Comment # 3
on bug 12118
from Pavel Sindelka
I'm afraid I've found a bug in AVPL matching - while strict matches do work as
described, Loose and Every match report success also when no AVPs of the same
name exist in both the PDU's AVPL and the match instruction's configuration
AVPL. So for both Loose and Every, (a=aaa, b=bbb) "matches" (c=ccc, d=ddd)
except that the result is an empty AVPL.
As I am definitely unable to fix this behaviour myself, we may end up with an
up-to-date Wiki describing features that Luis had in mind but hasn't
implemented properly, with a remark "currently works this much different from
what is described due to a bug".
So my question is how realistic it is to expect that someone will pick up the
bug I would eventually file on this, along with
https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=12161 and the not-yet-filed
bug on MATE's inability to extract any mgcp field (as I've seen at the Q&A site
and verified to be true), and make MATE work the way Luis had in mind. The
answer determines the amount of effort I'd spend on analysis and documenting of
the Loose and (likely) Every AVPL match behaviour.
You are receiving this mail because:
- You are watching all bug changes.