On Thu, 22 Feb 1996, Sean Berry wrote:

> A friend of mine mentioned that the failure to get the file might be related
> to a byte order difference between the HP and the i386 box.

Having hacked on rbootd quite a bit, I can say with some authority that 
it's not an endian-related problem on NetBSD machines.  This was fixed 
some time before the 1.1 release.

> [...]
> [rbootd exchange]
> [...]
> 12:25:04.254501 8:0:9:1:5a:9b > 0:20:af:3b:d0:82 sap f8 ui/C len=43
>                          0000 0006 0806 0903 0000 0000 0000 0803
>                          2c48 5053 3330 3020 2020 2020 2020 2020
>                          2020 2020 2009 5359 535f 4e
> 12:25:09.056789 rarp who-is 8:0:9:1:5a:9b tell 8:0:9:1:5a:9b
> 12:25:09.092431 rarp reply 8:0:9:1:5a:9b at lunchbox.ansci.iastate.edu
> 12:25:09.217835 lunchbox.ansci.iastate.edu.1023 > 255.255.255.255.sunrpc: udp 96
> 12:25:09.262304 arp who-has lunchbox.ansci.iastate.edu tell traveler.ansci.iastate.edu
> 12:25:09.421372 arp reply lunchbox.ansci.iastate.edu is-at 8:0:9:1:5a:9b
> 12:25:09.735545 lunchbox.ansci.iastate.edu.1022 > traveler.ansci.iastate.edu.694: udp 100

I'm not sure what this is, but you could start debugging by running 
rpc.bootparamd with the -d flag.  I'm assuming that it's started up 
already; edit /etc/netstart if it's not, it should be fairly obvious.  
Also, make sure that /etc/netstart and /etc/rc agree on the name of the 
variable used to say whether to start it or not.

--scott