Does anyone have any empirical evidence that the Big Todo patch actually
improves performance?
I tried a few tests, and didn't seem to find any difference in performance
between stock qmail and one that was patched.
One test included shutting down qmail, putting 5,000 messages (to local
recipients) into the queue, and the starting qmail and timing how long it
would take till the queue was empty.
The other test (a bit more complicated) measured the amount of time it
took qmail to "process" a message (move it out of the todo directory).
Both situations seemed to take the same amount of time. This is under
Solaris 2.6.
-Eric
