From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: weis Received: (from weis@localhost) by pauillac.inria.fr (8.6.10/8.6.6) id NAA25867 for caml-redistribution; Mon, 10 Jul 1995 13:28:06 +0200 Received: from concorde.inria.fr (concorde.inria.fr [192.93.2.39]) by pauillac.inria.fr (8.6.10/8.6.6) with ESMTP id MAA25728 for ; Mon, 10 Jul 1995 12:53:30 +0200 Received: from epsilon.qmw.ac.uk (epsilon.qmw.ac.uk [138.37.6.3]) by concorde.inria.fr (8.6.10/8.6.9) with SMTP id MAA24790 for ; Mon, 10 Jul 1995 12:57:04 +0200 Received: from canary.dcs.qmw.ac.uk by epsilon.qmw.ac.uk with SMTP-DNS (PP) id <05774-0@epsilon.qmw.ac.uk>; Mon, 10 Jul 1995 11:56:32 +0100 Received: from aux23.dcs.qmw.ac.uk [192.135.231.58] by canary.dcs.qmw.ac.uk (8.6.12/QMW-server-2.4s) with SMTP; poster "Richard.Bornat@dcs.qmw.ac.uk"; id LAA00966; Mon, 10 Jul 1995 11:55:12 +0100 X-Sender: richard@mailhost.dcs.qmw.ac.uk Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Mon, 10 Jul 1995 11:49:06 +0100 To: caml-list@pauillac.inria.fr From: Richard.Bornat@dcs.qmw.ac.uk (Richard Bornat) Subject: Compiling caml-light for the Mac Cc: keithc@dcs.qmw.ac.uk, sufrin@comlab.ox.ac.uk Sender: weis Bernard Sufrin and I have a proof editor - Jape - which was written in SML/NJ for SunOS and A/UX. I want to port it to MacOS now, and to Windows-based PCs as soon as maybe. I have already translated it into Moscow ML, and am enormously impressed by the speed of translation, the small size of the bytecodes, and so on. But speed of execution is disappointing - about six times slower than NJ. This doesn't matter most of the time, when the user is making small proof steps, because all they see is a slightly longer delay between steps. But we have some compute-intensive phases - for example, reading in a file of previously completed proofs and re-incorporating them into the system - where speed is essential. Mosml 1.30 isn't yet available for the Mac, though I know it's coming. Caml-light is ready, and it has compilation tools, and a profiler, and a large body of users, and I don't know what other wonderful features. To move now from Moscow ML to caml-light would be a relatively small step. I could even do it mechanically, as a first approximation. But would that make it possible to compile my program into fast-running C on the Mac? on Windows PC? or fast-running anything on either? I note that both camlot and camloo/bigloo know nothing about the alien environment of the Mac. Is the C they produce Mac-friendly, or is it very difficult to make it fit? So far as I know, nobody has experience of caml-light -> C for the Mac. (I should be delighted to be contradicted on this point!). I have MPW and CodeWarrior for MacOS. I could run camlot / camloo on a Sun and, through the miracle of fileserving, compile it with the Mac tools. Richard Bornat Department of Computer Science richard@dcs.qmw.ac.uk Queen Mary and Westfield College phone (+44) 171 975 5206 direct line Mile End Road (+44) 171 975 5200 department LONDON E1 4NS fax (+44) 181 980 6533 department UK