Comments by "" (@diadetediotedio6918) on "New Linux Kernel (It's Starting To Get Rusty)" video.
-
5
-
4
-
3
-
2
-
2
-
@anon_y_mousse
First, I was responding to your comment about novice C programmers being wiser than Rust ones, and responding in general as well.
Next, I didn't make any pointless arguments, the fact that you have little knowledge about economics does not make you oblivious to the consequences of this lack of knowledge, if a company the size of Microsoft (a literal corporation) and with the amount of information that she has (which is quite broad) is saying that most of her security bugs came from memory errors (and she is saying that), so we very likely have good reason to think prima facie that these languages create a favorable environment for this kind of bug. A lot of the linux kernel bugs are due to things like that too, many of which could have been avoided with a safer tool like Rust (or not, after all, we have to test things before we know they will work).
The argument that the creator of linux would be considering adopting the language in his dear kernel just because of "pressure" is completely irrational, that yes, it's just a cheap complaint and you're disregarding all the informational scope that Linus himself should already have considered when accepting to include the language in the development. He's not stupid, he's not doing things by simple pressure, that didn't even happen with C++, which has one of the most toxic communities I've ever seen in my life, and it wouldn't happen with Rust.
The point is, we use magnetic screwdrivers to avoid the mistake of dropping a screw, yet experienced people rarely drop a screw using an ordinary screwdriver. We use cars with smart sensors to avoid crashing, yet good drivers are unlikely to crash using a car without these sensors. We consider taking precautionary measures on our own future actions with insurance and the like, even though we are very confident that we will not go wrong, and we take precautions even though we are experts in our fields. Science has checks because even the greatest investigators can make obvious and crass mistakes. We use safety equipment even though good workers will probably never have an accident. Even the best doctors use better equipment even if the old equipment did the trick, just to lessen the possibility of an error.
There is absolutely no reason to believe that a programmer, even the greatest and most experienced programmer in the entire world (who, by the way, won't live forever), won't make mistakes, that's innocence, that actually borders on complete insanity, and using a tool that reduces the propensity for silly mistakes (and even serious mistakes) to occur is just a natural step in technological evolution.
And make no mistake, maybe Rust is not that tool that will prevent serious security errors from happening, maybe the language doesn't work for that and everyone who says it works is completely wrong, who knows? The idea is that if we never try something new, we will forever make the same mistakes. If Rust doesn't work, let's build something better and test it again, if that doesn't work, again, and again, we'll keep going, until we've found a way to effectively make working people's lives safer, even if for a small margin, that's progress.
2
-
2
-
2
-
1
-
@thewitheredstriker
This is because nowadays the amount of low-level programming is infinitely less than the amount of high-level programming, which wasn't exactly like that at the time (and let's face it, C didn't last that long, most companies in the 2000s had a good part of their plastered systems made in Java and languages other than C, even banking systems used COBOL and not C, it's not like C The programming language), not to mention that in that era most of the systems we use today were being written.
Nowadays for a language with the ambitions of rust to prevail and become mainstream, it first needs to be effectively adopted, which effectively won't happen if we stop using it because it's not established (that would be a vicious circle), re-writing systems from scratch is simply impossible, the cost is gigantic and would be impractical for the overwhelming majority of corporations, hiring new programmers too, and even training existing ones would be costly, which is precisely why adoption is not so fast ( the fact that Rust has a relatively high learning curve doesn't help with that). In terms of stability, on the other hand, the language appears to have few bugs and it doesn't seem to be the case that it is static, it has an active community and people are willing to use it, think about what Linux itself would have been like if people never had given it a chance in the first place?
At the end of the day, it all comes down to cost, and it seems big companies are starting to look at Rust as more cost-effective than not using it (we're seeing movement in big companies like Amazon, and Microsoft itself recently, and even Dropbox has a significant portion of its systems written in the language), every new technology not established is a gamble until it isn't, which is why the world moves slowly (e.g. we still don't have IPv6 in, I would say, more than 50% of places in my country).
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1
-
1