Is there any preference when sending data as to send it byte by byte or sending
a whole buffer (the whole file) at once?
Is the latter faster than the former for example?
This is a discussion on Send a byte or whole text within the Networking/Device Communication forums, part of the General Programming Boards category; Is there any preference when sending data as to send it byte by byte or sending a whole buffer (the ...
Is there any preference when sending data as to send it byte by byte or sending
a whole buffer (the whole file) at once?
Is the latter faster than the former for example?
Using Code::Blocks with MSVC++ 2010.
You can pass as many bytes as you like to send(), but you'd better pay attention to the return result.
If it's less than the amount you asked to be sent, then you need to call send() again, from the point where the previous attempt left off.
A bit like this.
Code:size_t num_to_send; char *buff; while ( num_to_send ) { int num_sent = send( sock, buff, num_to_send, 0 ); if ( num_sent > 0 ) { num_to_send -= num_sent; buff += num_sent; } }
If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.
If at first you don't succeed, try writing your phone number on the exam paper.
I support http://www.ukip.org/ as the first necessary step to a free Europe.
In addition to what Salem said: send as much as you have available to send, but obviously try to keep the size of what you must send small (if you have control over the protocol/its contents) - ie, a PNG or a GIF is better than a BMP.
recv as much as you can recv.
long time; /* know C? */
Unprecedented performance: Nothing ever ran this slow before.
Any sufficiently advanced bug is indistinguishable from a feature.
Real Programmers confuse Halloween and Christmas, because dec 25 == oct 31.
The best way to accelerate an IBM is at 9.8 m/s/s.
recursion (re - cur' - zhun) n. 1. (see recursion)
Thank you Salem and Cactus_Hugger!
I see you use a char *buff; in your code.
Does it matter as to which one of the 3 we use?
Code:char *buffer; char buffer[256]; char *buffer = new char[256];
Last edited by Ducky; 08-09-2009 at 05:31 AM.
Using Code::Blocks with MSVC++ 2010.
So long as it's allocated, no.
If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.
If at first you don't succeed, try writing your phone number on the exam paper.
I support http://www.ukip.org/ as the first necessary step to a free Europe.
May i ask whats the point of this line in your code Salem?
I suppose its for decreasing the length of buff.Code:buff += num_sent;
But if i do a test like this, i get completely random values for buf:
Code:#include <iostream> using namespace std; int main() { int i = 0; char * buf = new char[i]; while(i<15) { buf += i; i++; cout << strlen(buf) <<"\n"; } delete [] buf; return 0; }
Using Code::Blocks with MSVC++ 2010.
> buff += num_sent;
If you've got a buffer of 100 bytes, and send 10, that means you've got 90 left (that's what the -= is for).
It also means that you need to try the next send from 10 bytes into the buffer (that's what the += is for).
If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.
If at first you don't succeed, try writing your phone number on the exam paper.
I support http://www.ukip.org/ as the first necessary step to a free Europe.
Ah, ok thanks!![]()
Using Code::Blocks with MSVC++ 2010.