From mboxrd@z Thu Jan 1 00:00:00 1970 Received: (from majordomo@localhost) by pauillac.inria.fr (8.7.6/8.7.3) id WAA03117; Sat, 24 Mar 2001 22:49:09 +0100 (MET) X-Authentication-Warning: pauillac.inria.fr: majordomo set sender to owner-caml-list@pauillac.inria.fr using -f Received: from concorde.inria.fr (concorde.inria.fr [192.93.2.39]) by pauillac.inria.fr (8.7.6/8.7.3) with ESMTP id WAA03484 for ; Sat, 24 Mar 2001 22:49:08 +0100 (MET) Received: from scaup.prod.itd.earthlink.net (scaup.prod.itd.earthlink.net [207.217.121.49]) by concorde.inria.fr (8.11.1/8.10.0) with ESMTP id f2OLn7T11200 for ; Sat, 24 Mar 2001 22:49:07 +0100 (MET) Received: from SFOWVCHAK1L7 (pool0572.cvx7-bradley.dialup.earthlink.net [209.178.166.62]) by scaup.prod.itd.earthlink.net (EL-8_9_3_3/8.9.3) with SMTP id NAA19015; Sat, 24 Mar 2001 13:48:58 -0800 (PST) Message-ID: <000401c0b4ad$37dbd6b0$3ea6b2d1@SFOWVCHAK1L7> Reply-To: "Vijay Chakravarthy" From: "Vijay Chakravarthy" To: "Brian Rogoff" Cc: , "Arturo Borquez" , References: Subject: Re: [Caml-list] Why People Aren't Using OCAML? (was Haskell) Date: Sat, 24 Mar 2001 13:55:59 -0800 Organization: VerticalNet Solutions MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.50.4133.2400 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Sender: owner-caml-list@pauillac.inria.fr Precedence: bulk ----- Original Message ----- From: "Brian Rogoff" To: "Vijay Chakravarthy" Cc: ; "Arturo Borquez" ; Sent: Saturday, March 24, 2001 10:38 AM Subject: Re: [Caml-list] Why People Aren't Using OCAML? (was Haskell) > On Sat, 24 Mar 2001, Vijay Chakravarthy wrote: > > We are also trying to use Ocaml out here, mainly for prototyping purposes > > (at least thats the way I'll introduce it to the rest of the org.) > > Once they find that rewriting the prototypes in java takes 10 times the > > effort..... > > And runs at one tenth the speed :-). I laughed at this since one of my > Caml-hating colleagues recently suggested to my manager (who reads > this list :) that we rewrite some Java code in a low level language like C > or OCaml (!) for speed. > > > The main difficulties we have faced have been -- > > a) Education - Out here even for people skilled in scheme and some other > > functional languages, it is challenging to learn ocaml. > > What have the issues been? I hate to mention this again (and again, ...) but > I find that beginners and non-experts stumble over syntax a lot. Now, it is > clear that Daniel De Rauglaudre is too shy and modest to mention his Revised > syntax available with CamlP4. If your people have syntax problems (they do > go away with familiarity IME) then check it out. > I personally dont have problems with the syntax, but I have found that people take a little bit of time getting used to the syntax. The simplest way to fix this is not CamlP4 (IMHO), but rather to provide a lot of examples showing people how to do things they would commonly do. The three liner Markus posted in c.l.f for an echo server, for example, helps get buy-in when trying to convince people this is a cool language, and that happens primarily through the use of examples. I also understand that there are a lot of examples scattered around, but it would be nice to collect all of them in an easy to look at manner. I believe that Fredrick Lundh's python ematter book is a good role model in this area. > > My team found it easier to learn erlang, for example. However, people > > love the strong type checking, and the fact that if it compiles its > > likely to run correctly... The english translation of the horse book should > > help. > > Yes, how is that coming? I hope that ugly horse can get replaced by a > beautiful dromedary. > > > b) As I mentioned in a previous message, support for Windows is weak. Many > > This might be the kind of thing that the Consortium can help with. If > enough members want Windows they can kick in for an extensive thin binding > to Windows services. > Its not that we need binding to Windows services. All that is required is that the various packages are as easy to make and use on windows (without installing cygwin). This is mostly true, except for some critical packages like findlib, dbm, etc. A dbm interface on windows to Sleepycats berkeley db would be great. > > c) Package availability is fragmented. I have a good idea of various > > packages etc, but to a newcomer, there is no single place to > > track down code examples. Plus larger packages like Ensemble, Geneweb etc > > contain modules that would be useful in general, but > > that is visible only when one peruses the source code of such packages. > > Two issues there. First is that some tool like findlib or the Python distutils > should be part of the toolset. Second is that a big library structure like > the SML Basis library would be helpful. > > > On the other hand, the language and the libraries are excellent. > > Agreed. > > > Plus functional programming languages are IDEAL for the type > > of work we do, which is in the enterprise software space. > > I would have said that it's ideal for VLSI design software (any other EDA > hackers out there?) so maybe it's just ideal for everything? > > > BTW, I dont know if this is the right place for this, but are there any > > people interested in ocaml programming out here in the > > San Francisco area? > > Santa Clara (which hasn't had any blackouts!) area but there was an > attempt at a NoCal OCaml users meeting. Maybe the participants can tell > you how it went. It's great to hear that the number of industrial OCaml > programmers is growing! > > -- Brian Actually, we have two offices, one in SFO, the other in Palo Alto (right next to Xerox Parc). Vijay ------------------- To unsubscribe, mail caml-list-request@inria.fr. Archives: http://caml.inria.fr