But modules cannot do anything with it, you just cannot call an abstract callable object before it is known
That's what I asked for
And you gave your answer
MᏫᎻᎯᎷᎷᎬᎠ
Ilya
Like being an "implicit declaration" or sth
Implicit declarations are valid in pure C mode, but time has proven they are harmful, already
MᏫᎻᎯᎷᎷᎬᎠ
Implicit declarations are valid in pure C mode, but time has proven they are harmful, already
Why they are harmful?!🤔
MᏫᎻᎯᎷᎷᎬᎠ
Hmmm
GeekyLOAF
Hi
MᏫᎻᎯᎷᎷᎬᎠ
Hi
Artöm
Why they are harmful?!🤔
You see error on linking stage, but could see it on compile or even code editing stage
GeekyLOAF
Welp
Artöm
And you also can link wrong function
GeekyLOAF
I don't like talking Alot
Ilya
GeekyLOAF
Name: Mohammed
Age: 19
Location: Yemen
Artöm
Test src or stuff
MᏫᎻᎯᎷᎷᎬᎠ
Function with same signature may accidentally be in other tu
And what that has to do with declaration?!
GeekyLOAF
We don't need that here
It's kind of introduction I should keep a good image of myself
Ilya
Really?
Yes, you have function with one signature which is unknown at compilation time, it is called as a func with a different signature, and some parameters are just have wrong memory address or don't get destroyed after the function call.
GPF is the best what you can get.
Artöm
And what that has to do with declaration?!
I was wrong, implicit declaration at call point is outdated C feature
Ilya
I was wrong, implicit declaration at call point is outdated C feature
Yes and do you know WHY it is outdated?
Larry
How can you read 3 .csv files and shuffle them into two
MᏫᎻᎯᎷᎷᎬᎠ
I think there is a high chance to standardize a package manager for C++20
Francisco
I think there is a high chance to standardize a package manager for C++20
There isn't a plan for a standard package manager. There will be modules, which will allow making easier
Larry
Could someone show me the code using fstream
Artöm
I think there is a high chance to standardize a package manager for C++20
Chance is zero. I dont remember any proposal for package manager and C++20 is feature frozen
MᏫᎻᎯᎷᎷᎬᎠ
There isn't a plan for a standard package manager. There will be modules, which will allow making easier
That's the reason
Modules make packaging easier
MᏫᎻᎯᎷᎷᎬᎠ
But I think adding a package manager to C++ is not something related to (3 years cycle)
They can add it in 2021 or later
MᏫᎻᎯᎷᎷᎬᎠ
Or maybe standardizing an existing one
Bojan_Krdemn
hey, what is the best ide for very simple tasks?
Bojan_Krdemn
console output
Bojan_Krdemn
win 10, don't like compiling and running from command prompt (tried to learn spent day gave up)
Ilya
I think there is a high chance to standardize a package manager for C++20
There is not even a hint for it, although JFrog are in the committee.
For sure it will not be in 20th because there has been feature freeze for 20 already
Ilya
How can you read 3 .csv files and shuffle them into two
read all files line by line, shuffle lines and then write.
Ilya
hey, what is the best ide for very simple tasks?
The best IDE for simple tasks is CMake + your favorite text editor
klimi
win 10, don't like compiling and running from command prompt (tried to learn spent day gave up)
If you don't like compiling then don't use compiled language
klimi
No
Ilya
Oh No not Eclipse!
Talula
Ilya
Can you read?
Talula
https://code.visualstudio.com/docs/cpp/config-mingw
Bojan_Krdemn
Downloaded vsc
Talula
You don't have to compile on command line...
Anonymous
👌
Sergey
Hello, Could please someone help me with cmake? I'm interested is that possible to create Clean Architecture with cmake. In project I have 3 layers (data, domain, presentation). Data and presentation layers depends on 3rd librarys and freamwork. Domain is based on pure c++. Can project split into independed layers with cmake with public interfaces?
Anonymous
Anonymous
for college assignments. thank you
Dima
Pavel
For a moment I thought it is carved on a stone wall
Anonymous
Hello guyzz...is here any one use unity 3d
Anonymous
How long r u using
Anonymous
Hello guyzz...is here any one use unity 3d
/warn it's C/C++ chat and Unity is definitely an offtop here
Anonymous
Oh sry😊
Anonymous
hi
klimi
hi
klimi
what do you need rebel
I_Interface
klimi
xD
olli
Why they are harmful?!🤔
C had to use forward declarations because computers were slow and doing another lookup was too memory or cpu intensive.
C++ wanted to be kinda "compatible" and followed the same principle.
But then C++ also needs to support things like this, where not requiring forward declarations might alter the behavior.
#include <iostream>
using std::cout;
template<typename T>
void adl(T) {
cout << "T";
}
struct S { };
template<typename T>
void call_adl(T t) {
adl(S());
adl(t);
}
void adl(S) {
cout << "S";
}
int main () {
call_adl(S());
}
MᏫᎻᎯᎷᎷᎬᎠ
and Also
I got something for you :)
https://vittorioromeo.info/index/blog/fixing_cpp_with_epochs.html
olli
And it worked perfectly without declaring adl(S) above
yeah it does, but if you consider adl(S) implicitely declared the behavior is different.
e.g. if you add replace struct S {}; with
struct S{};
void adl(S);
If you do so both adl(t); and (adl(S()); will call the non template function.