Keep in mind that BSD didn't really have releases. When you called up for a tape, it wasn't made from some master tape, but rolled off from a system that had right version on it. I know Kirk told me this once when I was chatting with him about tapes, RMS and other things. Thje version control wasn't quite as strict as things are today, so I'm not surprised there's some variance in images from place to place around the net. We have SCCS, and multiple images. I think the best we may be able to do is to do the this was copied from that with these changes and produce a tree of inheritance... IIRC, SCCS has issues with moved and removed files that makes it hard to reconstruct things exactly with it. I know RCS and CVS had these issues. The historical unix git repo has

  remotes/origin/BSD-4-Snapshot-Development
  remotes/origin/BSD-4_1_snap-Snapshot-Development
  remotes/origin/BSD-4_1c_2-Snapshot-Development

branches. Also, version numbering was kinda hazy. Kirk has a big listing in his house of 4.5 BSD. This is post the first 4BSD release, but not the 5BSD release. They had thought they'd do a 5BSD, but they had all these contracts with 4BSD in them, so they were basically forced to do 4.1BSD instead, so the "4.5BSD" thing is basically an early version of what we know know as 4.1BSD.... So between these two quirks, I'm not surprised there's not an 'untainted' version of 4.1BSD around... I'm guessing the tape that has the July 1981 date on it was made in early 1982 and the extra files with the weird dates are just an artifact of when the tape was made and that people had used the 'master image' system in the mean time, if for nothing else than logging into and running the make tape script :)

Warner