> I am unhappy about this. It only works if everybody does
> "inner++" instead of "inner=TRUE". The latter is the more
> sensible thing to do for a boolean value.
I am definitely an ethereal newbie, but my impression was that
a dissector never specifically assigns any booleans like that.
Instead, the moment when a protocol dissector is called,
it's "protoname" filter automatically triggers.
Same thing seems to happen when an FT_NONE is added -
the filter (named as described) in the associated hf registration info -
is triggered.
So, it doesn't look to me that the change has to be scattered anywhere
across the dissectors.
>
> If you want to do this, the way you describe (define an
> "inner.count" hidden integer field and increment it every
> time) is much better. Either that, or find out some way to
> declare inner as FT_UINT?
This is something I am currently trying to do, but the problem here is
that mine is not the only proto with many similar primitives,
and I had to forsee the filtering desire of my dissector future user
to select specifically an outer frame with multiple inner frames,
and whatever similar protocol dissector wasn't designed like that,
there is no option for the user to do such filtering.
Please correct me if I am missing something obvious,
Vassilii
