Newsgroups: comp.lang.dylan
From: cyber_surfer@wildcard.demon.co.uk (Cyber Surfer)
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!gatech!swrinde!pipex!peernews.demon.co.uk!wildcard.demon.co.uk!cyber_surfer
Subject: Re: Dylan Availability [was: QuickSort function?]
References: <dsawyer-0903951150130001@ws1_101.randomaccess.com> <2878291254@hoult.actrix.gen.nz>
Organization: The Wildcard Killer Butterfly Breeding Ground
Reply-To: cyber_surfer@wildcard.demon.co.uk
X-Newsreader: Demon Internet Simple News v1.27
Lines: 43
X-Posting-Host: wildcard.demon.co.uk
Date: Sat, 18 Mar 1995 16:57:02 +0000
Message-ID: <795545822snz@wildcard.demon.co.uk>
Sender: usenet@demon.co.uk

In article <2878291254@hoult.actrix.gen.nz>
           Bruce@hoult.actrix.gen.nz "Bruce Hoult" writes:

> 1) a Mac toolbox interface or class library so I can write real programs.

That's a platform specific issue, and a platform I don't use,
so I can't comment on it.
 
> 2) speed.  Mindy and Marlais run at about the same speed as, say, interpreted
>    MacGambit.  Apple claim that Dylan will be as fast as C++.  I believe them,
>    and I'd like to try things in Dylan that rely on that speed.  Patrick Beard
>    (who I'd like to thank profusely for his work) notes in the MacMarlais
>    "0.5 CHANGELOG" file that the PowerPC version of a recursive fib(13) takes
>    two seconds vs eleven seconds in 68000 emulation.  Well, Metrowerks
>    Codewarrior C++ does the same thing more then 5,000 times faster.  That's
>    quite a speed difference to close.

Is MacMarlais a native code compiler? I don't know the C++ compiler         
you used, so I can't comment on it, but I seem to recall that some
of the "Dylan" implementations available at the moment are interpreters.
I'm pretty sure that I remember reading that Mindy is a byte-coded
implementation, in which case you'd need to compare it with a byte-coded
C++ compiler for a fair comparison.

In any case, these Dylan implementations are not the thing as the
Dylan implementations that will be available as commercial products.
The biggest difference from my point of view is that the Dylan language
is still a moving target, which I believe is the reason why Mindy is
short for Mindy Is Not Dylan Yet.

I'll be happy to use any Dylan implementation I can get my hands on,
whether it'll use a byte-code interpreter or compile to native code.
However, I'm reserving judgement of the language until then, and even
then, I'll judge it based on how fast it helps me code, not how that
code runs. The language is the issue for me, not the implementation.
If the language is useful to me, but the code is not fast enough, then
I'll look for a faster implentation. I'm sure I'll find one.

It's times like that I like to check the FAQ for a newsgroup, so I
can have an up to date list of implementations etc. If I've made any
incorrect assumptions or misremembered anything, then I apologise.
-- 
CommUnity: <URL:http://www.demon.co.uk/community/index.html>
