[PHP-DEV] RFC: uniqid default param

2003-02-10 Thread Marcus Börger
Hi,

the current implementation of uniqid set the more entropy default
true for CYGWIN and false for the rest. CYGWIN must use more
entropy because it does not produce new values after usleep(1)
necessarily. However usleep(1) should nowadays be very slow
compared to whatever php_combined_lcg() needs to do.

Shall more entropy be always true?

marcus


--
PHP Development Mailing List http://www.php.net/
To unsubscribe, visit: http://www.php.net/unsub.php




Re: [PHP-DEV] RFC: uniqid default param

2003-02-10 Thread Sascha Schumann
On Mon, 10 Feb 2003, Marcus Börger wrote:

 Hi,

 the current implementation of uniqid set the more entropy default
 true for CYGWIN and false for the rest. CYGWIN must use more
 entropy because it does not produce new values after usleep(1)
 necessarily. However usleep(1) should nowadays be very slow
 compared to whatever php_combined_lcg() needs to do.

 Shall more entropy be always true?

No, some users might depend on the return format (think of
database entries).

- Sascha

--
PHP Development Mailing List http://www.php.net/
To unsubscribe, visit: http://www.php.net/unsub.php




Re: [PHP-DEV] RFC: uniqid default param

2003-02-10 Thread Marcus Börger
At 23:08 10.02.2003, Sascha Schumann wrote:

   No, some users might depend on the return format (think of
database entries).


That was one of the reasons i only changed it for CYGWIN.


--
PHP Development Mailing List http://www.php.net/
To unsubscribe, visit: http://www.php.net/unsub.php