I did figure out a way that they might be getting
their passwd files.
It looks sort of like they are scraping postings from the usenet
archives, and putting together login names with systems from
back in the day.
Good catch, especially uucp lists and maps seem very relevant here.
All paths go through mimsy :-) later
mimsy.umd.edu so... a
late 80's UUCP map obviously figured heavily in the simulated networking.
By necessity: any retelling of 80's era networking will have so much
source information available from UUCP mapping projects and Usenet
and mailing lists, than any other more closed sections of the network.
(BITNET and later NSFNet are also up there in terms of preserved
message/map volume but by no means as huge). It's a little biased
but by no means wrong (especially considering my link to the
network was largely uucp back then too... I can easily dig up usenet
groups from the 80's that I overlapped with, but the bitnet-based
listserv's that were popular back then are far harder for me to find references to.)
I suspect Henry Spencer's preservation of usenet archive tapes at utzoo will
figure heavily in any retellings in the future. History is written by the
victors, it seems so! If victors == usenet nerds :-)
Tim.