Re:Adventure? (Score 2)

Zork was developed independently at MIT LCS (the Dynamic Modeling group, specifically), definitely *not* the same as the MIT AI lab, though they were in the same building. It was certainly inspired by Adventure / Colossal Cave, which swept the ARPANet in early 1977; Zork was the usual hyper-competitive MIT response, "we can do better than *that*."

Muddle was not a particularly distant cousin of Lisp; it anticipated quite a few things that later showed up in Common Lisp.

The FORTRAN version was complete as of the time the code was lifted from MIT-DM; the DEC engineer who did the translation did most of the work during the aftermath of the Blizzard of '78. It didn't keep up with later additions to the MIT version.

Its parser was an effective demonstration of how much understanding of imperative English you can cram into a very small piece of code running against a very constrained world. Actual computational linguists never came looking for advice, though.

The Infocom releases of Zork (1, 2, and 3) started with the original MIT game, but broke it into parts, with at least some new material added for each part. Thus, the sum of Zorks 1, 2, and 3 was quite a bit more game than the MIT version, but each of them was smaller than the original. Zork 1 had very little new material; Zork 3, quite a lot.