According to Todd Vierling <tv@NetBSD.ORG> :
> 
> I have not yet changed mknod(8) or ls(1) (except to try out a three part
> split on ls).  I'm just trying to address the cosmetic issues, which at the
> base, aren't even problems.  Restructuring dev_t as you proposed just adds
> marginal readability at the cost of kernel cruft and additional
> compatibility issues. 
> 
I think it's time the emphasis here got changed a touch (mudslinging 
being evidently ready to pop up): which of the two proposals - Todd's 
extension vs minor split - scales better to 64-bits?

I think we should expand to 64-bits now, but that may be a little 
premature given the most common architecture.  As a compromise, the 
mechanism that makes it easiest to upgrade to 64-bits should be 
favoured.

-- 
Lucio de Re (lucio@proxima.alt.za)
Disclaimer: I'm working at getting my opinions to agree with me.