Programming thread

DavidS877 said:
The 1-Wire is the 'slowest' which just means I kick it off and then come back 750mS later
Click to expand...
Well, almost. The bus transactions are also slow. It can take nearly 10mS to get a data readout as the bus is designed to be slow and large. My main loop will be unhappy with more than 1mS blocking. So I can find better 1-Wire libraries using non-blocking or timers, or make/update my own. Or, what I'm leaning towards is taking an ATTiny, slapping an Arduino Core on it and having it poll the bus and spit out results in async serial, which the Pico can then lazily consume.

And now, back to my day job, today involving CI/CD and Python, yay.
 
  • Like
Reactions: Rusty Metal Skull Gun
Evil Whitey said:
Any Ziggers in here? If so, what do you think of Zig, and would you recommend it?
Click to expand...
I looked at learning it a while ago as a low-level language and gave up because the documentation was shit and standards were changing or deprecated with some expected things just not having a replacement. It's probably in a better spot today though as that was 0.11 and it's at 0.14 now.
WielkiWapiesz2137 said:
The main selling points are excellent interoperability with C (you can import C files without any retarded wrappers), a build system that uses the language itself instead of some moronic DSL such as makefiles and convenient features such as defer, option types, etc. Its tooling is far from ready, tho, and it forces you to write boilerplate for example, you need to "consume" the return value of function which amounts to putting _ = before function calls otherwise it won't compile.
Click to expand...
The C interop and non-retarded syntax was what attracted me to Zig initially along with these claims:
  • No hidden control flow.
  • No hidden memory allocations.
  • No preprocessor, no macros.
Click to expand...
I fucking hate hidden magic stuff in my code, from times when I worked on Java Spring apps and the god awful dependency injection that went with them.

Comptime stuff in Zig also just seemed like a fancier preprocessor to me.

I learned a bit of Rust instead and all I'll say is that language can be annoying as fuck to write simple stuff in and treats the programmer like an idiot child (which is often true). It's still an okay programming language, but mostly because of its type system in my opinion.
 
  • Thunk-Provoking
  • Informative
  • Like
Reactions: Rusty Metal Skull Gun, Evil Whitey and ADHD Mate
WielkiWapiesz2137 said:
it forces you to write boilerplate for example, you need to "consume" the return value of function which amounts to putting _ = before function calls otherwise it won't compile.
Click to expand...
why the fuck
this is some of the most retarded bullshit i ever saw
oh wait i remember now it also errors if you have an unused variable
0/10 niggerlicious language with retarded overly pedantic compiler, avoid at all costs

just seems like one of those languages that attempts to "fix c" without actually knowing what c's issues are, so they end up wasting an unholy amount of work making a worse c
 
  • Like
  • Informative
Reactions: Bababooey Warlock, The Taxman, Marvin and 4 others
Creative Username said:
oh wait i remember now it also errors if you have an unused variable
Click to expand...
If only they had spent 3 minutes with C...

Code: 
$ cat main.c
__attribute__((warn_unused_result)) int seven() {
  return(6);
}
int main() {
  int i;
  seven();
}

$ gcc -Wall -Werror main.c
main.c: In function ‘main’:
main.c:6:5: error: unused variable ‘i’ [-Werror=unused-variable]
    6 | int i;
      |     ^
main.c:7:1: error: ignoring return value of ‘seven’ declared with attribute ‘warn_unused_result’ [-Werror=unused-result]
    7 | seven();
      | ^~~~~~~
cc1: all warnings being treated as errors

You can get all the errors you want. And more importantly, shut them up.
 
  • Like
Reactions: Rusty Metal Skull Gun, 306h4Ge5eJUJ and Creative Username
DavidS877 said:
If only they had spent 3 minutes with C...

Code: 
$ cat main.c
__attribute__((warn_unused_result)) int seven() {
  return(6);
}
int main() {
  int i;
  seven();
}

$ gcc -Wall -Werror main.c
main.c: In function ‘main’:
main.c:6:5: error: unused variable ‘i’ [-Werror=unused-variable]
    6 | int i;
      |     ^
main.c:7:1: error: ignoring return value of ‘seven’ declared with attribute ‘warn_unused_result’ [-Werror=unused-result]
    7 | seven();
      | ^~~~~~~
cc1: all warnings being treated as errors

You can get all the errors you want. And more importantly, shut them up.
Click to expand...
This is my favorite aspect of languages like C and JavaScript. The language has been separated from the tool (compiler, linter), allowing you to choose your tools and more freely express your own code style without unwanted nagging. Almost all nu-langs fail in this regard: Go, Rust, Zig, the tool is synonymous for the language.
 
  • Like
  • Thunk-Provoking
  • Semper Fidelis
Reactions: Polyarmory, Kuroyuri, y a t s and 5 others
Lard Shart said:
This is my favorite aspect of languages like C and JavaScript. The language has been separated from the tool (compiler, linter), allowing you to choose your tools and more freely express your own code style without unwanted nagging. Almost all nu-langs fail in this regard: Go, Rust, Zig, the tool is synonymous for the language.
Click to expand...
this is exactly it, the designers think that they are making an "ecosystem" and shit instead of just the first program that turns their gay language into machine code
 
  • Like
  • Winner
Reactions: Safir, Zeftax and Lard Shart
Any novel innovations in languages are already in a Lisp. I thought I was clever when I was trying to come up with my own language some pages back now, but I was pleasantly surprised to find that anything I could have possibly added was already in the Lisp I settled on (Clojure).

"Innovations" in languages these days are just about making them as pajeet-proofed as possible. Thus cargo/go get replacing NPM, thus a single source of truth replacing a spec and a variety of compilers and platforms, etc.
 
  • Feels
  • Like
  • Winner
Reactions: TroonsDid911, Marvin, Creative Username and 5 others
Evil Whitey said:
Any Ziggers in here? If so, what do you think of Zig, and would you recommend it?
Click to expand...
Whether Zig structs are passed by value or by reference is intentionally up to the compiler on a per-function basis because copy elision is hard or something, which causes all kinds of funny aliasing bugs and makes me think the author is an idiot.
 
  • Horrifying
  • Islamic Content
  • Thunk-Provoking
Reactions: Nitro!, The Taxman, Rusty Metal Skull Gun and 9 others
306h4Ge5eJUJ said:
Whether Zig structs are passed by value or by reference is intentionally up to the compiler on a per-function basis because copy elision is hard or something, which causes all kinds of funny aliasing bugs and makes me think the author is an idiot.
Click to expand...
if i understand correctly this sounds just like normal copy elision except it's UB to mutate anything passed by value because fuck you
this easily wins the kiwi farms niggerlicious programming language award in the category of "makes c++ look really simple and intuitive"
 
  • Agree
Reactions: The Taxman, Monad, ADHD Mate and 2 others
Evil Whitey said:
Any Ziggers in here? If so, what do you think of Zig, and would you recommend it?
Click to expand...
At the end of the day, language creators need to understand that what they're doing is essentially giving a programmer a loaded gun. You can cover the gun in warnings and all kinds of devices to try to make sure they use it properly, but you can't really stop someone who is determined to shoot their cock off with it from doing exactly that. Zig has a lot of interesting features, the C interoperability specifically looks neat and the built in defer/errdefer is nice, but the author seems to have gone quite a bit past "give them the features to not make jeetcode" and right into "you must write this my way because then you can't make jeetcode" which is just a false assumption. You can always make jeetcode, just like you can always find a way to shoot yourself with a loaded gun so long as it can actually fire.

My perfect example for this is how the language really, really wants you to do some nice error handling by trying to hogtie you into using try/catch, but any shit-covered jeet worth their caste is just going to copypaste some empty error discarding code block into every line requiring such and utterly ignore the intended result of that language feature. There's more, of course, but this is just the best example because it's something they already do a lot in other languages that seems to have not been considered or understood here.

So yeah it's not as bad as Go or one of those other fag languages but it's still kinda pointless as of yet. We'll see where it goes.
 
  • Thunk-Provoking
  • Agree
Reactions: Lard Shart, WielkiWapiesz2137 and Evil Whitey
Genuine question, why C/C++ had become the defacto language for most programs and Lisp hadn't got away with being used in user software as much as C?
 
  • Like
Reactions: Polyarmory
1) Indians can't write Lisp.
2) Lisp runs on a platform, (like how java runs on JDK, C# on .NET, etc.) which made it historically expensive to be used to compute things, so it wasn't universally spread like the ALGOL descendants. Only (relatively) recently did hardware get good enough to run Lisps. There were historically dedicated Lisp machines, but they were expensive.
 
  • Informative
Reactions: Bababooey Warlock, The Taxman, Polyarmory and 3 others
Marvin said:
This isn't really true.
Click to expand...
You're thinking of current software, done in high-abstraction languages such as Java or Python. Historically, most programs are written in C++ due to what it could do if you knew how to use the memory pointers well, and a lot of institutional push towards it's standards.

Still, I wish I had proper Lisp classes, that language is really cool.
 
  • Feels
Reactions: Creative Username
One Angry Computer said:
You're thinking of current software, done in high-abstraction languages such as Java or Python. Historically, most programs are written in C++ due to what it could do if you knew how to use the memory pointers well, and a lot of institutional push towards it's standards.
Click to expand...
Current software is most software.
 
You must log in or register to reply here.
Back