Hi Anders
On 9/28/06, Anders Broman (AL/EAB) <anders.broman@xxxxxxxxxxxx> wrote:
I think it sounds reasonable to have the dissector turned off as default
and definetly turned of if there is no
"ASN.1 type table file" as the table is "stating" how the protocol
should be dissected right?
I agree that the dissector should be turned off as default, especially
given its slightly unstabled nature. However, it would still be nice
to have it show up in the "Decode as ..." list. Even if you don't have
an ASN.1 type table file, you can still use it to dissect simple
ASN.1. It may not be as comfortable (because you don't get the nice
names), but it works good enough for a lot of uses. It certainly is a
huge improvement over the raw hex data :-)
Furthermore the default port should probably be zero
Yes, that would be a good idea. Can you register a protocol for port
0, and would that have the effect we are looking for?
I did manage to get my debugger working now (you have to do make
install, or it does not find the plugin...). It seems like the
function proto_register_asn1() is never called? Anyway, the relevant
condition is right at the end:
/* If we now have a PDU tree, register for the port or ports we have */
if (PDUtree) {
So the current behaviour is intentional, but it is also easy to change
if we agree on a better behaviour.
If you frequently use the asn1 plugin to dissect a protocol you should
consider using asn2wrs and create a "real" dissector for the protocol in
question.
I agree, but the protocol in question changes so often and is used on
so many platforms that this turned out not to be practical.
Thomas
