[mercury-users] SEC: UNCLASSIFIED:-

[Fergus Henderson]

On 29-Oct-2002, Ian.Sykes at defence.gov.au <Ian.Sykes at defence.gov.au> wrote:
> I want to port some turbo prolog code to Mercury.
> 
> The code was written on a DOS 6/i386 system to do real-time control (don't
>        panic - it's non-commercial general interest stuff) .
> 
> The code requires elapsed time, so the ElapsedRealTime part of
>        'time__times(ProcessorTime,ElapsedRealTime,IO_state,IO_state)'
>        appears ideal.
> 
> Can those in the know confirm that ElapsedRealTime delivers genuine
>        centiseconds as opposed to an in interpolation
> of some larger clock interval, as was the case with the DOS/386 platform.

The accuracy of Mercury's time__times procedure is determined by
the accuracy of the C function times() which is used to implement it.

The ElapsedRealTime returns a value in "clock ticks".
These are not guaranteed to be centiseconds.

I think that on some systems, the granularity of the timer may be more
than one "clock tick".

> Is there a better way of getting a more precise time interval?

Probably, but it depends on which platform you're using.
Which platform are you using?  I take it you're not still
running this on a DOS 6 system? ;-)

On most platforms, there are C library functions for this sort of thing
which you can access from Mercury using the C interface.

-- 
Fergus Henderson <fjh at cs.mu.oz.au>  |  "I have always known that the pursuit
The University of Melbourne         |  of excellence is a lethal habit"
WWW: <http://www.cs.mu.oz.au/~fjh>  |     -- the last words of T. S. Garp.
--------------------------------------------------------------------------
mercury-users mailing list
post:  mercury-users at cs.mu.oz.au
administrative address: owner-mercury-users at cs.mu.oz.au
unsubscribe: Address: mercury-users-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-users-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------