Untitled

I think thats the joke

Not sure if you know what safety is referring to

Attached: 1521476742401.png (620x463, 273K)

Fixed.

Attached: 15498226678127.png (1207x826, 116K)

gitlab.redox-os.org/redox-os/redox/issues/1136
weird, where's all the security bugs like a c os? surely i can complete the redox crash challenge

youtube.com/watch?v=iSmkqocn0oQ

>python, lua and javascript on the safe side
you must be joking.
those are dynamic languages with very little type safety and compiler checks to ensure correct code.

>javascript on the safe side
Are you blind?

based af
wonder what simon peyton jones thinks about rust

How does C++ provides less control/performance than C?

>some hipster substitute for javascript
oh misread that, but lua and python are stll wrong.

Kek.
On a serious note though, isolation of unsafe code is not a bad idea in itself.
Inb4 write small unsafe core in C and write the rest in a safe language.

also php doesn't have bad performance

>Inb4 write small unsafe core in C and write the rest in a safe language.
Isn't that Java? My teacher said Java is written in C++ and you should just use C++ because java sucks.

I would even claim that C++ is faster in some cases since function pointers can be replaced by templates at some ocassions

But Rust depends on the notion that it's an inherently safe language by design.
If that's not true, and it's just that all the unsafe code was swept under the rug for you, then why not just use C/C++ and stick to the STL?
Same idea, right?

Okay, the idea is that you do all the dirty work inside unsafe blocks. That unsafe code has certain invariants that, if broken, will cause undefined behavior. You then wrap the unsafe code in an API that makes sure that the users of said API can't use it in a way that breaks the invariants. This is where Rust's type system with the borrow checker, type traits, lifetimes, etc. comes into play. With them, you can go a loooong way without having to resort to runtime checking of invariants.

Since C++ doesn't have a borrow/lifetime checker, and concepts are still WIP, it's much easier to abuse the STL and other "safe" APIs to cause undefined behavior.

Although I hear that the C++ commitee is actually thinking about implementing some form of borrow/lifetime checking, which should be interesting.

Okay, I autismed too much, user out.

Standard ML's specification is provably type safe. That is why those whole-program optimizing compilers are so ridiculously aggressive and efficient. It should be off the list compared to shitty Haskell and Rust.

You can write performant Haskell, but it's just doesn't feel nice to do.

FORTRAN isn't faster than C. It's stack based which is inferior to register based/hybrid languages.
Stack based is safer tho, I'll give you that.

:^)

Attached: graph22.png (768x592, 14K)

C/C++ are the ultimate zoomer hipster languages

Why is this a bar graph?

Can you not read a 2d graph chart user?

Fortran is faster than C because of static memory.

C isnt a register language
you are using nonstandard features

>Runs in parallel
Oops sorry Rust guess you can't stay here

>Haskell low control and or performance
What kind of retard stores large ammount of data in lists?

Jow Forums plz tell me what does safety means in this context? i see people talkin shit about c++ what do they mean? any links i can reada bout it?

memeory safety, one of the main "features" of c/c++ is manual memory management, and when used im properly it leads to memory leaks, because not only are you responsible for allocating the memory you also have to deallocate it.

Fake and gay, assembly never crash

Literal childs trying to avoid responsibility use Rust

Assembly is never late, nor is he early, he arrives precisely when he means to!

if your CPU has bugs assembly can fuck up.
especially using undocumented opcodes
besides that your OS can always rape you no matter what you do

but the Rust compiler is strict and forces responsibilities on you, you could try and say that for some dynamic interpreter language and it would make sense

It forces it's CoC on me as well ;^)
miss me with that gay shit

[citation needed]

>#include

Wow so hard

rust-lang.org/policies/code-of-conduct

you don't have to join the rust team when using rust

What the fuck is this safe/unsafe memery by all you zoomers?

No programming language is safe because it has to run on physical hardware.

using rust is supporting ((their)) political agenda
it would be like a vegan saying its ok for him to buy meat and then he doesn eat it because he is a vegan.

but you can use Rust and still say nigger on twitter or whatever you enjoy doing without any CoC repercussions, so that analogy doesn't work

Attached: 1549822667812-2.png (1609x1197, 214K)

You're thinking of FORTH, retard. FORTRAN is indeed faster than C and is not stack based.

saved

A true man of taste

it is
it has more advanced type system

Attached: 4348.png (1207x826, 12K)

Check'd

Attached: preview.jpg (1024x422, 45K)

fuck off fucking tripnigger

Attached: fixed.jpg (1311x908, 160K)

The only thing Rust has done is show that C++'s complexity is justified for the most part.

safety is for javascript babies.

Why do rust programmers evangelize all around reddit and twitter and Jow Forums and shit with rust posts and downvote and spite and sage and comment, by the dozens, anything that doesn't praise rust?

They just absolutely crowd around "big opinion man just implied he used rust on twitter, a win for rust!!!" posts and "I forgot to free some memory I allocated!!! C and C++ are bad!!!" type of discussions? Are they all just fucking webdevs that went to degree-mill colleges and got scared of C and C++ because their dumb ass professors still think C++ is C++98? Because their professors are also shitty webdevs and java brainlets themselves that are scared of cleaning up their own memory? Because they need dedicated garbage collection threads and up-front 4GB heap allocations just to feel "safe" rather than discipline themselves to handle their resources properly and use things like unique_ptr and RAII and reference counting?

I've had people approach my C++ projects with "why not use rust?" comments even and people are making github issues with "have you considered rewriting your library in rust?" like the fucking plague.

github.com/search?o=desc&q=rust rewrite&s=&type=Issues

I've just never been pretty much forced to have distaste for the community around such a programming language with basically hype-driven-development. Rust is just yet-another-llvm-frontend that seems to need `unsafe` just to get any real work done when it interacts with all the C and C++ library bindings and wrappers that make it useful for anything besides being a toy language.

Attached: rtx2080ti-unboxing-2.webm (640x360, 1.46M)

Is fortran used anywhere other than calculation for science and mathematics?

misery loves company.
and rust faggots love e-begging for relevancy in every corner of the net.
Because nobody actually uses their garbage. They want so badly actual software and libraries, but offer nothing to the programmers making them.

>Is [language] used anywhere other than [its intended use cases]?

well, sure, tex was used to program a nasa rover once, doesn't mean it efficient or easy to do so, sorry for asking, gonna shitpost next time, discussion is for faggots anyways

(You)

How is C more unsafe then JS ?
Doesnt make sense in my mind
t. JS programmer

github.com/ansuz/RIIR

Going in Bestof folder.

What about Typscript though

>where's all the security bugs [in this toy os that gets used by dozen people for things such as: booting up and shutting down, maybe occasionally barely-testing a feature or two] like a c os [that gets used by millions of people for anything and everything between running a toaster to mining meme currencies on devices of visitors of kyrghyz yak-milking forum]?
you're welcome
now tell me why are all provably safe systems written in either C or Ada?

This, but also with C# at the same level of time spent as Java, but with a tiny bit less safety and C++ level control.

>Rust
>Performance
>Control
Rust forces you to pick one or the other, and it doesn't have the best of either one.

What is with this pervasive meme that C is faster than C++, they pretty much have identical performance in most cases (C90 is almost subset of C++). C++ can practically be faster because templates can replace void* or indirection in C functions (a classic example is std::sort vs qsort), but otherwise the performance is identical.

>otherwise the performance is identical
Most benchmarks say otherwise.

Which benchmarks? And I'm not doubting, I'm just curious, because I haven't seen these benchmarks online.

I know microbenchmarks are shit but benchmarks games shows them as being roughly equally matched: benchmarksgame-team.pages.debian.net/benchmarksgame/faster/c.html.

Furthermore C and C++ share almost identical semantic rules and they both share a given optimizer. I can't thing of a possible optimization for C that isn't available for C++.

C++ doesn't even have restrict.
Also, many C++ features have a negative impact of performance.

sure but c doesnt even have those features so how can that be a fair comparison?

This is the dumbest post I have read today.

C++ supports restrict as a compiler extension in all the major compilers and most C developers don't have restrict either, except for the 4-5 projects out there that actually use C99.

Otherwise RTTI can be disabled and exceptions are zero-overhead on x86-64 and may even be more performant than checking each individual error. What other features negatively impact performance?

>extension
Now this discussion is completely meaningless.

Its available in every major compiler and the vast majority of C projects use __restrict instead of restrict because it's more portable. Also where are those benchmarks?

HolyC is the only turing complete programming language

>Go's regex
I laughed

>python
>safe
This pic brought to you by a retard.

The "safety" axis is retarded, replace it with "level"

Lua is slower than Haskell tho.

how the fuck is lua as fast as C and C++

Doesn't safety mean "interaction with environment"?

C++ *is* faster than C when used properly, despite what Cniles desperately claim.
All that sweet extra information the compiler gets can potentially be used for further optimization.

You are confusing language semantics with their implementation.
Fortran's explicit no-aliasing policy makes it inherently faster.

>Haskell is normally as fast as Java or faster
are you on crack lmao idiomatic haskell is much slower than java

I think he is referring to LuaJIT

*blocks your path with rayon*

I write Rust code 8-9 hours a day an I admit that sometime a month of Rust could be done in 5 minutes of catastrophic Python.