I have been trying to detangle the history of Adventure — the Crowther / Woods game — so I know which ports I should be playing in which order for the All the Adventures project. This turns out to have been very complicated as information is spread over a tangle of websites and not all of it is accurate. This is my best attempt to sort everything together.
This document will likely get edited as I tangle more things out. Looking at the full list of modifications gives a small glimpse into the insanity.
I link to sources whenever I can.
In the 1975 academic year Will Crowther starts what we know as Adventure. [Source.]
All Crowther family testimony is consistent with the 1975-76 date range. Responding to a direct request via e-mail, Crowther (2001) dated his original “Adventure” to 1975, “give or take a year.”
— Dennis G. Jerz
Will Crowther stops work on Adventure. The game is noticeably incomplete (there is a sign mentioning maintenance in a lower section of the cave where advancing further causes the game to crash). [Source.]
The most likely timeline places Crowther ceasing work on his original game in early 1976.
— Dennis G. Jerz
Starting March, Don Woods discovers the source code and starts working on it. He produces a 250 point version [source] on his way to finalizing a 350 point version several months later.
I’m relying solely on memory which tends to be fallible (see above: the dwarf ‘vanishes’, not ‘disappears’) but my best recollection is that ADVENT.EXE first appeared on the PDP-10s at ADP (the old First Data in Waltham, Mass.) in 1977. It was an incomplete version which only had about 250 points worth of treasure. I seem to recall that there was nothing past the troll bridge but an ‘under construction’ sign or some such. I believe our copy came from WPI, but word at the time was it was developed at Stanford. Two or three months later we got the full 350 point game.
— John Everett
This version is then ported, in a fairly literal way, by Jim Gillogly into C. [Link to source code.]
The original 350-point version is separately ported July 1977 by Kent Blackett. [Source]
ADVENT.FOR: C REV. 17 ADVENTURES MODIFIED BY KENT BLACKETT ENGINEERING SYSTEMS GROUP DIGITAL EQUIPMENT CORP. 15-JUL-77 ORIGINAL VERSION WAS FOR DECSYSTEM-10 THIS VERSION IS FOR FORTRAN IV-PLUS UNDER THE IAS OPERATING SYSTEM ON THE PDP-11/70″.
This is followed by Bob Supnik, circa October (date given in his own source code). It has 366 maximum points and the cumulative efforts of Blackett and Supnik (probably) represents the first modification of the game past the Woods version.
There’s also a version of Bob Supnik’s port on if-archive, although I have been unable to confirm if it is exactly the same as the 1977 code.
The version I played was written in FORTRAN. One of the treasures was an African gray parrot in a pirate aviary, accessible by a rubber raft. There was also an office of some kind with the words “how do you spell relief?” on the wall, and after you read the message, you could use “Rolaids” to teleport to and from the well house. I think the magazine at Witt’s End might have been a copy of Byte magazine.
While this version has been lost, there is a downloadable Osborne port which may be derived from the same source although I have been unable to test it as of yet. The actual 1978 version seems to be a this archive.
David Long at University of Chicago starts on his own modification of Adventure (a process which goes on until at least 1980), although no versions from this year exist.
Peter Luckett and Jack Pike finish “Adventure II” by the end of 1978. [Source and executables]
David Platt writes yet a different modification of Adventure into a 550-point version. [DOS version]
Microsoft Adventure is released as a commercial game for TRS-80 computers, with both v1.0 and v1.1 dated as June 1979. Gordon Letwin, who wrote the Heath port from 1978, is the one responsible. [Info from the Game’s Manual] It includes some slight modifications which mirror the 1978 Heath version somewhat. [Downloads for v1.0 and v1.1]
This year has a “501 point” version of David Long’s code (“Version 5.2/2, October-79”) although it rather confusingly this might be a “side release” by an anonymous contributor. Even more confusingly the version might actually be intended as 500 points (?) and the vanilla compile goes to 496 points (??) although the version online is fixed to be at 501. [Source code] [501 point version compiled for online]
The David Long story gets very messy past this point into the 1980s and I’m not going into every variant from his code, but I should mention the last version from David Long himself was at 751 points. While this version was playable on Compuserve for a while, it now appears to be lost except for a remake by Carl Ruby into QBASIC.