[ / / / / / / / / / / / / / ] [ dir / animu / arepa / asmr / ausneets / pawsru / sonyeon / vg / wx ][Options][ watchlist ]

/tech/ - Technology

You can now write text to your AI-generated image at https://aiproto.com It is currently free to use for Proto members.
Email
Comment *
File
Select/drop/paste files here
Password (Randomized for file and post deletion; you may also set your own.)
* = required field[▶ Show post options & limits]
Confused? See the FAQ.
Expand all images

File (hide): 6e533f1e76c5538⋯.png (4.22 KB, 225x225, 1:1, rust.png) (h) (u)

[–]

 No.973095>>973113 >>973126 >>973585 >>974267 [Watch Thread][Show All Posts]

It keeps gaining steam. Apparently it's faster than C++, memory safe, and thread safe. Has anyone used it? What are your thoughts?

 No.973099>>975024

use c


 No.973100>>973103 >>974285


 No.973103>>973135

>>973100

Steve?


 No.973110

Fucking weak thread, OP. Apply yourself.


 No.973113>>973504


 No.973115>>973127

>Docs have a 'dining philosophers' game including Foucault and like.

>Community is pozzed and mostly duplicating the deliberate cult atmosphere of Haskell, where critics are the enemy and get dossiers compiled about them.

>ATS is better.


 No.973126

File (hide): 8a34fe8b0511963⋯.png (7.51 KB, 225x225, 1:1, 5432565636535634.png) (h) (u)

>>973095 (OP)

>Forgotten languages


 No.973127>>973130

>>973115

>where critics are the enemy and get dossiers compiled about them.

WTF


 No.973130>>973145

>>973127

a thing Don Stewart has done in Haskell. I haven't seen that in Rust but I've kept plenty of distance from Rust.


 No.973135

>>973103

lurk moar newfag


 No.973143>>973344 >>973384

They somehow figured out how to make a syntax worse than C++. Just use Haskell.


 No.973145>>973169 >>973503

>>973130

yet they allowed dibblego to be part of the haskell community for years


 No.973169

>>973145

I've seen the name but don't remember anything about him.

https://twitter.com/dibblego/status/1019177499853389824

seems like an interesting guy. he's too toxic now?


 No.973344


 No.973349

>>its great in my basement!

thats not how any of this works

first there have to be actually good implementations used in say few larger games

technically its interesting and actually has potential if utilized properly


 No.973384>>973503

>>973143

>They somehow figured out how to make a syntax worse than C++.

>Just use Haskell.

Joke of the century.


 No.973503

>>973145

What do you have against my boi Tony?

>>973384

>complaining about syntax

hello nodev.


 No.973504

>>973113

God damnit


 No.973585>>973617

>>973095 (OP)

Ultra-pozzed devs that ban and block anyone not falling in line, which is how they can keep spouting nonsense claims such as

>faster than C++


 No.973605>>973619

File (hide): f41a4b54f0869e1⋯.jpg (38.37 KB, 1023x683, 1023:683, d137937ca8ae87cd77750b9494….jpg) (h) (u)

I like Rust. it wasn't too hard to learn, and I'm honestly a better programmer because of the habits it made me adopt.

I'm also really enjoying the zero-cost futures, as well.

That said, I bypass the CoC, and contribute to the project pseudonymously, so they can't link my shitposting to my work. People that get to see my resume get my dox, but the rest of the community sees another _why_the_lucky_stiff and doesn't realize I would gas them all if I had the chance. I consider this a win/win.


 No.973611

>Linux having a CoC is bad

>Rust having one is ok though


 No.973617

>>973585

You're right nothing can be faster than a segfault.


 No.973619>>974267

>>973605

>zero-cost futures

You know anon just because they say zero-cost does not make it cheap. They literally copied the javascript concurrency model.


 No.973668>>974027 >>974332 >>974419 >>974967 >>975021 >>975686

File (hide): 32cbef1991bff6d⋯.webm (489.27 KB, 640x360, 16:9, rust.webm) (h) (u) [play once] [loop]


 No.974027

>>973668

the end of that is hard to watch


 No.974267>>974285 >>974336 >>975021

>>973095 (OP)

I'm learning it now. I'm not a big fan. For context I'm an engineer (civil, not software), and I use Python regularly (scrapers, data visualization, task automation), have no professional or academic background in software, and I have dabbled with C, Haskell, and JavaScript.

Syntactically Rust appears to be the mulatto offspring of JavaScript and C with most of its parents' warts. I know the argument "hurr, hurr, real devs don't care about syntax, and I'm a real dev who works at a help desk/ is in college." But there are already much better and more readable syntax paradigms in use by popular languages right now. I've never written one line of Ruby, but I can look at its source and have an immediate understanding of what's going on. Contrast that with curly braces hell where they have multiple meanings depending on context and semicolons because why should a compiler ever interpret a newline character in a way that is understandable to humans?

Obviously this was a "guys, we need to make it like JS to not scare the web devs" kind of decision, but that's a terrible reason. It's propagation of poor design decisions because we're used to them.

I would characterize the language rules as autistic and obnoxious. Keeping up with ownership and lifetimes is a pretty unusual Paradigm. I understand why it was done, now that I know more, but it still is a irksome way to do things. however, I can't think of a better way to handle problems like memory safety. The current Mantra of C coders which is to "just run valgrind," or "just gitgud, XD," is not in any way an actual solution.

There are a few places where rust does it right, though. The documentation is pretty good. The meta-programming tools are really excellent. At this time there is not really an excuse to not have a tool chain which automate processes like packaging libraries for distribution or having a convenient way to import new libraries, like crates. There are other languages that have this, but often it relies on some kind of third party platform. Rust did it right by making a stable, usable default to get everyone on the same page.

I am sticking with rust right now because I want the ability to do embedded software or tackle systems programming tasks if they come up. It would be nice to expand the skillset I've got outside of just scripting, and, frankly, I'm too late in my career to spend 10 years getting C experience to just *start* writing useful code. From where I stand, I do not see a better alternative to rust. I really wish that there were one. Crystal is dead in the water without a genuine threading ability. I don't see that happening anytime soon. Also, Nim is Okay, but ultimately it is a C transpiler. Somebody correct me if I'm wrong on that one. I don't see how you escape the really fundamental problems with writing C.

Show me a better embedded-capable language that doesn't require years/decades of practice and I'll jump the Rust ship.

>>973619

The JS influence is pretty heavy and a I think a very bad thing.


 No.974285>>974294

>>973100

I've been in mostly read-only mode since the end of February. My total posts since then on /tech/ have been lees than 10. I didn't create this thread. You would know that I put way more effort into the OP if you weren't a retarded LARPer.

Also I'm /pol/, not /leftypol/.

Anyways since I'm already here I might as well answer any Rust related questions (I have >3 years Rust experience).

>>974267

>this whole wall of retardation

wew

>Show me a better embedded-capable language that doesn't require years/decades of practice and I'll jump the Rust ship.

Unironically C. The only place where choosing C makes any amount of sense.

>JS influence

Yep you are retarded.


 No.974294>>974295

>>974285

>-t. works in a call center

Adults are talking here, junior. Go play outside.


 No.974295>>974302

>>974294

>LARPing

>12 btw XDDDDDDDDDDDDDDDD


 No.974302>>974303

>>974295

LARP? I can't tell if you're mindlessly parroting image board maymays because you're a bot or if you are actually too dumb to make original thoughts.


 No.974303

>>974302

>>-t. works in a call center

>Adults are talking here, junior. Go play outside.

>original thoughts

huh... really makes my thought noggins joggin.........


 No.974327

I feel like we've had this thread before...


 No.974332>>974390

>>973668

>can't free memory in C

the absolute state of modern programmers.


 No.974336>>974390

>>974267

>a great language for someone too dumb to learn c

Quite the ringing endorsement.


 No.974390>>974396

>>974332

They need their hand held. Your skills at manual memory allocation management will earn you more money, as your skillset becomes more rare.

If all the Visual Basic programmers learned Rust instead, we'd be in a better world. Can we try to push for that? They'd be so easy to teach how to love a less strict compiler and freeing a malloc or five if they never learned all the VB mistakes and laziness.

>>974336

the inheritance model is really nice, and I try to bring it with me to any project I work on, because it makes memory management so much easier once internalized.


 No.974396>>974974

>>974390

> memory allocation management will earn you more money

Lol no you will make less than an SF web developer.

>as your skillset becomes more rare.

Cobol and Fortran devs don't make shit on average.

>the inheritance model

inheritance is shit. composition all the way


 No.974419>>974448 >>974929

>>973668

Why would anyone listen to someone who talks and carries themselves like this?


 No.974448

>>974419

I know, right?


 No.974929

>>974419

It's not nice to make fun of retards tbh.


 No.974941>>974953

freeing memory for running programs is terrible idea, theres a myriad of reasons why everyone avoids it unless its some super optimized game engine thing


 No.974953

>>974941

>I have infinite RAM


 No.974954>>974956

C is garbage collected. Try allocating more memory than your machine has. Does it crash? No. C is garbage collected.


 No.974956>>974994

>>974954

>Try allocating more memory than your machine has

Memory is allocated in a lazy fashion. Allocate more memory than your machine has and then write 1 byte to each page and it will crash.


 No.974967

>>973668

the function free is too hard :,(


 No.974974>>974975

>>974396

>you'll make less than a web dev

that is literally the lowest paying career path in software development

>Cobol and Fortran devs don't make shit on average.

that used to be one of the highest paying career paths last I checked

> inheritance vs composition

you can actually use both retard


 No.974975

>>974974

>that is literally the lowest paying career path

Still better than you trivial c faggots.

>that used to be one of the highest paying career paths l

Only if you are the 0.001% of contractor

>you can actually use both retard

Yeah you can shit and piss at the same time


 No.974994

File (hide): 46f17866bc1956c⋯.webm (10.61 MB, 896x504, 16:9, ex-rust-developer-intervi….webm) (h) (u) [play once] [loop]

>>974956

You can tell the kernel not to overcommit tho. If you do that malloc will return 0 when you're out of memory.


 No.975011>>975013 >>975022 >>975681

What I love about Rust is the core language’s notion of guaranteeing memory safety (akin to what you’d get from a garbage collector) without an actual garbage collector, and all the performance benefits that entails for any sort of real time application where random pauses and stutters are not okay.

I also like that there are no nullable pointers (sum types are used instead). Seriously, nullable pointers need to die already (except for limited contexts like “unsafe” blocks).

What I don’t like about Rust is a language that is perhaps too complicated in ways that are unnecessary (C++ is way more complicated, but for “legacy” reasons --- Rust is a fresh start and has no such excuse), poorly designed “ergonomics” around their borrow checker stuff, vague and unhelpful error messages, and last but not least: that awful SJW CoC.

It’s been getting better in some ways, but worse in others --- particularly, nasty compiler bugs that will segfault your completely correctly written program, bugs where the borrow checker fails, etc. And to top it off, their management seems kind of lackadaisicle/dispassionate about even prioritizing fixes for these.

They’re more interested in their future-looking rewrite and research projects than actually fixing bugs. And of course, if you tell them that’s a really stupid way to manage a programming language, they’ll ban you for intolerance and “toxic meritocracy” via their CoC.


 No.975013

>>975011

>“toxic meritocracy”

Please tell me that's not an actual quote.


 No.975021>>975022

>>974267

>using rust for embedded, of all things

If you do any serious embedded work (e.g. on a resource-constrained microcontroller), you'll need to interact directly with hardware peripherals (intrinsically and unavoidably unsafe) and you wont call malloc very often, if at all (so memory leaks wont be your biggest problem). So what's the point of using rust?

>Show me a better embedded-capable language that doesn't require years/decades of practice and I'll jump the Rust ship.

C. Too hard for you? Or do you think rust would somehow save you from having to be careful about resource management on a resource-constrained system? Are you the thing in that video: >>973668 ?


 No.975022

>>975011

>Seriously, nullable pointers need to die already

No??? Pointers are the only types where null actually makes sense.

>It’s been getting better in some ways, but worse in others --- particularly, nasty compiler bugs that will segfault your completely correctly written program, bugs where the borrow checker fails, etc.

There are a few bugs in the borrow checker but they have been there for years. There really aren't that many new borrow checker bugs.

Also they are working on it. They are completely rewriting the borrow checker. (For some reason they have decided to not fix the bugs in the old version while working on the rewrite.) The current borrow checker operates on the AST, the rewrite will operate on the MIR (mid-level intermediate representation). Apparently it is way easier to do borrow checking on the MIR than the AST. Also the rewrite will come with some cool new features like upgradeable borrows, non-linear lifetimes and better error messages. The rewrite is feature complete since quite a while. You can use it on the nightly compiler. I think it will hit stable in version 1.31. So in about 12 weeks.

http://smallcultfollowing.com/babysteps/blog/2018/04/27/an-alias-based-formulation-of-the-borrow-checker/

http://smallcultfollowing.com/babysteps/blog/2018/06/15/mir-based-borrow-check-nll-status-update/

>And of course, if you tell them that’s a really stupid way to manage a programming language, they’ll ban you for intolerance and “toxic meritocracy” via their CoC.

The Rust devs are actually pretty good about taking criticism as long as you don't use mean words.

>>975021

upboated and sage negated


 No.975024>>975831

> Apparently it's faster than C++,

hurr

>>973099

> use c

no macros, weakly typed, outdated 70s language. no thanks


 No.975035>>975037

If you've ever written non-trivial program in rust, you'd know that it doesn't make it any easier than c to write proper programs. It just tries to make the compiler fail whenever you've done something against the rules. For a normal development model, this is a non-starter: you first hack together a barely working version of your program, analyse what works, then rewrite the crappy parts. Repeat until the code is of sufficient quality for your means (engineers for nuclear shutdown systems spend 90% of their time analysing, web devs stop after the first step). This is why we see so many "rewrite it in rust" projects: rust is a terrible tool for prototyping, but a great tool for rewriting working code with.


 No.975037>>975154 >>975166

>>975035

>LARPer tries to talk about software development

You think before you write code. Only LARPers shit out shit code and then have to rewrite it because it it shit.


 No.975154>>975657

>>975037

I'd like to see you tackle a project of any real size, like a 3D game, without re-writing anything as you go. You sound more like a LARPer than he does.


 No.975166>>975657

>>975037

>first we list all the requirements, then we design all the control flow graphs, then we hand it off to the code monkeys to write it

>why are you so far behind schedule? We already did the hard work of designing, all you had to do was translate it into working code

>what do you mean our design documents were shit and you had to rewrite everything?

Is rust the perfect waterfall programming language?


 No.975169>>975657

You fucking rust spammers are the worst. GTFO.

Faggots not welcome here.


 No.975657

>>975154

stfu LARPer

>>975166

nice strawman

>>975169

>this rust thread is the worst on /tech/ right now

the whole of /tech/ is a fucking shitshow. sage negated btw


 No.975665>>975666

File (hide): b890d9553cf8555⋯.jpeg (42.25 KB, 640x636, 160:159, arson-selfie-cat.jpeg) (h) (u)

Can the Rustfags die already? We are never going to use pozsoy in a million years. Sooner we will remove shared library of Rust and gimp the execution of any Rust code.


 No.975666>>975667 >>975727

>>975665

>We

>we

Oh no! A bunch of LARPers won't use Rust. Rust is finished!


 No.975667

>>975666 (checked)


 No.975681

>>975011

>particularly, nasty compiler bugs that will segfault your completely correctly written program

That's the opposite of what i would expect from a language that goes on about how memory safe it is.


 No.975686>>975692

>>973668

What the fuck is that thing?


 No.975692

>>975686

your mum


 No.975727>>975729

>>975666

Sperging out will not make me use Rust. Rather it will push it away more than it already is.


 No.975729

>>975727

What makes you think I care about you not using Rust?

LMAO you are just a retarded LARPer.


 No.975831>>975836

>>975024

What? C has macros.


 No.975836

>>975831

Really really bad hacked in preprocessor ones: https://en.wikipedia.org/wiki/Hygienic_macro




[Return][Go to top][Catalog][Screencap][Nerve Center][Cancer][Update] ( Scroll to new posts) ( Auto) 5
68 replies | 5 images | Page ?
[Post a Reply]
[ / / / / / / / / / / / / / ] [ dir / animu / arepa / asmr / ausneets / pawsru / sonyeon / vg / wx ][ watchlist ]