Comments by "" (@diadetediotedio6918) on "Rust Is Coming to The Linux Kernel and I'm Not Sure How to Feel" video.

@jondoe6608 I think at some point you need to leave the old stuff behind, this strikes me as more a feeling of nostalgia than a rational thing, if C can't become as safe as Rust in 50+ years, he most likely won't become much safer now (mainly because the changes would have to be severe). Bringing in a new language has the benefit of not having to maintain compatibility with so many thousands of existing systems, it allows you to advance in places where you wouldn't be able to do with C alone, and frankly it feels like a lot more sticking with tape. problems than to actually solve.
14
My view is more to the contrary, I have never seen the rust community being toxic, I see them being incredibly boring and sometimes literal missionaries, but I can assure you that I have never seen a community more destructive than the C++ community (the C community is certainly much better which makes me have a lot more affection for C developers). In the end what should prevail is free software, and change should always be welcome to free software when it provides sufficient benefits.
7
You can do that, but creating a kernel is quite a task, I don't think many people would create a new kernel from scratch in literally any language, that would be almost like asking it to be born dead.
6
There is a lot of discussion that is caused by C++ devs, luckily or unfortunately I have become an absolute hater of this language over the years due to its ABSOLUTELY TOXIC community. The point is simple, people want to use the things they know and pour their hate on everything again that tries to improve the situation, this has been like this since the world is world and it won't change now, the thing is that we Sometimes we have to throw away old patterns and tools and use new ones as they match our needs, it's like the thing of using a magnetic screwdriver because it makes it harder to lose the screws, you can still do all that with a conventional screwdriver but why would you? C++ devs keep trying to push all these bandages and all this duct tape over the language and saying that this made the day, and that programmers should get used to it and that anyone who doesn't is a bad programmer, that's just a way crazy to think. Whether or not Rust is suitable for the kernel we'll find out soon enough, but people shouldn't stop using better technologies just because of grouchy old people full of resentment.
5
@Appley Systems This is just duct tape and saying "programmers must adapt", your solution is basically nut.
4
@autistadolinux5336 In fact, I didn't mean it. Traits are quite common in some languages and it's not a new concept at all, what's new about rust is its memory management, something I've never seen done by any language before it (I mean the way it's done) . Out of very few resources like this the language could be compared with literally any other, monads, immutability, discriminated unions, these are concepts present in dozens of languages out there, immutability by default and motion semantics (by default) is one of the things which makes me think it breaks a little bit, but not much, with C-based languages (just see for example the way most C-based languages simply copy the values during assignment, this is not a default behavior in Rust), which is why I don't think it has that much to do with C++.
3
@donaldmickunas8552 I think as long as we fight for free software it won't die, don't you? There are pitfalls in the "OSI model" that we unfortunately fall into over time, but with luck we can still keep it alive for years to come with enough effort
2
@autistadolinux5336 I wouldn't say it's C++ like in fact, I've had the "chance" to work with C++ for quite some time, and while I'd say Rust isn't close to C, it's not exactly close to C++ either. The truth is that Rust draws from many different sources, including a very strong influence from functional and multiparadigm languages (even Ruby gets into this), and while it has serious similarities to C++, the same could be said for most modern languages out there. In my view, Rust is much more of a functional language than an object oriented one for example, it simply has different ways of dealing with things (e.g. the notion of immutability by default), and in some ways I would say it is a language unique in itself. That said, I only use rust when necessary, I mean, you don't need to use a low-level language 90% of the time, C# or even JavaScript are quick and easy solutions to use. If I needed to make a small low-level project fast I would probably use C for the simplicity and tactileness of the language, and if I needed something more robust and secure I would use Rust, languages are just tools and we need to get the ones that serve us the most.
2
I am amazed that there are still people like you who criticize language without providing any evidence for these criticisms, when it has already demonstrated itself in the fields where it needed to demonstrate
2
@vernongrant4710 It's like, "Just never crash your car" or "just don't be sad", as if that's how things work and even the most capable people don't make mistakes.
2
@vernongrant4710 I have no idea, I just criticize foolish behavior, I don't necessarily defend Rust.
2
I personally find the term quite embarrassing, but I see the rust community as less of a cult than the C++ community for example. The "rustacians" are incredibly healthier, like friends who tell you not to drink, they can be annoying but they're not always wrong.
2
@anon_y_mousse This kind of thinking is absolutely destructive and irrational, just to be clear. Programmers have been making safety mistakes for as long as the world has been, that's not because they're "dingos" or any crap like that, the same way people get into accidents driving and using hammers and nails and all sorts of stuff, well, obviously these accidents would not have occurred if these people were careful what they did, but accidents are called that because they are accidental and they happen, so if you can avoid it by using a safer tool, it would be silly not to.
2
@Tachi107 I just read the article, the way you put it sounded a lot more "aggressive" than the article itself actually was. They were describing very specific situations of language deficiencies, something that I really believe should be taken into account and corrected, and they also pointed out solutions, including pointing out that there were contradictions in kernel developers about using unconventional solutions. I think you're making a big deal out of a teacup if that's your only source of criticism, but it's a really interesting article and it provides interesting insights indeed.
2
What are the issues with rust that could be some kind of impediment to its use here?
1
@autistadolinux5336 To me C++ seems like some kind of modern evil religion, and in fact arguing with C++ devs reminds me a lot of arguing with Java devs about how the language is "superior and cross-platform, and it's much better than C#", there is a great resistance to change in both communities that seems more like something coming from old people resentful of the world than a really rational thing about how technologies are just tools and we should use what suits us best. I mean, while some people are stuck with these concepts, good programmers should be able to see the beauty and usefulness in the most diverse languages for different situations. Just so you know, are you br?
1
@vernongrant4710 So why are you defending your position right now? Sounds a little contradictory, don't you think?
1
@anon_y_mousse So honey, that's it. If you weren't convinced by these extremely obvious things I said, what would convince you? Absolutely nothing I could say in a youtube comment. Your mind is plastered and full of resentment, and that's just it, see how you yourself assumed that people "would have no idea what they do" when all I said is that a better tool provides a better quality in your service , you are only defending your position ideologically. I can literally think of an ASM programmer advocating that C would be "a problem and that compiled code is not always predictive, that you need to master hardware programming at the lowest possible level and that if so you will never make mistakes" and that should not adopt it, just as I can think of a doctor refusing to use new safer, less brittle syringes because "a good doctor would never let the needle break in a patient's veins", or some pharmacist saying that the new methods of mixtures are for "inexperienced pharmacists, and that if you are an excellent pharmacist you will never need these new methods". At the end of the day it's just that, an irrational resistance to change. Thank God enough people didn't think like you and actively found new, better and safer ways of doing things, because if we depended on people like you humanity would still be hunting with stone tools, it's that simple.
1
@Raleighthrbub123 very fine pointing
1
The first thing I need to ask is, why do we need bitfields in this case?
1
@Tezla0 Lol, so why don't you just use C? If Linus absolutely needed bitfields (something I've read "talks" of him saying should be avoided whenever possible) he wouldn't have considered using Rust.
1
I don't think that would be a problem, C will probably never stop being part of the kernel at some point, and they will probably only use rust on the critical parts that don't require an extreme level of insecurity, and at the same time can benefit from management memory clean and a fearless concurrency
1
@Tachi107 And what is the memory stability problem that rust has?
1
@Tachi107 Don't worry, I meant it sounded aggressive, in the sense that it sounded a lot more provocative than it should have, but I can understand what you mean. Well, I personally think that Torvalds, being the creator of Linux and the one who names the kernel itself, has had enough time to consider all these issues, and frankly, I imagine this is going to be more of a test than an absolute implementation, then we will be able to see in practice whether or not these language problems will really prove to be insoluble problems, or if they will just be considerations to improve it over time, I believe it is insufficient to drop an entire language that has so much power, like this (there is a lot of more rust than borrow-checker, safe concurrency is one of its big selling points too, and that's quite interesting for a kernel). I think in terms of standardization this isn't a big problem either, I mean a good part of the reason why C/C++ are such "static" languages is because they aren't willing to change quickly to solve their problems, the rust developers have been very helpful over the years and the language has been making great strides (from the article itself, it seems to me that the current barrier to solving that question is much more a proper formalization of the problem and the solution than a unavailability of wanting to improve, which is extremely positive in this case). And as for low-level stuff needing to be implemented, that might be an unpopular opinion, but I also don't think that's a big issue, Rust is known for its stability and it seems to me that if it's possible to solve problems using this, it's probably a good solution (mainly in view of the control and fine-tuning that this brings). Don't get me wrong either, I've been a C# programmer much longer than I've been a rust programmer, and I also have my taste in C (an absolutely beautiful language indeed), and my dislikes with C++, I just try to see how many of the criticisms made of language are effectively criticisms of impediment and how many are, how to say, just a very immediate reaction. In the end, I'm hoping that the addition of the language will make the kernel more powerful and not worse at all (considering all that people have been saying about Rust, such as the recent statements by the Microsoft guy, and the fact that Torvalds himself has said that the language won't do any harm to the kernel, I believe it's a perfectly valid belief, these people wouldn't say that without reason).
1
It's the old question of resistance to any change, I don't blame you, I just don't see how to respect this irrational fear.
1