No subject

Fergus Henderson fjh at cs.mu.OZ.AU
Sun Feb 14 22:29:16 AEDT 1999


Received: from mail5.microsoft.com (mail5.microsoft.com [131.107.3.121]) by mulga.cs.mu.OZ.AU with ESMTP
	id AAA06291; Sat, 13 Feb 1999 00:13:01 +1100 (EST)
Received: by INET-IMC-05 with Internet Mail Service (5.5.2524.0)
	id <1WGTPXGN>; Fri, 12 Feb 1999 05:12:59 -0800
Message-ID: <39ADCF833E74D111A2D700805F1951EF0F00B9A7 at RED-MSG-06>
From: Don Syme <dsyme at microsoft.com>
To: "'mercury-users at cs.mu.OZ.AU'" <mercury-users at cs.mu.OZ.AU>
Cc: "'conway at cs.mu.oz.au'" <conway at cs.mu.OZ.AU>
Subject: RE: [mercury-users] Threads
Date: Fri, 12 Feb 1999 05:12:58 -0800
X-Mailer: Internet Mail Service (5.5.2524.0)
Status: O
Content-Length: 1760
Lines: 49


Are there any docs on this?  Will Mercury threads map down to Win32 threads
on Windows systems (I presume so).

There's no rush - I'll look forward to the release.  I plan to spend some
time later this year comparing, contrasting and enhancing support for Win32
programming with Mercury, Haskell and Ocaml.  Threaded interface programming
will be a significant part of this.

Thanks & cheers,
Don

> 
> 
> On Thu, Feb 11, 1999 at 09:43:31PM EST, Ralph Becket wrote:
> > What's the status of the work on adding threads to Mercury?
> 
> RSN. The runtime engine can be compiled to be thread-safe (using
> a grade containing .par, eg asm_fast.gc.par). I have a bunch of
> stuff which will go into the extras directory including 'spawn':
> 
> 	:- pred spawn(pred(io__state, io__state), io__state, io__state).
> 	:- mode spawn(pred(di, uo) is cc_multi, di, uo) is cc_multi.
> 
> which starts up the given closure as a coroutine. There is also
> stuff for semaphores, etc.
> 
> Probably won't make it in until mid march - now rather than being
> a self-determining postgrad, I'm a tutor in the department, and my
> teaching committments will be taking all my time for the next little
> while.
> 
> Thomas
> -- 
> Thomas Conway <conway at cs.mu.oz.au> )O+
> To a killer whale, otters are like hairy popcorn -- Paul Dayton



More information about the users mailing list