Comment # 12
on bug 11322
from Hadriel Kaplan
(In reply to Pascal Quantin from comment #10)
> That still sounds as a hack to me, but I do not have any strong opinion.
> Given the comments found in packet.h for dissector_try_uint_new(), returning
> 0 still seems incorrect.
I think it was due to historical reasons - the Lua API only did the old
dissector model originally, and I guess most people didn't even return any
value in their script dissector.
But anyway, yeah it does feel like a hack. The more I think about this the
less I think there will be other scripts like this anyway - I'll try and find
more TCP-based dissector scripts out there to see what they do.
At the very least, what I'll do is have the dissect_lua() function in
init_wslua.c perform the same type of assertion check, but generate some
detailed explanation of why it's not legal. (and I need to update the wiki page
to describe it in more detail)
You are receiving this mail because:
- You are watching all bug changes.