Not so pessimistic - you just write the code. There are so
many Alpha's these days - people will complain when something
is wrong. Testing is not the right way to get working programs -
you just write them correctly to start with.
You may be able to write code that is bug free from the very beginning,
and you may even be willing to release code that hasn't been tested to
users --- but you'll forgive me if I hold myself to higher standards.
Well, of course my remark was semi-joking. I should have added a :-).
But on the other hand it was semi-serious.
Some situations are common, and easily tested; other situations have
a very low probability of occurrence. (But there are quite a lot of such
low-probability events, and the total likelihood is non-negligible.)
If one programs with the idea that everything must be right from the start
then testing is superfluous. One might, or might not, actually do some tests.
On the other hand, if testing is seen as a necessity, required to catch
bugs, then one only obtains a program that works right in the common cases.
E2fsck has a built in regression test suite, ...
Yes - and this just demonstrates this point.
I stumbled over a bug in 1.09, and in the 1.10 release notes you write
Fixed rare bug in mke2fs where if the user had a very unlucky number
of blocks in a filesystem (probability less than .002) the resulting
filesystem would be corrupt in the last block group.
almost as if it were unfair to encounter a bug with so small a probability.
But of course, with a million users, hundreds will encounter it.
So - I do not believe in testing (of programs that do not depend directly
on hardware behaviour).
On the other hand, I do have an alpha and some big disks, and am quite willing
to mail you the results of some regression tests in cases where you are unable
to do them yourself.
All the best - Andries