chicken-and-egg, this situation is. Did anyone stop to consider that the reason why some of these utilities aren't being rewritten is because the unix old guard doesn't want to support it in the toolchain? If this guy is the wall that you'll eventually have to crash your car into, I'm not going to get behind the wheel in the first place. Why waste my time?
Look; the compile limitations are real. I'm not saying that it should be supported. But, attitudes like the one are not productive. What is it about operating system devs and these aggressive, mean, anti-social, self-important personalities? A response worded like this should not be tolerated by the community. It would have been just as easy and clear to say "We don't have toolchain support for Rust, and we're really not interested in supporting it due to its compile time and architecture support issues. <One sentence explaining compile time issues>. <One sentence explaining architecture support issues>. We can revisit this discussion in the future if/when these improve."
- quotemstr 6y agoI got the same vibe from the email. It reads just like bigcorp turf defense. When a smart but curmudgeonly component owner just doesn't want to accept some outsider's change, there's no end to the technical objections that component owner can raise against the outsider's change --- but these objections are just a smokescreen, and addressing them is futile, because the real problem is that the component owner has done a thing certain way for a long time and some attempt by an outsider to shake things up triggers all the owner's ancient territorial defense instincts. The actual merit of the outsider's proposal doesn't matter: the problem is social, not technical. In a situation like this, when the intrepid outsider does manage to exhaust the maintainer's objections, the maintainer will switch to just spewing unfalsifiable FUD, to inventing further ridiculous objections, or just simply ignoring the outsider's change entirely. There's no technical fix for a curmudgeonly maintainer who really doesn't want to entertain new ideas. In the corporate world, to resolve this problem, you usually have to go up the management chain and explain to leadership that certain people are making important work impossible. (You should try every possible alternative means of persuasion first, of course.) In the FOSS world, you just have to fork as a last resort.
- alqx 6y ago> In the corporate world, to resolve this problem, you usually have to go up the management chain and emphasize to leadership that certain people are making important work impossible. In the FOSS world, you just have to fork. Typical Machiavellian talker who overestimates his own abilities, derails a working project and moves on to the next company after two years to continue the damage there.
- jacquesm 6y agoWhat a load of bullshit. Theo has had more staying power than the Rust community collectively to date as a maintainer of a mission critical OS. That's a lot of responsibility and I personally very much value his defense of what matters a lot to me: backwards compatibility. Not breaking stuff is the prime directive at that level. Suggesting that your rewritten, non compliant, non-drop in replacements and associated toolchain should be part of the default is hubris at a level that does not befit nor benefit the Rust community. If you want to reach that level then you need to commit for real, and solve those issues that Theo rightfully marks. Even though this was written in 2017 it is just as applicable today: either you replacement is as good on all fronts as the thing it replaces and has some additional benefits replacement of something battle tested makes no sense.
- 013a 6y agoThat really wasn't the vibe I was getting. There probably are legitimate technical issues with integrating Rust. That's fine; it happens. Maybe Rust will be better in the future (this was written in 2017, so maybe it already is). Or, maybe Rust will never be a good candidate for BSD. Any of these outcomes are possible. I take issue with his tone, which is strongly reminiscent of Linus' management of Linux. He's welcome to be mean, but I would think long and hard about joining a community with a leader with such a disposition (I don't know Theo's history, so I can't say whether this is a pattern). Being kind when running a project, especially of the scope of an operating system/kernel, is very difficult. Not everyone has the disposition to do it. You'll get Rust-fiends and Go-phers and all sorts of people who haven't contributed a single line come in and make wild suggestions that don't make sense, every day. You'll get HackerNews armchair commentators like me who think their opinion matters. In the face of this, its understandable to respond with scorn. But; enviable leaders instead form measured responses. You never make it personal. You never shut the door on suggestions. Open Source, especially at BSD's scale, is not a one man army. Thousands of people have helped with that project, and its creator no longer has the blank check to be a dictator. An excerpt from Bezos' 2010 Princeton commencement address [1]: > I decided to do the math for my grandmother. I estimated the number of cigarettes per days, estimated the number of puffs per cigarette and so on. When I was satisfied that I’d come up with a reasonable number, I poked my head into the front of the car, tapped my grandmother on the shoulder and proudly proclaimed, “At two minutes per puff, you’ve taken nine years off your life!” > I have a vivid memory of what happened next and it was not what I expected. I expected to be applauded for my cleverness and arithmetic skills. “Jeff, you’re so smart. You had to have made some tricky estimates, figure out the number of minutes in a year and do some division.” That’s not what happened. Instead, my grandmother burst into tears. I sat in the backseat and did not know what to do. While my grandmother sat crying, my grandfather, who had been driving in silence, pulled over onto the shoulder of the highway. He got out of the car and came around and opened my door and waited for me to follow. Was I in trouble? My grandfather was a highly intelligent, quiet man. He had never said a harsh word to me, and maybe this was to be the first time? Or maybe he would ask that I get back in the car and apologize to my grandmother. I had no experience in this realm with my grandparents and no way to gauge what the consequences might be. We stopped beside the trailer. My grandfather looked at me and after a bit of silence, he gently and calmly said, “Jeff, one day you’ll understand that it’s harder to be kind than clever.” [1] https://www.theladders.com/career-advice/jeff-bezos-to-princetons-class-of-2010-its-harder-to-be-kind-than-clever https://www.theladders.com/career-advice/jeff-bezos-to-princ...
- zerpelin68 6y agoCensorship of opinionated entities speaking in a manner you disagree with is so popular right now how can you be wrong? The 'community' meaning my cloud/bubble/echo chamber as usual I guess? Theos take was refreshing honesty in my bubble.
- 013a 6y agoHe's allowed to say whatever he wants; I'm allowed to believe that its unprofessional for a community member of his stature; and I'm allowed to decide whether I would feel comfortable contributing to a project ran by someone like that. None of this is censorship.