Comment # 45
on bug 7893
from Michael Mann
(In reply to comment #43)
> (In reply to comment #42)
> > I think heuristics are the better way to go and I have some ideas,
> > especially if packet data can be used.
>
> Heuristic vs Directly Information about payload type? Heuristic is not
> optimal and can generate bugs (try open file in Wireshark, some of them are
> incorrectly recognized (mostly PDFs, compressed archives)).
I guess I was thinking:
1. Try "heuristics" if payload type isn't already registered.
2. Try "heuristics" first, then if failed, explicitly try payload type.
3. User preference in rtp to determine 1 or 2 (if we can't decide if 1 or 2 is
definitely the way to go or just like the flexibility)
So far only BT would be involved in the heuristics, and the heuristics could be
as simple as checking for a preexisting value in a packet (use p_add_proto_data
in BT dissector before rtp dissector is called then the heuristic would just
use p_get_proto_data and return false if not present)
You are receiving this mail because:
- You are watching all bug changes.
