Related articles |
---|
Mathematica-like Language Implementation lojedaortiz@interlink.com.ar (Nicolás) (2000-02-17) |
Re: Mathematica-like Language Implementation mauer@math.uiuc.edu (Andrew Mauer-Oats) (2000-02-19) |
From: | Andrew Mauer-Oats <mauer@math.uiuc.edu> |
Newsgroups: | comp.compilers |
Date: | 19 Feb 2000 00:29:21 -0500 |
Organization: | University of Illinois at Urbana-Champaign |
References: | 00-02-088 |
Keywords: | design |
"Nicolás" <lojedaortiz@interlink.com.ar> writes:
> I am in the course of developing a general purpose system programming
> language for Linux. My original idea of this language was for it to be
> a system programming version of Mathematica's programming language.
> [...]
> Now, I am looking for any kind of advice or help on how to implement
> this kind of language from someone who has used Mathematica.
This may sound discouraging, but it is really intended to indicate the
depth of the problem I think that you are interested in. (I assume
that you are not really trying to invent the next "C"... apologies if
you are.)
Wolfram knew a lot when he wrote Mathematica. Your project is really
going to be to learn something of why what he did is "right" (IMO,
anyway). I have a hardcopy of a transcript of a talk he gave on the
design of Mathematica, and it is very informative. You can probably
find it on the web if you look.
What I believe you will learn is that Mathematica is very closely
based on lisp and CLOS. Learn about those. (Random (and probably not
important) question to consider: "Where are common lisp macros in
Mathematica"?)
This leads to learning about Olin Shivers' scsh, the scheme
shell. (Which is actually probably not as deep of a pool as all of
Common Lisp, and probably a better starting place.)
To end on an optimistic note -- either CL or scsh could be a good
platform to start from if you were interested in doing even more.
/Andrew Mauer-Oats/
Return to the
comp.compilers page.
Search the
comp.compilers archives again.