Newsgroups: comp.object,comp.lang.smalltalk,comp.lang.ada
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!news.mathworks.com!news.kei.com!world!bobduff
From: bobduff@world.std.com (Robert A Duff)
Subject: Re: Has C++ had its day?
Message-ID: <DAMMwI.1ML@world.std.com>
Organization: The World Public Access UNIX, Brookline, MA
References: <3q7tb1$1ct@holly.csv.warwick.ac.uk> <DAJ73H.35M@world.std.com> <3sc5g3$hvl@gensym1.gensym.com> <dewar.803869433@gnat>
Date: Fri, 23 Jun 1995 13:02:42 GMT
Lines: 15
Xref: glinda.oz.cs.cmu.edu comp.object:33445 comp.lang.smalltalk:24937 comp.lang.ada:31643

In article <dewar.803869433@gnat>, Robert Dewar <dewar@cs.nyu.edu> wrote:
>Actually these days, machines are getting very fast, and compilation
>rates are getting very comfortable.

Yes, machines are getting faster, but they're still not fast enough.  To
me, fast enough means truly interactive response: I can make a change to
a large program (like adding a subprogram to a heavily-used package),
and then have the system recompile, relink, (and maybe run all the
regression tests) so fast that I'm not tempted to get up and get another
cup of coffee (i.e. about a second or two).  Preferably .2 second, so I
don't even notice a delay.  Today's computers are still not *that* fast,
so there's still value in incremental compilation / incremental linking
/ smart recompilation / etc -- tools that make the job happen faster.

- Bob
