List and telco guys,
A pretty large number of people have been testing the new h245 dissector on
a rather large
number of captures with very positive response.
(==it dissects all tested packets properly)
Many thanks to Martin Regner and Danilo Kempf for support, bug reports and
problem capture files.
Therefore I have checked in changes to make ehtereal build the new h245
dissector by default.
As before, since h225 is lacking, ehtereal can not detect what is h245
automatically and one has to do
DecodeAs... to specify what is h245.
Only three instances remain where a construct is not understood in the
dissector. This will change shortly
and these NOT_DECODED_YET macros will be replaced with proper code.
As a handwritten dissector for a semi-complex protocol it works already
better than could be expected
and that is why it is now enabled by default.
With it checked in properly it is my hope that there will be feedback on how
to prettify the output
in the tree pane and the COL_INFO field of the packet list pane.
Who wants to take over maintenance of this one now? Or when it is more
complete? Andreas?
As for those expressing interest in implementing h225,
the patterns used to translate the descriptions of various h245v9 constructs
into calls to the per api now in ethereal
is pretty trivial/straightforward and a small protocol like h225 should not
be too hard to implementn now since it is
1, pretty small compared to h245
2, most of the nessecary per helpers are already implemented and examples of
the transform already exists in
packet-h245.c
have fun and report whats broken
ronnie s
