[Joerg Mayer <jmayer@xxxxxxxxx>]

On Wed, May 01, 2002 at 03:12:38AM -0700, Guy Harris wrote:
> Well, there's still a fair bit of noise - the biggest offenders at this
> point are:
> 
> 	the generated dissectors for some GIOP protocols - fixing that
> 	"right", may be hard, as the generator might not know whether
> 	the variable in the boilerplate it generates will be used, so it
> 	might just have to flag *variables* with _U_ (it's generated
> 	code, so that sort of ugliness doesn't bother me all that much);

I'm thinking about another solution, but I don't know how portable it is:
add the following to the compile flags:
-Wall -W ... `test -r .nowarn && cat .nowarn` ...
and put a file .nowarn with -Wno-unused as its sole content into the
directory. This makes two assumptions: 1) `...` is supported by standard
make (it works with gnu make) and 2) that the dissectors in question are
located in a directory of their own (which they curently are).

> 	the DCE RPC stuff, and NT services built atop it;

I've asked Ronnie about them and received feedback. Once I've integrated
his feedback I'm going to ask Tim about the remaining ones.

  ciao
    Jï¿œrg
--
Joerg Mayer                                          <jmayer@xxxxxxxxx>
I found out that "pro" means "instead of" (as in proconsul). Now I know
what proactive means.