On Thu, 5 Sep 2002, Bill Fumerola wrote:
> On Wed, Sep 04, 2002 at 01:23:59PM -0700, Guy Harris wrote:
> > On Wed, Sep 04, 2002 at 03:39:42PM -0400, Matthew Smart wrote:
> > > The patch to 0.9.6-current is trivial: just added the source file
> > > to Makefile.am and Makefile.nmake and two new files, packet-netflow.h
> > > and packet-netflow.c.
> >
> > Checked in.
>
> that is unfortunate. mine[1] is a lot better and supports all netflow
> protocols. the one checked in is really incomplete and just adds a bunch
> of text fields. it is missing a lot of support. as noted later in this
> thread, it is missing support for juniper specific information - i have
> that information from juniper directly.
>
> i sat on submitting it until it was complete, but if incomplete dissectors
> are ok (as they evidently are), please remove this one and commit mine
> which just has a few fields implemented a little more lazily then i'd
> like.. but is otherwise pretty complete.
Ummm, Bill, how do you think Microsoft got where it is?
Incomplete dissectors are very useful, as they help people begin to
understand what is going on. CF, the NTLMSSP dissector, which is actually
wrong, but help enormously.
Regards
-----
Richard Sharpe, rsharpe@xxxxxxxxxx, rsharpe@xxxxxxxxx,
sharpe@xxxxxxxxxxxx
