["Peter J. Creath" <peter-ethereal@xxxxxxxxxx>]

> You might want to check out the current CVS version of 
> "dissect_xdlc_control()" - it now takes, as an argument, a structure 
> containing hf_ values for named fields for the subfields of the 
> control field, so it no longer directly calls the routines to format 
> bitfields.

Ah, good point.

> Your changes might still be useful for dissectors that have 
> bitfields that aren't named fields, however.

For bitfields with more than named fields, I think my changes are 
very helpful.  For completeness, you can of course use 
bitfield_string to represent named fields as well as more complicated 
bitfields.

And even if you're only doing named fields, these changes might be 
preferable.  They provide a convenience routine to generate the 
entire string of bits set (rather than doing it by hand).  And 
perhaps more importantly, they provided a routine to generate the 
subtree showing the bitfield detail (which I personally find helpful).

	-P