On Thu, Apr 14, 2005 at 08:40:52AM +0200, Joerg Sonnenberger wrote:
> Which keeps the question why the initial correction did not happen.
Turns out this was a bug fixed in OpenBSD a while ago. Basically it
checked whether it sent a query and reverted to normal daemon mode
if it didn't. The first run of the main loop was fine, the second run
(after a timeout of 0) not.
I've only fixed a related problem of long timeouts after the settimeofday,
it should work fine now.
Joerg