Bill Studenmund wrote:

> Also, don't boot monitors require an ffs of the host's endianness? Or is
> that totally bootblocks-specific? For machines where the "chosen"
> endianness is different from the vendor os, we would make it hard to
> switch OS's.

Half and half at least for the pmax.  The ufs implementation is in
the bootblocks, and the first stage is fairly tight on space at the
moment.  At a guess, I'd say byte-swapping wouldn't fit in the existing
framework.

Don't some arch's contain enough smarts to read ffs filesystems in PROM
though?  I thought OFW could...

Simon.