Re: Hi, I'm creating a new language, anybody wants to help?

Nils M Holm <Sunrise!nmh@dialup.nacamar.de>
5 Nov 1999 01:35:21 -0500

          From comp.compilers

Related articles
Hi, I'm creating a new language, anybody wants to help? nop39545@mail.telepac.pt (Joćo Paredes) (1999-11-03)
Re: Hi, I'm creating a new language, anybody wants to help? adonovan@imerge.co.uk (Alan Donovan) (1999-11-05)
Re: Hi, I'm creating a new language, anybody wants to help? Sunrise!nmh@dialup.nacamar.de (Nils M Holm) (1999-11-05)
Re: Hi, I'm creating a new language, anybody wants to help? jandk@easynet.co.uk (Jonathan Barker) (1999-11-05)
Re: Hi, I'm creating a new language, anybody wants to help? jshof@my-deja.com (1999-11-16)
Re: Hi, I'm creating a new language, anybody wants to help? nop39545@mail.telepac.pt (Joćo Paredes) (1999-12-04)
Language vs. run-time Re: [...] creating a new language [...] mihai@cs.wisc.edu (Mihai Christodorescu) (1999-12-07)
| List of all articles for this month |

From: Nils M Holm <Sunrise!nmh@dialup.nacamar.de>
Newsgroups: comp.compilers
Date: 5 Nov 1999 01:35:21 -0500
Organization: Compilers Central
Keywords: design

> [Not to be unduly discouraging, but the first question I'd ask is "Why
> does the world need yet another programming language?" -John]


Because ...


- most existing languages simply copy the features of some
    other language(s) and (sometimes) decorate it with a
    different (and sometimes new) syntax.


- only very few language designs include really new ideas
    and even fewer languages are *based* upon new ideas.


- creating compilers improves one's compiler-writing
    practice. Compilers (even simple ones) are complex
    programs and practice (frequently) leads to higher
    quality programs.


- if people would not keep inventing languages, we would
    still consider assembly language a problem-oriented HLL.


- there always exists a better solution (language) for a
    specific class of problems.


And finally, the world's need for a new language is only one
factor. IMO, there are so many uninteresting languages, *because* they
are implemented only when they are needed. I think, it is more
important to *play* with problems and develop own ideas.


Designing a language and implementing a compiler are fascinating tasks
and I would encourage everybody who is interested in this area to make
his/her own experiments, even if it does not immediately lead to a
functioning compiler.


Bye,
nmh.


(Sorry, could not resist.)
--
Nils M Holm <nmh@dialup.nacamar.de> [Please use Reply-To:]
http://www.homepages.de/home/nmh/ -- The home of the T3X compiler
[Oh, I agree that writing compilers can be very educational, and there
are interesting experiments yet to be done in language design. But I
don't see the point in yet another mutant Fortran/Pascal/C. If you
implement an existing language, you have the great advantage of having
lots of existing code in that language to use for testing and
benchmarks. -John]





Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.