From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail4-relais-sop.national.inria.fr (mail4-relais-sop.national.inria.fr [192.134.164.105]) by walapai.inria.fr (8.13.6/8.13.6) with ESMTP id pBEGtfHm009333 for ; Wed, 14 Dec 2011 17:55:41 +0100 X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AvwAAM7U6E7B/BfSkWdsb2JhbABDqGGCaAEBAQEJCwsHFAMigXIBAQQBOEABBQsLGAkWDwkDAgECAUUGDQEHAQGHdgK3PYN5iBAElHSFTYxe X-IronPort-AV: E=Sophos;i="4.71,353,1320620400"; d="scan'208";a="123371155" Received: from msa01.smtpout.orange.fr (HELO msa.smtpout.orange.fr) ([193.252.23.210]) by mail4-smtp-sop.national.inria.fr with ESMTP; 14 Dec 2011 17:55:35 +0100 Received: from [192.168.1.102] ([83.199.117.71]) by mwinf5d49 with ME id 94va1i0031YWU4b034va35; Wed, 14 Dec 2011 17:55:35 +0100 Message-ID: <4EE8D506.1040800@frisch.fr> Date: Wed, 14 Dec 2011 17:55:34 +0100 From: Alain Frisch User-Agent: Mozilla/5.0 (X11; Linux i686 on x86_64; rv:8.0) Gecko/20111105 Thunderbird/8.0 MIME-Version: 1.0 To: Adrien CC: Jonathan Protzenko , Martin DeMello , Gerd Stolpmann , caml-list@inria.fr References: <4EDE33A0.6070004@gmail.com> <1323760512.9833.9.camel@samsung> <4EE711FB.5020602@frisch.fr> <4EE83C26.7090108@frisch.fr> <4EE86D90.6080409@gmail.com> <4EE87976.4030604@frisch.fr> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: [Caml-list] Some comments on recent discussions On 12/14/2011 02:37 PM, Adrien wrote: > Actually, I think that you should have used the "/etc/alternatives" > symlinks: /usr/bin/gcc points to /etc/alternatives/FOO and you can make this > FOO symlink point to the /usr/bin/BAR binary that you want. The problem is that flexlink.exe (and ocamlopt.exe) are Win32 executables. They cannot follow Cygwin symlinks. Of course, I had /etc/gcc symlinked to gcc-3.exe through /etc/alternatives, but it did not work. > I don't think it would be possible to live without a C toolchain simply > because we use C libraries all the time. It depends on who is "we". I can imagine that library developers still need a C toolchain but release binary packages that don't. > I'm quite interested in the ability to create .cmxs files without a C > compiler and can already picture me using it. I've also noticed Benedikt's > ocamlnat work. Would it be usable to script native-code applications? > Maybe with less requirements? FWIW, LexiFi's application is distributed together with flexlink.exe and ocamlopt.exe, and it can recompile and dynamically load user-defined plugins without any other external tool. (Our clients don't need to install anything else to write, compile and run native OCaml code.) Benedikt's work on ocamlnat also includes a similar direct code generator as ours(to avoid the external assembler); I don't think it comes with a COFF file emitter, though. But yes, ocamlnat can be used to script native-code applications. Alain