As Christian Brunner said:
>
>
> On Tue, 12 Jan 1999, Gilbert Ramirez Jr. wrote:
>
> > I haven't looked very closely at your code yet; is it pretty easy to stick
> > in a different data-link type so that (eventually) token-ring packets can
> > be generated as well as ethernet? (like the switch statement in
> > dissect_packet() )?
>
> It isn't to complicated. Every header is implemented as a widget,
> with some common access functions. Just look at gen-rfc1042-w.c.
>
> Which RFC is the one for token-ring encapsulation ? (I thought that you
> can use RFC1042 packets on a token-ring network.)
True, RFC1042 applies to token-ring. But the 802.5 MAC header contains more
fields than the ethernet header. The paragraph "For IEEE 802.5" in RFC1042
explains it:
Access control: 1 octet
Frame control: 1 octet
Destination addr: 6 octets
Source addr: 6 octets
Routing Information Field: 0 - 18 octets (yes, 802.5 has a variable-length
header, which makes writing filter programs more interesting).
(and TR has no length field in the header)
Perhaps it would be useful to disassociate 802.3 and 802.2 in the dialogue
box for RFC1042 encapsulation.
--gilbert
--
Gilbert Ramirez Voice: +1 210 358 4032
Technical Services Fax: +1 210 358 1122
University Health System San Antonio, Texas, USA
