I think it should make 3.0, it seems quite critical and counter intuiative. I would however like to know if others on this list agree with the idea before making a jira. (although logic escapes me as to why we wouldn't want it ^_^ i've made a lot of nice errors recently - see the 270 patches's thread context classloader bugs for example - and I'd prefer a concensus on it)
-----Original Message----- From: Eitan Plotnik [mailto:[EMAIL PROTECTED] Sent: Monday, April 28, 2008 4:30 PM To: [email protected] Subject: RE: Issue in EndpointReference.equals() method Chris, Do you think this one should be fixed also in official release? should I file a Jira? Would appreciate if you could please upload a patch once you have the fix. Thanks, Eitan. Chris.Twiner wrote: > > I agree with the reason: > > // > // compare wsa:Address - EXCLUDING the host. this allows us > // to compare EPRs that have equivalent IP addresses and > // machine names/localhost > // > > given that localhost could also be the machine name and the ip > address, or be multi homed. > > However it seems to be perfectly valid to run the same client (with > presumably the same url source path) from two seperate machines or > indeed multiple clients from a single machine. > > Its exceptionally difficult to make hostname lookup work in general > but I think its a valid enough requirement for the client to provide a > decent hostname: > > InetAddress addr = InetAddress.getLocalHost(); > > shouldn't be that hard to use. > > And IF need be have a simple rule for "localhost" vs the actual local > host name and default interface. For multi-homed servers/clients its > normal to require them to choose an IP address / hostname. > > Its a simple fix but there may be something I've overlooked. > > PS (thanks for finding this Eitan, I'll certainly be fixing it on my > local version) > > -----Original Message----- > From: Eitan Plotnik [mailto:[EMAIL PROTECTED] > Sent: Monday, April 28, 2008 10:49 AM > To: [email protected] > Subject: Issue in EndpointReference.equals() method > > > I have a test client registering for notifications. > When I run the client from 2 different machines at the same time, > using the same server, some of the notifications are sent to the wrong client. > > I noticed that EndpointReference.equals() method explicitly doesn't > compare the IP address part. > > Does the test client must generate a unique (machine based? > timestamp?) subscriber uri part after the IP address? > > Thanks, > Eitan. > -- > View this message in context: > http://www.nabble.com/Issue-in-EndpointReference.equals%28%29-method-t > p1 > 6930964p16930964.html > Sent from the Muse - Dev mailing list archive at Nabble.com. > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > > -- View this message in context: http://www.nabble.com/Issue-in-EndpointReference.equals%28%29-method-tp1 6930964p16940189.html Sent from the Muse - Dev mailing list archive at Nabble.com. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
