[Home] [Groups] - Message: [Prev in Group] [Next in Group]
10748: Re: [MUD-Dev] Properties of computer languages
[Full Header] [Plain Text]
From: Chris Gray <cg@ami-cg.GraySage.Edmonton.AB.CA>
Newsgroups: nu.kanga.list.mud-dev
Date: Thu, 17 Jun 1999 19:27:26 -0600
Organization: Kanga.Nu
[Caliban Tiresias Darklock:]
> Yeah, I know, don't design inefficiency in -- it will happen in the
> implementation. But I'm talking about BIG inefficiency. To give you an
> example, here's a small portion of what I'm improving on:
Hmm. Do I hear my name being called? :-)
Before reading that paragraph, I was going to suggest leaving out most
of the language and having *only* recursion. Could be truly inefficient.
Back when I was TA-ing an intro comp. sci. course, I was supposed to be
teaching recursion. I did an example that showed using recursion to do
addition and multiplication (using the addition). One student, on the
next assignment (not really related to recursion), used my example
routines in his program, and couldn't understand why his job ran out
of time on the mainframe. Oooops.
On the more serious side, and going on Caliban's example, which I've
snipped, I'd suggest as some goals for the language:
- as few special cases as possible
- no gratuitously added special syntaxes
- an absolute minimum of "magic values"
All of the %# things in Caliban's example violate all 3 of these.
_______________________________________________
MUD-Dev maillist - MUD-Dev@kanga.nu
http://www.kanga.nu/lists/listinfo/mud-dev