Another thing Günter did in PDa is introduce t_time for the clock functions instead of using 'double' directly. Any reason not to include t_time in vanilla?
+typedef long long t_time; EXTERN t_clock *clock_new(void *owner, t_method fn); -EXTERN void clock_set(t_clock *x, double systime); -EXTERN void clock_delay(t_clock *x, double delaytime); +EXTERN void clock_set(t_clock *x, t_time systime); +EXTERN void clock_delay(t_clock *x, t_time delaytime); EXTERN void clock_unset(t_clock *x); -EXTERN double clock_getlogicaltime(void); -EXTERN double clock_getsystime(void); /* OBSOLETE; use clock_getlogicaltime() */ -EXTERN double clock_gettimesince(double prevsystime); -EXTERN double clock_getsystimeafter(double delaytime); +EXTERN t_time clock_getlogicaltime(void); +EXTERN t_time clock_getsystime(void); /* OBSOLETE; use clock_getlogicaltime() */ +EXTERN t_time clock_gettimesince(t_time prevsystime); +EXTERN t_time clock_getsystimeafter(t_time delaytime); EXTERN void clock_free(t_clock *x);
.hc
----------------------------------------------------------------------------
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hans-Christoph Steiner wrote:
Another thing Günter did in PDa is introduce t_time for the clock functions instead of using 'double' directly. Any reason not to include t_time in vanilla?
can't think of one.
fgmasdr IOhannes