hi guys, i am using non blocking i/o on a socket udp, but next i want to use again the blocking i/o, so how can i make it? which options should i use on setsockopt?

Are you using some kind of library/class for this? Because C/C++ doesn't support "non-blocking" natively.

Are you using some kind of library/class for this? Because C/C++ doesn't support "non-blocking" natively.

i used setsockopt with the SO_RCVTIMEO option

i used setsockopt with the SO_RCVTIMEO option

Which Library? These are not native to C/C++, maybe I could recommend using another library that doesn't support non-blocking... And why in the world do you want to use blocking? If you want to block something from happening, do it manually... if you're using GUI, disable the controls on Window, if you're using server side, put it in a while loop while waiting for process to finish or if you're using it on CLI, do the same...

i am doing a server for download and upload of files during the download, i want a non blocking recvfrom, but after the download i want a blocking recvfrom

Suckless software mostly sucks and is not usable by a normal human, only by crazy people like us

musl rocks?

musl rocks?

In terms of C standard library it is good, but also C is not always the best tool for every single task and suckless enthusiasts seem to think it is. LMDB and musl and some other minimalistic projects are very nice, but sometimes I would rather use C++ (which they call bloat without even considering that I do not use the hosted implementation)

#include<stdio.h> int main() { int a = 1; if(a--) printf("True") ; if(a++) printf("False") ; } What will be the output of this code?

its true

hi guys, how can i divide a project in multiple files?

hi guys, how can i divide a project in multiple files?

By making libraries... depending what you're using to program, which language?

in C

Hello guys, does any one know why float, double, and long double have the same precision of 7 decimal places in MSVC?

Hello guys, does any one know why float, double, and long double have the same precision of 7 decimal places in MSVC?

They have different precision

Hello guys, does any one know why float, double, and long double have the same precision of 7 decimal places in MSVC?

It's not

I tried them all and each time after compilation, my value is truncated to only 7 decimal places in the console.

import <iostream>; import <numbers>; int main() { double pi = 0.12345678901234567890; std::cout << pi << std::endl; } After compilation, 0.1234567

I tried them all and each time after compilation, my value is truncated to only 7 decimal places in the console.

use std::setprecision

use std::setprecision

Ok

also you have f in end of your double literal, that makes it float literal

also you have f in end of your double literal, that makes it float literal

Yeah, I added that only for testing purpose. Forgot to take it off before posting. Sorry

import <iostream>; import <numbers>; int main() { double pi = 0.12345678901234567890; std::cout << pi << std::endl; } After compilation, 0.1234567

Ha Ha. It's not right way to check precision

in C

Distribute things in libraries, your main program is always small, put other function in different libraries.

Ha Ha. It's not right way to check precision

You see, I am studying from a book so I try to create possible error scenarios in my codes to see how everything works. Used this approach for testing value ranges for the integer types and it worked well. However, despite the sizeof function indicating that double has twice the bytes of float and as much byte as the long double, all three floating point data types happen to have a precision of 7 for the microsoft visual c++ 20 that I am using

/duolingo

You see, I am studying from a book so I try to create possible error scenarios in my codes to see how everything works. Used this approach for testing value ranges for the integer types and it worked well. However, despite the sizeof function indicating that double has twice the bytes of float and as much byte as the long double, all three floating point data types happen to have a precision of 7 for the microsoft visual c++ 20 that I am using

You you want testing something - do it under debugger to see exact value. Moreover you can use std numeric_limits to get different properties of numeric types

use std::setprecision

std::setprecision works

You you want testing something - do it under debugger to see exact value. Moreover you can use std numeric_limits to get different properties of numeric types

Sure

try to avoid libevent because it is bit of a legacy libev is a library for asychronous programming using a callback funcion so you are expected to do your networking with <sys/socket.h> but note that libev has slightly better performance than the others because it is more low level and close to the kernel libraries which are usually the fastest thing you can get out of your system

SO_RCVTIMEO doesn’t make your networking non-blocking it just sets a timeout for each packet (think of it as a portion of a message) so your server won’t wait forever if your a client went offline suddenly you should use (epoll on linux or kqueue on BSDs including MacOS) or use a cross platform library that works everywhere also read the messages above where I explain it to someone else before ☝️

epoll isn’t really hard to use as you think on linux try writing this in the terminal “man epoll” and scroll down to the examples section that might help you note that epoll is for linux only so you have to use kqueue for BSDs (including MacOS) so unless you are planning to make your own asynchronous networking library for your project (which is a very good practice for learning), it would annoying to do so because you would have to write two versions of your code, one for linux and one for the BSDs

a cross platform library does that for you *

in C

By using header files, Just search about it

I tried using the debugger but I didn't get anything useful. However, using the setprecision function, I saw that despite having each different floating point type value printed with a precision of 7 decimal places, it appears that it is the doing of microsoft. The setprecision with an argument of more than the maximum precision of any given floating-point type produces gibberish after the precision limit. So 7 decimal places is the standard for all floating-point data type in MSVC or so it seems.

hi guys, how can i divide a project in multiple files?

it is a very good practice not to make your C files too long for readability and understandability assume that you have a function called int foo(){...} in a file called “foo.c” then you want to call from another file called “main.c” firstly the main should only exist in main.c (you can call this file whatever you want btw) you can create another file and call it “foo.h” when it would contain only the declaration of the function so instead of writing the whole function again in “foo.c” you would write only the following: int foo(); only with no main function in “foo.h” then in main.c before the main function you write: #include “foo.c” //with the double quotes or you can simple copy and paste: int foo(); into main.c before the main function then if you like to compile it you do: gcc foo.c main.c note that the order that you write main.c and foo.c doesn’t matter the compiler will always look for the main function then do its work

Hey was wondering how to write c/c++ code to read and write from a doc or excel....or basically any sw with an input output tab any help?

a cross platform library does that for you *

boost asio

std::string getMessage() { char length[4]; fread(length, 4, sizeof(char), stdin); uint32_t len = *reinterpret_cast<uint32_t*>(length); char message[len]; fread(message, len, sizeof(char), stdin); std::string content(message, message + sizeof(message) / sizeof(message[0])); return content; }

could someone help me ajudar.com this function

the error is on that line

char message[len];

char message[len];

VLA is gnu extension. You can directly read into std::string. Also you are violating strict aliasing rule when reinterpret casting char[] to uint32_t. You need to use std::bitcast or std::memcpy std::string getMessage() { char length[4]; fread(length, 4, sizeof(char), stdin); std::uint32_t len; std::memcpy(&len, length, 4); std::string content(len, '\0'); fread(message.data(), len, sizeof(char), stdin); return content; }

VLA is gnu extension. You can directly read into std::string. Also you are violating strict aliasing rule when reinterpret casting char[] to uint32_t. You need to use std::bitcast or std::memcpy std::string getMessage() { char length[4]; fread(length, 4, sizeof(char), stdin); std::uint32_t len; std::memcpy(&len, length, 4); std::string content(len, '\0'); fread(message.data(), len, sizeof(char), stdin); return content; }

thanks brow =D

hi guys I don not know what went wrong with that lambda add function. can you help me with this bug?

#include<iostream> #include<string> using namespace std; class book{     public:         int price;         string name; }; int main(){     book a,b;     a.price=1;     b.price=1;     auto add=[](book& a,book& b)->decltype(a){         book c;         c.price=a.price+b.price;         return c;     };     book cc=add(a,b);     cout<<"cc price is"<<cc.price<<endl;     return 0; }

boost asio

he wanted a C library boost is for C++

Hey was wondering how to write c/c++ code to read and write from a doc or excel....or basically any sw with an input output tab any help?

libxls can parse xls and csv

libxls can parse xls and csv

So all sw have different libraries for this?🤔🤔

#include <iostream> using namespace std; int main() { long int t; cin>>t; while(t--) { int n,c=0,t=0; scanf("%d",&n); While(n>=0) { t=n%10; if(t==4) { c++; } n=n/10; } cout<<c<<endl; } return 0; } Why this code is taking input for more times than that t value

#include <iostream> using namespace std; int main() { long int t; cin>>t; while(t--) { int n,c=0,t=0; scanf("%d",&n); While(n>=0) { t=n%10; if(t==4) { c++; } n=n/10; } cout<<c<<endl; } return 0; } Why this code is taking input for more times than that t value

If you enter t as 3 but it getting 4 inputs ?

#include<iostream> #include<string> using namespace std; class book{     public:         int price;         string name; }; int main(){     book a,b;     a.price=1;     b.price=1;     auto add=[](book& a,book& b)->decltype(a){         book c;         c.price=a.price+b.price;         return c;     };     book cc=add(a,b);     cout<<"cc price is"<<cc.price<<endl;     return 0; }

Why have you put ->decltype(a) over there in lambda?

You could have just used, book instead of that

also if you insist on using decltype, that's not the way

have a look at std::declval

I found the error. The compiler think the a in decltype(a) is a in (book& a,book& b),not the a I defined earlier in line one “book a,b” .

I found the error. The compiler think the a in decltype(a) is a in (book& a,book& b),not the a I defined earlier in line one “book a,b” .

yes

#include<iostream> #include<string> using namespace std; class book{     public:         int price;         string name; }; int main(){     book a,b;     a.price=1;     b.price=1;     auto add=[](book& a,book& b)->decltype(a){         book c;         c.price=a.price+b.price;         return c;     };     book cc=add(a,b);     cout<<"cc price is"<<cc.price<<endl;     return 0; }

This is wrong because your lambda will return a reference (decltype(a) is book&). It doesn't matter whether you use the parameter a or the variable a in main. The return type will be a reference and returning a reference to the local variable c is a bad idea because c will be destroyed after the lambda returns

have a look at std::declval

The problem is not with decltype. And you shouldn't be using declval either here The problem can be fixed by either simply returning book or by using std::decay_t on the type returned from decltype

The problem is not with decltype. And you shouldn't be using declval either here The problem can be fixed by either simply returning book or by using std::decay_t on the type returned from decltype

Yeah right, declval is not necessary here

#include <iostream> using namespace std; int main() { long int t; cin>>t; while(t--) { int n,c=0,t=0; scanf("%d",&n); While(n>=0) { t=n%10; if(t==4) { c++; } n=n/10; } cout<<c<<endl; } return 0; } Why this code is taking input for more times than that t value

While(--t)?

But if I define “book aa” in main and use decltype(aa) It tends to be no error

But if I define “book aa” in main and use decltype(aa) It tends to be no error

Because then decltype(aa) would be book and not book&.

The problem is not with decltype. And you shouldn't be using declval either here The problem can be fixed by either simply returning book or by using std::decay_t on the type returned from decltype

The problem here can be solved by simply omitting the return type

Hello everyone I'm Sasikumar https://www.linkedin.com/in/duttaluru-sasikumar-9477ab188 Everyone let me connect in LinkedIn.

The problem here can be solved by simply omitting the return type

Omitting the return type will not work if he is compiling on a C++11 system and if there are multiple return statements.

Omitting the return type will not work if he is compiling on a C++11 system and if there are multiple return statements.

С++11 is legacy

С++11 is legacy

Lots of companies still use a C++11 or a C++14 compiler. The move to C++17 itself hasn't happened yet

Lots of companies still use a C++11 or a C++14 compiler. The move to C++17 itself hasn't happened yet

llvm uses C++14 Google uses C++14

nothing stops from using c++17 or newer

you can use c++17 and c++11 at once (i mean you don't have to rewrite the legacy code to C++17 to use C++17)

I don't remember the new standards relaxing the existing restrictions.

But I know many examples in which the new standards even make restrictions stricter

llvm uses C++14 Google uses C++14

No. Big companies don't migrate easily to the latest and the greatest. If you think they do, then you haven't worked in one. The problem is not that the compilers may have ABI compatability issues. The problem is that migration takes a long time and it has to be tested thoroughly. They can't take your word for it. At the end of the day it's a business and they would be happy to use an older version as long as it serves them. Google uses different versions for different teams. Fuschia OS team uses C++17 because when they started out many compilers already were offering C++17 support . The tensor flow team still uses C++11 as does the Nano team in GCP devops tools. C++11 is still being widely used as is C++14. C++17 is yet to see adoption at the same scale which would most likely happen soon as C++17 has had sufficient time to mature and be tested.

#include <stdio.h> unsigned int factorial( unsigned int n); int main () { int number = 5; int z =factorial( number ); printf("factorial of %d = %d \n",number,z); } unsigned int factorial( unsigned int n) { if(n==0) return 1; return n * factorial(n-1) ; }

#include <stdio.h> unsigned int factorial( unsigned int n); int main () { int number = 5; int z =factorial( number ); printf("factorial of %d = %d \n",number,z); } unsigned int factorial( unsigned int n) { if(n==0) return 1; return n * factorial(n-1) ; }

Which c++ smtp client you guys would prefer to send email? I am translating libcurl examples but it’s c

Which c++ smtp client you guys would prefer to send email? I am translating libcurl examples but it’s c

Poco is a pretty common one by also pretty Java-ish as far as C++ is concerned

in C++, is there any kind of mutex that works like in Rust? in Rust we can only access the data by locking the mutex, and its guaranteed to be unlocked at the end of the scope

Poco is a pretty common one by also pretty Java-ish as far as C++ is concerned

Thanks, I got libcurl working now

guys

I have a problem, who can help me solve it?

I have a problem, who can help me solve it?

What's up?

Hi, how to make write function show integers in C? Any idea?

Hi, how to make write function show integers in C? Any idea?

what?

write function from the library unistd.h