From: larry@JPL-VLSI.ARPA
Subject: "Brain-dead programming"
Date: Wed, 16-Apr-86 00:24:10 EST [thread overview]
Date: Wed Apr 16 00:24:10 1986
Message-ID: <8604161007.AA01445@ucbvax.berkeley.edu> (raw)
Getting rid of the worst programmers sounds like a good idea. However, you
have to remember that most people have a mix of ability, interests, and
experience that varies widely. The person who is a bad choice for new program
development may be perfect for debugging and rewriting old code.
In my experience the brightest programmers are often the worst to use in a team situation. They may be able to produce a working program ten times as fast as
anyone else, but when it needs changing no-one but they can do it--and they
won't because they're working on this desperately urgent job. And often they
have put in so much clever code that even they can't figure it out a year later.
I'm all for well-though out standards. I think Ada is one such; every part of
it has been tested in production languages except the tasking model, and that
model is general enough to cover most applications. Nor do you have to use it;
if you want semaphores or monitors you can roll your own, or get someone else's
package which you can tailor if you need to. Further, the mix of features is
pretty good in Ada--I don't get a "kitchen-sink" feel when using it.
As for "tools for lazy minds," I feel the more the better as long as they don't get in each others way. The better tools multiply the effectiveness of the
good programmer. But there is also a danger, of course; a power-tool can do
that much more damage when used by a poor worker.
Larry @ jpl-vlsi
reply other threads:[~1986-04-16 5:24 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox