Extra Sleeping Info...

When using the way of sleeping that uses permanent loops, where does the number of seconds to sleep for go?

Also, when I try to compile (although it may be because I don't know where to enter the seconds) my compiler tells me that:

FILE NAME - CONFIDENTIAL: 56: `now'
undeclared (first use this function)

FILE NAME - CONFIDENTIAL: 56: `
seconds' undeclared (first use this function)

FILE NAME - CONFIDENTIAL: At global
scope:

FILE NAME - CONFIDENTIAL: 61: parse
error before `else'

Post the code.

Post some code
Don't forget to read the posting guidelines at the top of the forum

#include <time.h>

void sleep_seconds ( long seconds ) {
clock_t limit, now = clock();
limit = now + seconds * CLOCKS_PER_SEC;
while ( limit > now )
now = clock();
}


this is the code that i am having trouble with...

Yours compiles correctly on my machine. Post the entire code, it looks like there might by a global variable name conflict.

So much for mentioning the posting code notice....

Your code compiles fine here with gcc, which compiler are you using?

Quote:

---

...I don't know where to enter the seconds...

---

Do you have a main() function? All C++ programs need main().

Then, to call your sleep_seconds() function for 100 seconds, you would put this in main():

sleep_seconds(100);

Or,
sleep_seconds(x);

Actually, you can call your sleep_seconds() function from any other function. You don't have to call it from main().... But, let's assume that your program isn't that complicated. :)

just a general question: is it better to use CLOCKS_PER_SEC or to just know what you're using and use your own constant?

which is better:

Code:

---

limit = now + seconds * CLOCKS_PER_SEC;

---

or

Code:

---

limit = now + seconds * 1000;

---

I use the latter in most of my coding...

> I use the latter in most of my coding...
And yet the former is better...

>>And yet the former is better...

that's what my question is... but that still leaves the obvious: why is it better?

It's more portable. On some systems, CLOCKS_PER_SECOND may not be 1000.

So, it's just a good habit to make all your programs as standard and portable as possible.

On most systems, the execuitable is going to be the same using 1000, or CLOCKS_PER_SECOND. So, the final program won't be "better". But the source code will be more standardized.

Quote:

---

Originally Posted by ale_jrb

#include <time.h>

void sleep_seconds ( long seconds ) {
clock_t limit, now = clock();
limit = now + seconds * CLOCKS_PER_SEC;
while ( limit > now )
now = clock();
}


this is the code that i am having trouble with...

---

Replace <time.h> with <ctime> and add "using namespace std;" as the first line in your function.