[2005-03-06 01:58] Phillip Rhodes said:
| Brent Verner wrote:
|
|
| > whoo-freakin-hoo!!!! I've done it! Sorry, I just had to
| >share the excitement with someone,
|
| I totally empathize. I've spent all evening trying to get RMI to
| work using a naming provider other than the stock rmiregistry that
| ships with the JDK. And after many hours of dead-ends and
| wasted time, it's finally working.
Heh. That sounds a little familiar...
| And in case anybody is wondering "why isn't rmiregistry good enough?"
| Tthe main reason I wanted something else is because you
| can only bind to rmiregistry from the machine where rmiregistry is
| running. So you can't have one central registry with remote objects
| from disparate machines registered into it. To do that, you have
| to use another naming provider. Either that, or run
| rmiregistry on every machine that hosts remote objects, and then
| use yet another mechanism (a meta-directory?) to lookup the
| various rmiregistrys to go lookup the actual stubs.
...but a lot more complicated!
My breakthrough happened when I (finally!) realized where/when
things were being executed on the server and on the client, though
I'm sure I only have just enough understanding to have my problem
solved :-)
I'm curious what you are doing. Are you exporting objects from
various machines onto a single/central registry?
| > because my computer seems
| > a bit unimpressed :-P
|
| Mine too. Even my room-mate's cat doesn't seem to care. She just sits
| there looking at me with that "Dumbass, I could have had that working
| hours ago" look.
LOL! My old monitor started looking at me that way at some point
last night, and I got all paranoid and went into the other room :-)
cheers.
Brent
_______________________________________________
Juglist mailing list
[email protected]
http://trijug.org/mailman/listinfo/juglist_trijug.org
