First files checked in

My project at sourceforge has been approved. So I have uploaded first source code:
http://diymiddleware.cvs.sourceforge.net/diymiddleware/
There are three files there:

• Server1.java. You can run the server (does not require any arguments). It will try to bind to the port 22221 and listen for incoming client connection. Server never stops on its own, so you have to terminate it forcibly.
• Client1.java. Non-runnable API-like client for the server. In order to use it, you need to instantiate it, passing host name of the server (DNS or IP address) as the argument. It will create a thread, which will be used to call your callbacks when you want to receive messages. Method "close" should be called in the end, it interrupts the thread. Three methods of this class correspond to the commands that the server understand:
1. consume. Sends "REC" command. Takes regex pattern and callback. When there is a message sent, part of which matches the pattern, callback gets called (by only one thread per client instance).
2. stop. Sends "STOP" command, canceling current "REC" subscription.
3. send. Sends "M" command, which actually publishes message to the server.
• Test1.java. Test program. Helps obtain some benchmarks. Currently, it tries to send 100 thousand messages of specified length to the server, using specified number of senders (each sender is an instance of Client1 class). There are also specified number of receivers (each receiver is an instance of Client1 class too). Senders send messages concurrently. Each sender sends equal number of messages, with total number around 100,000. First sender includes a string "P1P" into message, second one includes "P2P" and so on. First receiver subscribes to patter "P1P", second one - to "P2P" and so on. Hence, if number of senders more than number of receivers, some messages will not be received by anyone. Test measures 3 things:
1. Average sending rate (messages per second). This is total number of sent messages divided by the total time spent on sending (only sending, not delivering).
2. Average message delay. Every sender inserts current timestamp into message. Every receiver reads it and calculates delay.
3. Maximum message delay.