"Chris G. Demetriou" <cgd@cs.cmu.edu> writes:

> I'm not entirely sure I believe that.  1.2.1 reported the same errors
> for me with a Jaz on a BusLogic 542B, but -current 'did the right
> thing' including figuring out a geometry.  (the geometry it figured
> out looked fake, but the sectors per unit value was correct, and i
> could believe that it looks fake in the way that it does to make DOS
> maximally happy...)

I think -current has the flag set so the device isn't asked what geometry
it has...  I know there are entries in the quirks table for both the ZIP
and JAZ drives.

--Michael