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.
E2fsck has a built in regression test suite, and I don't release an
e2fsprogs release without first running it through that test suite.
This is necessary, since a screwup on my part may mean that a large
number of users may lose their data, and I'm not willing to risk that.
Even with my paranoia about doing release, I've occasionally released
e2fsprogs release that weren't 100% stable. Most of the problems have
been in the mke2fs program, mainly because I don't test mke2fs carefully
enough in my regression test suite. Fortunately bugs in mke2fs aren't
as critical as bugs in e2fsck, but it's still (to my standards) not
acceptable. I hope to do better in the future.
I doubt you'd be willing to find *anyone* who does software development
for a living who would be willing to advocate shipping code without
testing it. (In fact, I was talking to an acquantance who works in
quality assurance, and I got her to start spasmodically clenching her
hands and reaching for my throat when I cracked a joke about the
software engineer's motto --- ``It builds, SHIP IT!!'')
In addition, the > 2GB changes requires kernel mods, and even Linus has
gotten in trouble when he has released kernels without testing them
first. Perhaps you can really write bug free code "to start with",
including kernel patches. If that's really true, then you're a better
programmer than I.
But I doubt it. :-)
- Ted