Re: language evaluation metrics?

Charles Fiterman <cef@geodesic.com>
24 Nov 1998 22:26:08 -0500

          From comp.compilers

Related articles
language evaluation metrics? kcooper@cs.ubc.ca (1998-11-19)
Re: language evaluation metrics? jk@icok.net (Jimmy Kerl) (1998-11-21)
Re: language evaluation metrics? kcooper@cs.ubc.ca (1998-11-21)
Re: language evaluation metrics? dwight@pentasoft.com (1998-11-24)
Re: language evaluation metrics? cef@geodesic.com (Charles Fiterman) (1998-11-24)
| List of all articles for this month |
From: Charles Fiterman <cef@geodesic.com>
Newsgroups: comp.compilers,comp.lang.misc
Date: 24 Nov 1998 22:26:08 -0500
Organization: Geodesic Systems
References: 98-11-098 98-11-110
Keywords: design

Kendra Cooper wrote:


> I'm actually working on a requirements specification language....
>
> But when people develop new languages, how do they know they have
> "everything" in the language? (programming language or other)?


First the power of a language comes from what it can't do not what it
can. The power of pure functional languages comes from the absence of
side effects. The power of Java is that you can't escape the type
system and write a virus.


Languages need to start small and be expandable. To make them fully
expandable I propose using the following string notation.


Strings are marked two ways [] . Strings become something else only
when some language construct says so.


if x < y
    print
        hi sailor
        want a date


The material under the print is simply a string. The matereal under
the if starts a string but context causes it to be cast to code. If x
>= y at compile time the string under the if is discarded at compile
time.


The sample statement is lexically equivalent to


if x < y [print [hi sailor\nwant a date\n]\n]


except for line number annotation.


Post a followup to this message

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