data corruption with current (maybe sis chipset related?)

[Don]

> it's not really appropriate to -current any more.  Bosko's is
> preferred, because it doesn't disclose the nature of the bug,
How, exactly, is this preferred? :)

-Don