On 2011-06-16 20.08, Mark Benson wrote: On 16 Jun 2011, at 18:55, Mark Wickens wrote: I've yet to dip my toe properly in the emulated water There is a sneaky way around the SIMH-can't-talk-to-host limitation: Stick SIMH in a VM :) That way it can talk to the host running VMware, just not the IP stack on the VM itself. Ah, that's clever. Although I'm still thwarted given that atoms don't have the required VM instructions. You are correct, Atom CPUs lack the VTx component required for true Virtualisation. For some confounded reason Intel thought it might not be needed in a Netbook CPU ;) In my case I could easily run SIMH on my Mac Pro (I take it it'll compile on OS X? Maybe need darwin-ports or something like that?) - it has dual ethernet interfaces and one is redundant t the moment :) === So, do I need more than one ethernet card in the Linux machine to run a DECnet bridge out to HECnet? I'm not aware if DECnet and TCP/IP will play nicely on the same LAN or if I need a separate one for DECnet connections? DECnet and TCP/IP are just two different protocols on ethernet. No issues or problems with that. In fact, you also have at least ARP also running on the same ethernet (yet another protocol). No, you do not need to have two interfaces on a bridge machine, unless possibly if you want the machine that runs the bridge software itself to also be participating in the DECnet network as a node itself. The bridge program should compile just fine on Linux systems. If there are any issues, just let me know. There is a makefile, so just typing make in the directory where you have the files should produce the binary. Then you just need to figure out where to connect to, setup the config file, and start running. The separate issue then is set configuration of DECnet itself on whatever (VMS?) machine you use. Johnny