Hello! I just wrote rthread library that makes distributed computing look like writing multi-threaded programs. Here is a simple example that runs print_hello function remotely and prints the name of the remote operating system. *** import rthread import time def print_hello(): import commands print "Hello from %s" % commands.getoutput("uname -o") return 10 r=rthread.start_remote_thread(print_hello,(), remotehost="hostname.domain") time.sleep(2) print "Finished: %s" % (r.status==rthread.status_finished) print "Returned: %s" % r.retval *** The only thing required from remote hosts is ssh server and executable rthread_server (included in the package) in PATH. (The library executes "ssh -C -c blowfish remotehost 'rthread_server'", and uses stdin and stdout of that process as a communication channel.) There is no need for interface definitions, no need to open communication ports, no need to copy remotely run function code to remote hosts. The library is working (three examples are included in the package), but it still lacks some features, especially related to serialisation of objects. If you are interested, please check: http://www.cs.tut.fi/~ask/rthread/index.html Yours, Antti Kervinen