From: "David Reid" <[EMAIL PROTECTED]> Sent: Tuesday, January 08, 2002 11:51 AM
> But the patch didn't actually do anything to prevent the problem we saw as > it was in NO way related to a fork(). The issue is that if we call a > pool_clear we'll run the cleanup's anyway, as we should. That's right. This is why my last patch dropped ALL cleanups of the fd, for fd 0, 1, and 2. I agree my first patch was of dubious value :) Bill
