Re: RMI binding bug

1998-11-19 Thread Douglas T. Brown
You are correct to pick up the typos in my note. The syntax that I got to bind was in fact "rmi//... rather than "rmi:// ... As I have since read various sources, I have found the syntax for the argument of bind and rebind variably as: ("rmi://host/name", obj) or ("//host/name", obj). Further e

Re: RMI binding bug

1998-11-19 Thread Chris Kakris
"Douglas T. Brown" wrote: > > 1) The correct syntax for the first argument string for rebind is: > "rmi://host/name", whereas I had only "//host/name". > > 2) More interesting is that localhost does not work. Thus, > "rmi//127.0.0.1/HelloServer" does not work, nor does > "rmi//localhost/Hello

Re: RMI binding bug

1998-11-19 Thread Douglas T. Brown
Thank you all for your help on this. I seem to have it working now. There were two issues which I'll try to make explicitly clear for the mail list archive and future reference: 1) The correct syntax for the first argument string for rebind is: "rmi://host/name", whereas I had only "//host/nam

RE: RMI binding bug

1998-11-18 Thread dmarsh
Just a pointer: for /etc/hosts (or the equiv in NT (windir)\system32\drivers\etc\hosts) (assume your I/F IP is 10.12.14.16) --- 127.0.0.1 localhost 10.12.14.16 thehost.mydomain.org mypc --- the order is somewhat important (ip a

Re: RMI binding bug

1998-11-18 Thread Douglas T. Brown
Carl, You wrote: Is it not "rmi://xxx.xxx.xxx.xxx/name" ?? First, let me thank you for your quick response. Second, let me say, simply: It is. :-) I appreciate your help, and now I'll just quietly slip back under the nearest rock! Thanks again, -DTB

Re: RMI binding bug

1998-11-18 Thread Matt Welsh
The issue here is that the RMI registy is checking that the client's hostname matches the local hostname on bind() and rebind() calls. In addition the IP addresses are checked for a match. The logic is: InetAddress localhost = InetAddress.getLocalHost(); Class class = Class.f