Client distribution, installation, and configuration

When someone wants to use White Orb they go to our web page and click "download" (or whatever).  A self-extracting archive will be downloaded to their computer.  When this is executed a "whiteOrb" directory will be created containing the following files and directories:
  To start the client the user simply types "orb".

The orbLoader program

In order to logon to White Orb you need to know where the server is running, and you also need to have the correct version of the client.  This is what the orbLoader program is for.

When the user types "orb" the orbLoader program will start up and read a public file somewhere on the internet that contains something like this:

serverURL=rmi://whiteorb.com/orbServer

The location of this file will be stored in the orbSettings file (using a hardcoded default value).  So we have to decide some location on the internet where we can put a "serverLocation.ini" file that we know we won't have to move around.  As long as this file is available, we can move the server to wherever we like - the client will find it.

Once the orbLoader has found the serverLocation.ini file it will use the given URL to connect to the server.  From here on it will use the Server for everything.

It will now ask the server which client version is the latest, and check the local client version.  If the local client version is out of date it will tell the user something like "Your WhiteOrb client is out of date - click here to update it".  Then it will download the newest orbClient.jar file and replace the local one.

That's why orbClient.jar is seperate from orbLoader.jar.  The orbLoader.jar classes will execute the updating of orbClient.jar.  So the orbLoader.jar will NOT be auto updated, and should be perfect from start :o)

This orbLoader mechanism will save us from having to worry about which client versions we have distributed.


Henrik Kniberg

Last updated: