Hi,
>From Pascal:
> GMR-1 dissectors, like GSM-A dissectors, can be seen either as separate
> protocols belonging to the same family, or as a single protocol with sub
> dissectors:it depends a bit on your point of view :) Personnally I see them
> as a single one (like Sylvain) and was rather happy with the '.' separator
> instead of '_' one.
Excatly the way I see it. Especially since for example "RR" is not
even a protocol ... it's mostly a collection of elements that are
going to be reused at several place in different ways. ( In CCCH, RACH
and DTAP for example ...)
> For example at the beginning in Wireshark there was a single packet-gsm_a.c
> file that became huge and was split into several files in r25915 and r25917.
Yes, excatly. And here I tried to anticipate a bit better because I
expect some of those files to become pretty big. For eg the RR
dissector is only ~ 25% complete and it's already in the top 20% of
the biggest dissector in # of lines. The BCCH dissector only has 2
segment type fully supported out of few dozens and is not that far
behind. DTAP is a bit short right now but that's only because I took a
shortcut and forwarded most messages to the GSM dissector (that's
mostly correct but not entirely and it'll require a fair bit of code
to make it 100% correct) ...
>From mmann78:
> I'll revert the GMR-1 display filter name changes and update
> checkdisplayfilter.pl to accept periods in the "base" filter name, but I
> will note their existance in another "FILE LIST" grouping in case other
> dissectors don't have a valuable reason to do so.
Thanks.
Please beware to not just revert the commit. As it turns out you fixed
some typo at the same time (where the filter name was ._ from previous
hesitation between . and _ ... I can send you a diff if you'd like).
Cheers,
Sylvain
