Related articles |
---|
lcc intel backend? nick@nsis.cl.nec.co.jp (1993-10-07) |
Re: lcc intel backend? compile time? rds95@csc.albany.edu (1993-10-13) |
Re: lcc intel backend? compile time? graham@pact.srf.ac.uk (1993-10-13) |
Re: lcc intel backend? compile time? cliffc@rice.edu (1993-10-13) |
Re: lcc intel backend? compile time? rds95@csc.albany.edu (1993-10-13) |
Re: lcc intel backend? compile time? pardo@cs.washington.edu (1993-10-20) |
Re: lcc intel backend? compile time? henry@zoo.toronto.edu (1993-10-20) |
Re: lcc intel backend? compile time? tchannon@black.demon.co.uk (1993-10-21) |
Re: lcc intel backend? compile time? henry@zoo.toronto.edu (1993-10-22) |
Newsgroups: | comp.compilers |
From: | henry@zoo.toronto.edu (Henry Spencer) |
Keywords: | performance |
Organization: | U of Toronto Zoology |
References: | 93-10-041 93-10-085 |
Date: | Wed, 20 Oct 1993 22:59:16 GMT |
pardo@cs.washington.edu (David Keppel) writes:
>There is a story, perhaps apocryphal, that by substantially
>*increasing* the compile times, you can also increase productivity,
>since people take far more care with their changes and thus go through
>many fewer compile/edit/debug (:-) cycles.
This is just another version of the tired old argument that people wrote
better letters with quill pens, because writing was so much more work that
they had to think about it more carefully and made fewer mistakes than
with fountain pens. (Substitute typewriters, word processors, etc.)
What better tools *actually* do, is make individual differences more
prominent. Sloppy writers/programmers/etc. do indeed turn out sloppier
work, because they're not forced to stop and think. But the *better*
writers/programmers/etc. can spend more time getting things really right,
instead of having to settle for second best because just getting to that
point is hard enough.
--
Henry Spencer @ U of Toronto Zoology, henry@zoo.toronto.edu utzoo!henry
--
Return to the
comp.compilers page.
Search the
comp.compilers archives again.