On Tue, Dec 18, 2001 at 02:17:38PM +0100, Bernd Becker wrote:
> the fix to ethereal_gen.py from 1.16 to 1.17 broke the generated code
> a little bit. First a small syntax error sneaked in,
> col_clear(pinfo->cinfo, COL_INFO)) .
> ^
> But what this change actually does is delete the INFO column, because
> nothing is inserted in the column in the subdissectors. Leaving it
> untouched just keeps the info inserted by the GIOP dissector, which is
> OK. So I am backing it out whith a comment that something useful should
> be generated for the info column. I will try to come up with something.
>
> The changes in packet-coseventcomm.c and packet-cosnaming.c
> just reflect your change to ethereal_gen.py. After Frank Singleton
> finishes his current work with ethereal_gen.py it might make
> sense for him to regenerate the two dissectors from the IDL.
>
> Additionally I have a small fix to packet-giop.c that resets the
> COL_PROTOCOL to GIOP if none of the heuristic subdissectors
> succeeded. Right now each subdissector is setting the COL_PROTOCOL.
> If none is successful, the protocol name of the last one called
> remains in the protocol column, which is wrong and confusing.
Checked in.