Path: utzoo!mnetor!uunet!husc6!ukma!gatech!purdue!i.cc.purdue.edu!j.cc.purdue.edu!pur-ee!uiucdcs!uiucdcsb!robison
From: robison@uiucdcsb.cs.uiuc.edu
Newsgroups: comp.lang.misc
Subject: Re: First Languages (yet again)
Message-ID: <170500013@uiucdcsb>
Date: 22 Feb 88 20:01:00 GMT
References: <4022@ames.arpa>
Lines: 24
Nf-ID: #R:ames.arpa:4022:uiucdcsb:170500013:000:1027
Nf-From: uiucdcsb.cs.uiuc.edu!robison    Feb 22 14:01:00 1988


> It would be nice if a physicist (for example) could stick primarily to 
> physics and not have to learn a second discipline (computer science). 

This is becoming increasing unlikely for physicists.  The problem
is not user interfaces or computer grammar, but good algorithms.
For example, most programmers know that linked lists are often
much quicker to manipulate then arrays.  How many physicists know
what a linked list is?  They probably don't need to know the gory
details of pointers, but should know about asymptotic complexity
analysis.  (I don't intend to sound condescending to physicists.  Some
knowledge of physics is useful in CS, for example simulated annealing
and ray tracing.)

Some friends and I conjecture that a lot of supercomputer time 
is spent running poor algorithms.  Anyone have any empirical
evidence or anecdotes?

Arch D. Robison
University of Illinois at Urbana-Champaign
	
CSNET: robison@UIUC.CSNET
UUCP: {ihnp4,pur-ee,convex}!uiucdcs!robison
ARPA: robison@B.CS.UIUC.EDU (robison@UIUC.ARPA)