Message-ID:	<19981112103840.8972.qmail@hotmail.com>
X-Originating-IP:	[195.44.9.66]
From:	"a j" <aj664 AT hotmail DOT com>
To:	djgpp AT delorie DOT com
Subject:	Re: allegro and pctimer?
MIME-Version:	1.0
Date:	Thu, 12 Nov 1998 02:38:40 PST
Reply-To:	djgpp AT delorie DOT com

>From: Diego Dominguez Cazorla <ddomingu AT mlg DOT cit DOT alcatel DOT es>
>To: "'djgpp'" <djgpp AT delorie DOT com>
>Subject: allegro and pctimer?
>Date: Thu, 12 Nov 1998 08:45:16 +0100
>Reply-To: djgpp AT delorie DOT com
>
>	Hello!!!
>
>	I have a program that does some processing stuff and another that acts 
as 
>a front-end. The front-end uses allegro (and bgui ;) and calls (with 
>system(..) ) the processing one.
>
>	Inside the processing program I use only allegro's timer again to 
measure 
>time used by each processing algorithm and write it down to a text file 
to 
>allow the front end to show them. (Of course I disable timer in the 
>front-end before calling it, and enable it afterwards).
>
>	The problem is that it seems to me that allegro's timer is not enough 
>precise (number shown change a bit each execution, even after several 
times 
>to have cache stabilized). I tried with pctimer and things seemed to be 
>better, but it WOULD hang up the system if called from the front-end.
>
>	Do you know if it is possible to have alllegro, and (with allegro's 
timer 
>disabled) have a another timer to do the measuring? Do you know how 
exact 
>is allegro's timer?



try using the uclock()  from the standard C lib that comes with DJGPP
its resolution is about 1 million and something ticks per second.


______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com

- Raw text -