Now that REAL alternatives exists, will this shit be relevant in the next decades for NEW projects (not just legacy system that are too big and critical to be rewritten in a new language)?
Now that REAL alternatives exists...
Connor Bennett
Other urls found in this thread:
cs.virginia.edu
github.com
twitter.com
Jackson Morales
never mind I figured it out
Justin Perry
bump
Cooper Wright
>inb4 rust
no sweetie that's not an alternative ;)
Luis Wright
Give me one technical problem C can solve Rust cannot.
Robert Hernandez
Having a stable and well-defined ABI.
Brayden Morgan
Levi Robinson
You mean like not having this
printf("%d %d\n", ++i, --i);
be UB?
Ian Nguyen
I think C is useful for learning about how programming languages work. There are so many fucking idiots in my C class that can't figure out how pointers work. But outside of school I really don't see a purpose for it at this point.
Ian Long
That has literally nothing to do with what I said, you fucking moron frogposter.
Jason Hall
Austin Jenkins
>an OS for ants
Wow, great alternative.
Landon Morgan
why yes I do program in rust.
how did you know?
Cameron Allen
Agreed, everyone should use this instead
Adam Morris
Why not use go?
Anthony Jones
github.com
>Note that C++ still doesn't have a stable ABI, and it took C decades to get one.
>Note that rust does support shared libraries. What it doesn't support is mixing them from different compiler toolchains.
Literally a non-problem. It's really funny given how Rust manages to be safer than C without it.
Evan Rogers
>>Note that rust does support shared libraries. What it doesn't support is mixing them from different compiler toolchains.
And that's fucking not acceptable. Why do I have to literally recompile my entire system when the Rust compiler gets updated?
Also enjoy not being able to have modules.
Ayden King
>Also enjoy not being able to have modules.
Dude what? Include files are a fucking mess. A proper module system with features like namespaces and visibility configuration is the right way to go.
Gabriel Watson
Dynamically loadable modules, you idiot.
Caleb Price
again???
Brayden Cruz
Because he wants to use julia
Ethan Ortiz
Static linking is better.
Nathaniel Price
>And that's fucking not acceptable.
100% of developers do perfectly OK with that, because 0% of them are autistic enough to compile different shared libs on different compilers. Also, dynamic linking is retarded.
Gavin Murphy
No it's not. I don't want to have to recompile/relink absolutely fucking everything every time something changes.
Also, it doesn't solve the issue with programs that have user-defined dynamically loadable modules.
Oliver Davis
in what scenario would this even matter
the 1 time rust updates it's compiler in 3 years?
and even then, you don't need to update you retard
Benjamin Russell
.a exists, you retardo.
Jack Long
And the overall Rust ecosystem currently does not make use of them. Even then, if the Rust compiler updates, they literally all become useless.
>the 1 time rust updates it's compiler in 3 years?
It's far more often than that.
>and even then, you don't need to update you retard
Yes, you fucking do. Not every program is going to be so nice and only target old versions of the compiler.
All of you retards seem to have no idea how the wider ecosystem works and how distros work.
Jason Brown
isn't abi just compiled api
Ian Wright
>if the Rust compiler updates
Which will happen less than once a year.
Luke Hughes
Hunter Turner
Now imagine Debian, a linux distro with like 25k packages or some shit.
Some of those are logically one of those packages that have been split up, so lets be conservative and say 10k.
Do you fucking know how long it takes to compile that many packages? Especially with a language that compiles as slow as Rust does. If everything was static linked, and then something like libc was updated, it would literally take years of compute time to do that.
People don't have years of compute time to solve such retarded problems. That's why dynamic linking is good; something gets updated, you just need to recompile the one thing.
Static linking just does not scale. Only proprietary cucks and people who only focus solely on their own single program think it's a good idea for static linking to be widespread.
The Linux userspace API/ABI is actually extremely fucking stable. Their internal APIs are not, but that's not important.
Grayson Butler
Rust does have a stable ABI, it's called extern "C"!
Parker Rodriguez
>the 1 time rust updates it's compiler in 3 years?
Dude, they update once a month. And considering there have been security vulns found in the standard library a couple of times, at this moment in time, YOU HAVE TO UPDATE.
Blake Howard
>Now imagine Debian, a linux distro with like 25k packages or some shit.
Compiling 25k .a files within a year? If you think that's unfeasible, you're a deluded moron.
Carter Davis
this
julia >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> rust
Jayden Gutierrez
C ABI is not a part of the standard. Rust is free to define ABI too but there is no compelling reason for it. Try again.
Adam Wilson
>within a year
No, every time a package is updated, every other package that depends on it transitively needs to be recompiled/relinked too. The Rust ABI breakage is just an extra layer of retardation on top of that.
Lincoln Jenkins
It's not part of the ISO C standard, but is properly standardized by each various platform. The standards committee clearly understands that this is one of the most important features of C, and makes an effort to not make platforms need to break their ABIs.
Levi Butler
>every time a package is updated
No. Every time the Rust compiler is updated. Once again, .a files exist. How the fuck did you find your mom's tits to suck on them?
Luis Scott
>I have no idea how linking works
Good one.
Michael Russell
>.a files exist
Fucking cringe. This is why I stay away from Jow Forums. Any of you lot actually work in software?
Brody Sanders
What alternatives?
Jaxon Jenkins
You made it crystal-clear I know better than you how it works.