Path: utzoo!utgpu!water!watmath!clyde!rutgers!mcnc!decvax!decwrl!sun!amdcad!ames!pioneer!eugene
From: eugene@pioneer.arpa (Eugene Miya N.)
Newsgroups: comp.lang.misc
Subject: Re: Perfect language features (Discussions of languages)
Message-ID: <3928@ames.arpa>
Date: 14 Jan 88 06:23:41 GMT
References: <3352@ihlpf.ATT.COM>
Sender: usenet@ames.arpa
Reply-To: eugene@pioneer.UUCP (Eugene Miya N.)
Organization: NASA Ames Research Center, Moffett Field, Calif.
Lines: 36

In article <3352@ihlpf.ATT.COM> nevin1@ihlpf.UUCP (00704A-Liber,N.) writes:
>If you were designing the 'Perfect Programming Language',

Oooohhh nnoo! (Mr. Bill) sounds like you would end up developing Ada(tm).
Again.......

>what features would you put in (and what features would you leave out)?

I like iterators in CLU and icon generators.  It really depends on what
the language is used.  I think we need to develop a perspective of
"little languages" as Bentley calls them, rather than make a single
perfect language.

I tend to like overloading, and dislike certain aspects of function,
prefix, and parameter passing models of languages.  I also recognize
dangers and would like certain expansion tools, if I had such a language.

There have to be features which can both have implicit as well as
explicit parallelism (not based on a port or shared memory model).

Keep side-effects down. (Sort of like saying minimize GOTOs, a
Motherhood statement).

>Do these features limit the ease of use or applicability of
>the languages they come from?

I would not base a language on Pascal block structure.
All languages restrict who you use machines.  I think you have to
iterate the development of any language.

From the Rock of Ages Home for Retired Hackers:

--eugene miya, NASA Ames Research Center, eugene@ames-aurora.ARPA
  "You trust the `reply' command with all those different mailers out there?"
  "Send mail, avoid follow-ups.  If enough, I'll summarize."
  {uunet,hplabs,hao,ihnp4,decwrl,allegra,tektronix}!ames!aurora!eugene