Compile a code

>compile a code
>14 warnings
>program works fine

do you fix the warnings or do you ignore them Cred Forums?

"must be a glitch."

g++ -w shitcode.cpp

I fix them since warnings trigger an autismal instinct to make the compiler produce 0 complaints, i need perfection

>fix warnings
>program doesn't work anymore

>a code
How many codes have you programmed in your life?

who cares, the code is perfect
also
>give me succ maki you hot slut

>do you fix the warnings or do you ignore them Cred Forums?
pretty much all coding guidlines everywhere state that you have to fix all warnings.

"Just ignore them." - Enterprise dev probably

...

They're informational only. Not exactly a problem but might lead to one (buffer overflow, mixture of spaces and tabs). Basically keep the code to yourself or you're going to learn what a hotfix is

>program works fine
That's because they're warning and not errors.

maki maki gimme that facki

Why the fuck wouldn't you fix them?
A good compiler warns you about undefined behavior or ambiguous shit that might be understood differently on other platforms.

I don't understand the obsession with maki.
She's cute, but she's just a side character on love-live and barely gets any screentime.
Also, she has literally zero characterization other than being naive and still believing in santa.

go away nico

this is Cred Forums.. a bunch of basement dwelling neckbears, wtf do you expect ?

>Open a C# testbed using selenium framework
>800 fucking warnings on compile because they didn't surround element definitions with the right pragma
fucking amateurs

I guess it would depend what the warning is. Sometimes I'll be doing something weird with switch-case fallthroughs and early returns and stuff and the linter will think there's an unhandled case or something, but due to some precondition of the way the method is called it's not actually possible to end up in that state. I'd probably ignore something like that, whereas I'd listen to something that said unreachable code or whatever.

>shitting on stylecop
You take that back you dirty cowboy coder

It's a really specific thing to do with the way the testbed was written, and due to how the selenium framework defines elements if you're not programatically searching for them
It's a dumb quirk of the framework in C# and not an issue with stylecop

Well in C if you call a function with the wrong type of data (like char * instead of char) it will only trigger a warning and will compile just fine.

If you have a warning that you don't understand, you've fucked something up (wrong include, wrong cast of data,...)
If you know what the warning mean and you know it's the compiler talking shit, you're fine.

TL;DR Your compiler is smarter than you

No merge unless the code is warning-free

Except in the case of silly/stupid warnings, but I disable those in the build script already.

Any particular reason you chose selenium over Coded UI or White (given that you're working in C#?)

she's feminine and a slut, which is good

It depends on the language honestly. In Haskell, most things are errors, and only syntactical or stylistic things are warnings.

In C, almost nothing is an error, and everything is a warning even if it causes undefined behavior and will almost surely segfault your program.

So in C, you absolutely have to fix those warnings, because warnings are how the C compiler communicates to the programmer that their program is probably wrong - whereas in Haskell, warnings are how the compiler communicates to the programmer that their program uses bad style.

From your post it sounds like you're a newfag who thinks that LL is about the anime. Lurk more newfag.

it depends

this.

don't even think about making a PR unless your code compiles warning-free with -Wall -pedantic and valgrind with --leak-check=full doesn't report any memleaks

src/substr.c: In function ‘substr_extract’:
src/substr.c:68:15: warning: assignment makes pointer from integer without a cast
out->o_len = k;
^

Usually, compiler warnings remind you that your code doesn't do what you think it does.
I forgot the array index in this example.

>don't even think about making a PR unless your code compiles warning-free with -Wall -pedantic and valgrind with --leak-check=full doesn't report any memleaks
This is why all good projects have CI, so your PR runs through all the relevant tests automatically. No merge unless the test frameworks report no errors

Secretly veiled Maki thread?

I compile with -Werror, so I fix the warnings.

>being a shitty programmer

If you knew what the fuck you were doing this would never happen.

...

You didn't fix the warnings, you must made them go away.

It depends on the warning.

>secretly

feel desu

>tfw you pull a UR Maki

>being so bad at programming you break your code fixing warnings
Do you just randomly pick one of the recommended solutions?

Do a formal proof of your code and come back.