I agree that individually selecting _every_ broken TCP/IP stack workaround
can only lead to madness. Even breaking it down by vendor would still
lead to some kind of mental injury.
It would seem that if neither 0 nor all possible configurations are
acceptable, what about two possible configurations: compatible with
all known broken TCP/IP stacks, and not.
Think about the special case of a Linux-only network. A node in a
Beowulf cluster could be reasonably expected to never talk to anything
that wasn't running exactly the same Linux kernel revision, and they
might care about extra packet routing latencies or protocol overhead,
even if that overhead was local to the computing node.
Of course if the TCP/IP stack workarounds don't produce measurable
performance changes then the whole idea does not merit discussion, and
such discussion should be dropped until such performance problems appear.
-- Zygo Blaxell, Linux Engineer, Corel Corporation, zygob@corel.ca (work), zblaxell@furryterror.org (play). It's my opinion, I tell you! Mine! All MINE! Size of 'diff -Nurw [...] winehq corel' as of Mon Apr 19 14:14:02 EDT 1999 Lines/files: In 16568 / 137, Out 8736 / 111, Both 20566 / 208- 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/