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

Re: C as Universal Standard

    Date: Thu, 17 Jan 91 16:32 EDT
    From: GODDEN%RCSMPB%gmr.com@relay.cs.net

    Bob Kanefsky <Kanef@charon.arc.nasa.GOV> says:
     > Americans thinking that English is the only language anyone should
     > have to know is almost as bad as UNIX programmers thinking that
     > C should be a universal standard.

    I agree.  As everyone knows, it should be Common Lisp.

Dr. industrial productivity, the former DARPA chief Craig Fields, made
statements in his talk at AAAI-90 indicating that he had no idea about the
productivity and flexibility issues that distinguish Lisp from C.  He called
Lisp a ``failed language'' (for academics) in response to a query by Patrick
Winston pointing out that Lisp was a high productivity programming language.
Those who understand these issues should convey these understandings to
research funders and managers who have been buying into C and unix dogma.  

The toughest argument, which naive programmers, managers, poverty-stricken
academics latch onto, is price performance ratios, cheap MIPS and expensive
LISPs, Lisp workstations, Lisp programs.  Of course, if one is writing trivial
programs (the definition of programming in C), then why bother with Lisp?

A success of C and UNIX is the ability to share many readily available, and
often free, programs.  Even if Lisp is a vastly better vehicle for sharing
(the promise of common lisp), the Lisp community (vendors) has failed to
develop effective infrastructure for code sharing.  Developing this
infrastructure would make Lisp much more popular.  Of course, this would be
facilitated by a common lisp standards for system definition that:

	* provides a system construction tool

	* forces logical pathname usage

	* provides an interface to advertise systems for automatic
	distribution and updating

	* enforces reasonable package use

One such defsystem is in the works at MIT.  You could get a copy from