Both MAXC's were PDP-10 re-implementations, and ran TENEX. So the basic system is still around, not sure if they had any interesting local hacks (well, probably PUP support; MIT tried to put it in MIT-XX, so it may still exist on thats backup tapes). Interesting; I also have a large collection of maps: http://mercury.lcs.mit.edu/~jnc/tech/arpanet.html#Maps all with hi-res versions (click on the thumbnails). Some of the ones at the CHM I don't have, but the coverage time-wise is similar. I also have a modest collection of hosts.txt files, ranging from Jul-77 to Apr-94. The first 'Assigned numbers' RFC was #739, from November 1977. It never contained host addresses. There were a very few early RFC's which contained host adresses (#226, #229, #236), but pretty quickly host addresses were done via the hosts.txt file, distributed from the NIC. (Given the churn rate, using RFC's didn't make sense.) Early RFCs about this are #606 and #627. There were a bunch of RFC's that reported on 'host status' (e.g. how their software was doing), but their goal was different. PS: A number of people are leaving out the definite article before 'ARPANET'; this seems to be popular these days (especially in the UK it seems, not sure why), but it is incorrect. Also, the correct spelling is all capitals (check e.g. through old RFCs). Until of course the AP gets their hands on it (I'm breathlessly awaiting their announcement that the U.S. President's reference is to be referred to as the 'white house'). Noel