I say all of this to disagree that we should self censor and try to
guess what Linus will say yes too and not discuss or propose things
we aren't sure he will put into the kernel. Linus is finite, as you
say. He may not have the vision to see how a perticular idea in its
infancy would fit. But he seems to be willing to let ideas percolate
untill it is clear whether they are compatable with his vision of
Linux.
It's not self-censorship. It's we as a Linux kernel development
community becoming more intelligent about Linux kernel development. It
takes a long time for a million monkeys to come up with the works of
Shakespeare, and it sounds to me like you're argueing for a model where
people on this list should be just monkeys flinging code at Linus, who
then accepts or rejects the occasional brilliant piece of code.
In fact, it doesn't work this way at all, even today. I think you'll
find that the more successful kernel developers (i.e., the ones which
have written and contributed code which by and large *does* get accepted
into the kernel) do talk amonst each other, and not just simply rely on
Linus as the final gatekeeper.
I often float ideas on the list, and when Alan or Mingo or Larry or
David comment about whether or not it's a good idea, I *listen* to them.
Often they have suggestions which I hadn't thought of, which very much
improve part of the subsystem. This is not self-censorship, it's
teamwork.
Over time, we've developed a mutually shared aesthetic of what's
considered good kernel programming style, and what's not. When Larry or
Alan says something is ugly, and I go and look at the code in question,
I generally find that I agree with them. When the subject of devfs came
up at Lunch at the last Linux Expo, more than few "famous kernel
hackers" (sct and hpa; and a few others I can't remember, probably
dmiller and Alan Cox) unanimously agreed with me that it was a really
ugly thing.
Now, maybe we're all in space, but we've collectively written or been
responsible for a fairly large part of the kernel. (As Linus himself
stated in the last week's NYT Sunday Magazine article about Linux, he's
only responsible for writing maybe 5-10% of the kernel.) Maybe, just
maybe, we know something about clean kernel coding style. So when one
or more folks who have contributed quite a bit to the kernel takes the
time to look over a patch or a suggested idea, and makes some
constructive criticisms, don't you think it might be worthwhile to
listen, and not just depend on throwing shit at Linus and hoping he
knows how to reject the shit?
- Ted
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/