axiom-developer
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Axiom-developer] Re: bootstrap Meta/Boot


From: Ralf Hemmecke
Subject: [Axiom-developer] Re: bootstrap Meta/Boot
Date: Sun, 12 Aug 2007 01:26:19 +0200
User-agent: Thunderbird 2.0.0.6 (X11/20070728)

On 08/10/2007 06:13 PM, address@hidden wrote:
BTW, why do you think it is so essential that the underlying language of
Axiom must be LISP?

I don't. In fact, we've had discussions about using Aldor all the way
down.  I don't have the time to rewrite everything in Aldor.

But you have time to replace BOOT by LISP.

But it already is in Lisp. Boot is just syntactic sugar for Lisp which limits
what you can write and obscures the syntax.

So you don't need to use BOOT yourself. Write new functionality in LISP
or in whatever you like. Document BOOT.

Writing Boot to generate Lisp is like writing Fortran to generate Aldor.

Hmm, I would rather believe it is like "writing Aldor to generate Fortran".

Or do you want to say that BOOT is in some sense "lower level" than LISP? Or are you speaking of expressibility? If BOOT allows include or call LISP routines then I would claim, the expressibility issue is a non-issue.

If you are a Fortran programmer you won't see why that's a problem.

If a Fortran programmer uses Fortran to generate Aldor programs, I
believe that the output only uses a subset of Aldor and that
mathematical ideas that can easily be expressed in Aldor are invisible
in the fortran code as well as missing from the generated Aldor code.

If you are an Aldor programmer you can't imagine why that isn't a problem.

I am an Aldor programmer and I don't see a problem with Fortran
programmers generating Aldor. That's not the point. I would simply guess
that it is much harder to program in fortran something as complex as
Aldor's LibAlgebra. I would even bet that the code would be much longer.
And not only the code. There would be more need to explain the tricks in
human readable form.

But you can't make Fortran programmers into Aldor programmers.
They have to want to learn.

Agreed. But I also fail to convince LISP programmers to "want to learn"
Aldor.

If I now start a branch and convert Axiom to Haskell,
or Assembler would you object?

Nope. It's your time. If you want to rewrite 23 years worth of work
into another language, go for it.

Thank you.

Would you be interested?

Nope.

Can you believe that there are people who are not interested in a
LISP-only Axiom?

The "human readable" language is English.

I haven't count, but there is quite a whole crowd of people who rather
speak Chinese than English. All they can do with an (english) literate
program is to read the source code. Let's hope they are fluent in LISP.

I'm focused on getting the interpreter, compiler, and other internal
machinery into a clean, uniform, literate form in one already existing
language. My focus is on organizing and explaining what is already there.

Yes, go for it. But don't oppose to people who don't want to read
lengthy LISP code with or without documentation.

Ralf






reply via email to

[Prev in Thread] Current Thread [Next in Thread]