Newsgroups: comp.object,comp.lang.c++,comp.lang.ada,comp.lang.cobol,comp.lang.clos
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!gatech!newsfeed.pitt.edu!uunet!in1.uu.net!mole-end!mat
From: mat@mole-end.matawan.nj.us
Subject: Re: C++ not OOP? (Was: Language Efficiency
Message-ID: <1995May20.140250.19426@mole-end.matawan.nj.us>
Organization: :
References: <dewar.797512974@gnat> <3mbmd5$s06@icebox.mfltd.co.uk> <dewar.800831371@gnat>
Date: Sat, 20 May 1995 14:02:50 GMT
Lines: 39
Xref: glinda.oz.cs.cmu.edu comp.object:31235 comp.lang.c++:129342 comp.lang.ada:30164 comp.lang.cobol:3437 comp.lang.clos:2977

In article <dewar.800831371@gnat>, dewar@cs.nyu.edu (Robert Dewar) writes:
> Umm! yes, I think I am familiar with rapid prototyping (I am one of the
> principle investigators on the Griffin contract ...

Uh, Professor Dewar (it _is_ still `professor,' is it not?), some of us
would find things more convenient if you would stick with 80-character
lines ... please?

> As for small machines not being fast enough, there is no big problem in
> producing a C++ compiler that would compile at several hundred thousand
> lines a minute on a typical notebook computer with a decent amount of
> memory ...

Very often, the problem is neither the compiler nor the hardware but
the operating environment.  I've used Borland C++ in a networked
DOS-Windows-UNIX setup, and found it faster when my working files were
across the (NFS/TCP-IP) network on the UNIX `server' than when they were
on the local machine.  For that matter, I've found it faster to do many
file-oriented things on the DOS/Windows machine _from_ UNIX across the
network.

The file system on DOS and older versions of Windows is that bad.  On
the other hand, Windows-for-WorkGroups fights us when we try to run the
NFS client and server software, and its file system is no faster.

I know at least one cfront package that runs four or five times faster
on the `ufs' (was-one-the-fast-file-system) than on the original SysV
file system--even when the machine with the original SysV filesystem had
a system clock twice as fast and had twice as much main memory as well.

It _hurts_ to see the micro world make every mind-numbing mistake that
the mainframe and mini worlds made--and _never_ look that way for the
solution.  It's as though they felt that the older machines weren't
`real computers.'
-- 
 (This man's opinions are his own.)
 From mole-end				Mark Terribile
 mat@mole-end.matawan.nj.us, Somewhere in Matawan, NJ
	(Training and consulting in C, C++, UNIX, etc.)
