Messages in this thread | | | Date | Wed, 5 May 1999 08:54:00 +0200 | From | Kurt Garloff <> | Subject | Re: Ken Thompson interview in IEEE Computer magazine (fwd) |
| |
On Tue, May 04, 1999 at 04:50:10PM -0700, Ted Rolle wrote: > FWIW -- I've heard much the same thing from one of the developers of > FreeBSD. Instead of killing the messenger, why don't we just accept the > criticism as valid (and true) and fix it?
Because the critisism is not specific. It he had said: SCSI implementation in Linux sucks (it does), OK we would accept it and try to fix it.
Just saying it has a way to go may be true but doesn't help anyone.
I appreciate accepting critisism, but: Just saying, hey let's work on it does not help anybody. Working on WHAT? Linux has 11 MB gzipped source code!
Saying it's unreliable (and even more unreliable than W*n) is something very strange, which has to be explained, before i can accept it. Compared to a HA Unix cluster, Linux might be unreliable, yes, but ...
I'd suggest sb to contact him and asking him * where he sees the weak points in Linux * what makes him think it's unreliable
THEN we could start to work on improving Linux wrt to his critisism.
> This attitude of "he-doesn't-know-what-he's-talking-about" is appalling > and will never produce a quality Linux. Especially considering its > proponents and the target of the criticism. > > Let's get a clue, smell the coffee, and go to work!
He certainly know what he's talking about. But the statement on Linux in this interview was not his main point but rather a side note and not made with too much care. (I don't say without reasons or without knowledge, please note that.)
Let's ask him for specific problems ...
Regards, -- Kurt Garloff <garloff@suse.de> SuSE GmbH, Nürnberg, FRG Linux kernel development; SCSI driver: DC390 (tmscsim/AM53C974) [unhandled content-type:application/pgp-signature] | |