Re: Linux 2.0.4 (and a small patch, d*mn)

Andrew E. Mileski (aem@nic.ott.hookup.net)
Tue, 9 Jul 1996 06:20:44 -0400 (EDT)

> > Dejan Ilic (aka "a bug's worst nightmare") whipped me into shape - I now
> > always make sure every file can be compiled correctly using every
> > available option. Others should take up this philosophy too. It doesn't
> > prevent run-time errors, but it does prevent silly typos.
> > -- Andrew E. Mileski
>
> Hey, Linux has about 90 boolean config options (guessed, not counted).
>
> If poor Linus has to combine all 2^90 config options (my calculator
> can't display that number) and test them, don't you think we would
> still be waiting for 0.99.10?
> -- Georg v.Zezschwitz

First off, it is _NOT_ Mr. Torvalds's responsibility to test anything.
It is the developer's responsibility because Mr. Torvalds
- is way too busy already
- is not necessarily familiar with your code
- probably doesn't have the required equipment

Secondly, only the config options in the file(s) you are working
on need to be tested (for compile-time errors). Run-time errors are
a completely different matter, and are impossible to exhaustively
check in most cases (in other words, do your best).

--
Andrew E. Mileski
mailto:aem@ott.hookup.net      My home page http://www.redhat.com/~aem/
Linux Plug-and-Play Project Leader. See URL http://www.redhat.com/pnp/


Red Hat Software sponsors these pages - I have no other affilitation
with Red Hat Software, and I have never used any of their products.