Will try this on W$ once I get back home in couple of hrs. thanks ashish
On Fri, Feb 12, 2010 at 3:37 PM, Emmanuel Lecharny <elecha...@gmail.com> wrote: > Hi gys, > > yesturday I tried to fix the Datagram bug (DIRMINA-732). I think it works on > linux, I have no idea for windows (if someone can give it a try), and I know > it fails on Mac OSX. > > The problem is that if we register the 'any' address (0.0.0.0) on a datagram > acceptor, and try to get a session from it, like the way it's done on the > failng test : > > .. > InetSocketAddress bindAddress = new InetSocketAddress( > AvailablePortFinder.getNextAvailable()); > acceptor.setHandler(new TestHandler()); > acceptor.bind(bindAddress); > IoSession session = acceptor.newSession(new InetSocketAddress( > "127.0.0.1", AvailablePortFinder.getNextAvailable()), > bindAddress); > ... > > we get an error, as the bindAddress is considered as a Inet4Address, when > the acceptor is bound on a Inet6Address. As we store the address in a > hashMap to check if they are equals, we have an issue, because whatever you > do, a Inet4Address will always be different from a Inet6Address; I hacked > the code atrociously by using a String translation for this address, but it > does not work on Mac. > > Now, I'm dubious about the test. Why should be grab a session from the > acceptor ? > If someone has some idea about what's wrong about this test, that would be > great ! > > > -- > Regards, > Cordialement, > Emmanuel Lécharny > www.nextury.com > > > -- thanks ashish Blog: http://www.ashishpaliwal.com/blog My Photo Galleries: http://www.pbase.com/ashishpaliwal
