Rust 1.27 Released With SIMD Improvements

[jacob]

Or like how falling off a bike when you're a total rookie is a problem that wouldn't exist with a training wheel right? C programmers don't need hand-holding or a third wheel dragging their performance behind.

It's not even only about performance, for real programmers, it's just shameful to rely on a training wheel, you know.

That's a profoundly ignorant comment. Type soundness and (partially) provable memory soundness don't negatively impact performance, quite to the contrary in fact. As for "real programmers", I'm sorry but software is engineering like any other and fulfilling the l33t pride of some teenager is emphatically not one of its goals. Aircraft designers are expected to do more than just make pride statements, they must follow rigorous design methods that involve PROVING (not just claiming) that certain criteria are met. Good programmers know this and they are sufficiently knowledgeable in computer science (not just coding) to understand the problems and the theoretically correct solutions that are available. Kids and incompetent idiots play "real programmers" instead because they don't know any better.

[Weasel]

That's a profoundly ignorant comment. Type soundness and (partially) provable memory soundness don't negatively impact performance, quite to the contrary in fact. As for "real programmers", I'm sorry but software is engineering like any other and fulfilling the l33t pride of some teenager is emphatically not one of its goals. Aircraft designers are expected to do more than just make pride statements, they must follow rigorous design methods that involve PROVING (not just claiming) that certain criteria are met. Good programmers know this and they are sufficiently knowledgeable in computer science (not just coding) to understand the problems and the theoretically correct solutions that are available. Kids and incompetent idiots play "real programmers" instead because they don't know any better.

How about you look it up?

https://en.wikipedia.org/wiki/Real_P...27t_Use_Pascal

Quote: Since then, the computer folklore term Real Programmer has come to describe the archetypical "hardcore" programmer who eschews the modern languages and tools of the day in favour of more direct and efficient solutions—closer to the hardware.^[4]

And you were talking about ignorance? My "Real programmer" term is not an opinion, it's a known term. So keep your bullshit to yourself.

Also related: https://en.wikipedia.org/wiki/Hacker_culture


Terms like "software engineer" and "software developer" are used by those incapable of calling themselves programmers, but they want to still sound cool and important.

90% of those who do so only do it for a paycheck as part of job, that's as far from a real programmer (or prodigy) as you can get, nobody gives a shit of such "software developers", and I mean literally, they are replaceable, expendable and unimportant (nobody even knows their names, unlike the famous prodigies). They write code so that morons can understand, so why would they be special if a MORON CAN UNDERSTAND?!??

Real programmers think about code all the time not forget about it as soon as their job is over. Paycheck is nice but code is most important. They invent "clever" algorithms that the "software developers" find a nightmare to understand (because they suck) and either use a library without understanding shit about it or prefer idiot-proof "maintanable" code (for a monkey) even if it's 5x times more bloated or less efficient.

"Software developers" plug a few library calls (incl. standard library) together, see "it works" and think their job is done. They get paid after all, that's what matters right? How can this be a "real" programmer? Sounds more like a "real salary man" to me. This is the 90% or more of the trash code in existence. And users are the ones who suffer at the poor code -- users who really don't give a shit if Joe Smith got his paycheck for the bad code he did, and neither do they gain anything from it. Worse than that is code that is outsourced (cheap) to certain countries, you should know what I'm talking about.

For a real programmer, programming and the output code is the most important thing, anything else is secondary.



And now to the context at hand: most C++ code is garbage, and even C++ haters admit it. That is, the code is garbage, not necessarily the language. Code where you see almost every god damn line a call to a library or using some standard library construct. You know this guy is incapable of inventing anything on his own. He's a code plumber, not a programmer. Standard library-ridden C++ code is just as piss poor as Rust is, but you don't need "unsafe" blocks in C++ to use it properly. In fact, that's the default.

[jacob]

How about you look it up?

https://en.wikipedia.org/wiki/Real_P...27t_Use_Pascal

Quote: Since then, the computer folklore term Real Programmer has come to describe the archetypical "hardcore" programmer who eschews the modern languages and tools of the day in favour of more direct and efficient solutions—closer to the hardware.^[4]

And you were talking about ignorance? My "Real programmer" term is not an opinion, it's a known term. So keep your bullshit to yourself.

Also related: https://en.wikipedia.org/wiki/Hacker_culture


Terms like "software engineer" and "software developer" are used by those incapable of calling themselves programmers, but they want to still sound cool and important.

90% of those who do so only do it for a paycheck as part of job, that's as far from a real programmer (or prodigy) as you can get, nobody gives a shit of such "software developers", and I mean literally, they are replaceable, expendable and unimportant (nobody even knows their names, unlike the famous prodigies). They write code so that morons can understand, so why would they be special if a MORON CAN UNDERSTAND?!??

Real programmers think about code all the time not forget about it as soon as their job is over. Paycheck is nice but code is most important. They invent "clever" algorithms that the "software developers" find a nightmare to understand (because they suck) and either use a library without understanding shit about it or prefer idiot-proof "maintanable" code (for a monkey) even if it's 5x times more bloated or less efficient.

"Software developers" plug a few library calls (incl. standard library) together, see "it works" and think their job is done. They get paid after all, that's what matters right? How can this be a "real" programmer? Sounds more like a "real salary man" to me. This is the 90% or more of the trash code in existence. And users are the ones who suffer at the poor code -- users who really don't give a shit if Joe Smith got his paycheck for the bad code he did, and neither do they gain anything from it. Worse than that is code that is outsourced (cheap) to certain countries, you should know what I'm talking about.

For a real programmer, programming and the output code is the most important thing, anything else is secondary.



And now to the context at hand: most C++ code is garbage, and even C++ haters admit it. That is, the code is garbage, not necessarily the language. Code where you see almost every god damn line a call to a library or using some standard library construct. You know this guy is incapable of inventing anything on his own. He's a code plumber, not a programmer. Standard library-ridden C++ code is just as piss poor as Rust is, but you don't need "unsafe" blocks in C++ to use it properly. In fact, that's the default.

I know perfectly well what a "Real Programmer" refers to, thank you very much, and I used the term with all the mockery and disdain I feel towards it. The idea that "hardcore" programming (e.g. eschewing rigorous modelling and advanced verification methods) is somehow more "efficient" is pure ignorance. It's often only claimed by those who don't know any better and take bizarre pride in writing brittle code that only works by accident. People whom I genuinely respect as programmers, say Linus Torvalds for example, are as removed from your definition of "Real Programmer" as it gets. Linus likes his C because it is, or at least was until very recently, the only language other than assembly sufficiently low level to write to raw hardware, which is what he does, but he's famously stringent in the discipline and design of his code. In one of his memorable rants he actually publicly humiliated one "real programmer" by pointing out how ignorant he was to not understand the place and benefits of more advanced languages. He says he likes Haskell, but it's obviously not something to be used for kernel development. So sorry but I'm not sorry, I will keep laughing at "real programmers" and reserve my respect to "real computer scientists" who understand the theory of computation, type theory and static analysis methods, and are able to put them to good use.

[Falcon1]

It goes both ways - this very ownership model makes ordinary code much more difficult to write and anywhere you have parent-child relationship you have to use a special library because the ownership model has this flaw that it's impossible to express parent-child relationships. Add lifecycle annotations and you get a Rust code that is a lot harder to write and interpret. That's why I switched back to C++. Rust will never be as productive as C++.

From:
https://www.quora.com/Which-language...i-Alexandrescu

I think that's a matter of perspective. First: It saves you time in testing and debugging. So yes more time spent in programming but less in testing and debugging. Second: As with OO programming you need to get used to the paradigma. You're simple forced to think about variable lifetime and "ownership". Once you get the hang of it you'll get faster.

[brrrrttttt]

You're always forced to think about that if you want to write correct code...