To the best of my very small knowledge, what happens is this:
patches are installed to NVRAM, then the open firmware environment 
variable boot-command is changed to X bootr. bootr is the command in 
NVRAMRC that does things based upon the boot-device variable. You'll 
notice that it does a pattern match on ide0, ide1, enet and scsi and the 
keyboard, installing various patches to things as they are used. This is 
the reason that you have to reset-all when you change your boot-device 
settings. So basically boot is an alias to bootr which is a forth 
subroutine in nvram-rc that calls the "real" boot command after hopefully 
doing the right stuff to make the "real" boot be able to do its thing. 


On Thu, 12 Dec 2002, Markus W Kilbinger wrote:

> >>>>> "Henry" == Henry B Hotz <hotz@jpl.nasa.gov> writes:
> 
>     Henry> At 1:26 AM +0100 12/12/02, Markus W Kilbinger wrote:
>     >> setenv boot-device ide/@0
>     >> setenv boot-file netbsd
>     >> reset-all
>     >> 0 boot r
> 
>     Henry> What is the significance of "0 bootr" vice "boot"?
> 
> In System Disk's installed OF patches this kind of function (I don't
> know anything about forth!) seems to be defined and '0 bootr' is set
> as 'boot-command'. So, it seems to be a kind of 'boot' patch for OF
> 2.4 on beige g3v2....
> 
> Markus.
>