> As a point of historical reference: I confirmed a working 38400 on an SS1
> for the SunOS 4.0.3c release.

Point of hysterical reference: My little SS1/IPC-esque hybrid clone here has
had a "working" 38400 for the 4 years I've owned it ... (under SunOS 4.1.4).

The *only* time I get silo overflows is when there's heavy CPU-intensive tasks
going on, i.e. like when INN is doing expiration/overview expiration/etc.
processing twice a day.

Somebody should find Brian Katzung and draft him in to help us out ... he
really did a lot of nice work on the SunOS 4.1.x zs driver, some of it even
got folded into the SunOS 4.1.4 driver if I remember correctly ...

	- Greg