Subject: Re: 4000/60 nearly there!
To: Michael Kukat <michael@camaronet.de>
From: Jamie Lovick <jalovick@inettech.net.au>
List: port-vax
Date: 03/15/1999 17:22:58
On Mon, 15 Mar 1999, Michael Kukat wrote:
> > > 10.x.x.x and 127.x.x.x -> Class-A network, Netmask 255.0.0.0
> > > 172.16.x.x - 172.31.x.x -> Class-B network, Netmask 255.255.0.0
> > > 192.168.0.x - 192.168.255.x -> Class-C network, Netmask 255.255.255.0
> > 
> > But, this assumes the network is classful (which is an incorrect assumption
> > most of the time these days.. ). I guess if it has to make as assumption,
> > this is the way to go..
> 
> That's the problem why RARP doesn't do the job right today. We should use bootp
> (support is in libsa, maybe i find an easy way to include bootp support for
> root FS.)
> 
> > In the end tho, it seems to have worked ok despite this now that Im using
> > the 486 as a bootparam server
> > 
> > > 2) Patch the kernel, that it does bootp when mounting the root fs (maybe
> > > i try it somewhen, because my whole network works with bootp here,
> > > including linux and windoze boxes).
> > 
> > This would be the ultimate way to go.  Would this then mean we can
> > dispense with bootparamd altogether?
> 
> I just want to turn off my bootparamd and stop starting "rarp -f" after every
> boot (about 1 every month :-)
> 
> Nobody really needs bootparamd for other things, some old systems use it for
> booting, but why shold NetBSD stay an old system in this case. Bootp is much
> better and used very much today. (Windoze in DHCP mode uses bootp also).

I use bootparamd, as I have a number of Sun 3's which I need it for.

But I also class bootp as usefull .. :) (for the reason you mentioned
above).

Seeya

Jamie


--
Jamie Lovick  | Phone : +618 8242 3811 | Email: jalovick@inettech.net.au |
IT Consultant | Fax   : +618 8242 3822 | Anet : jalovick@amiganet.org    |
InetTech      | Mobile: +614 1479 1681 | Doof : jalovick@doof.org        |