The original Colossal Cave Adventure consists basically of a Fortran source file and a textual data file. These files would often travel from one installation to another via paper printouts: printed out at one site, typed in by hand at another.

The lines of WOOD0350’s Fortran source (intentionally or not) never exceed 80 columns regardless of your tab stop:

$ detab -4 advent.for | awk '{print length}' | sort -n | uniq -c | tail -4
  48 77
  52 78
  35 79
   3 80
$ detab -8 advent.for | awk '{print length}' | sort -n | uniq -c | tail -4
  58 77
  62 78
  39 79
   3 80

But the data file fits within 80 columns only with a tab stop of four. With an eight-space tab stop, four lines of the data file exceed 80 columns:

$ detab -4 advent.dat | awk '{print length}' | sort -n | uniq -c | tail -4
  55 71
  59 72
  67 73
  69 74
$ detab -8 advent.dat | awk '{print length}' | sort -n | uniq -c | tail -4
  59 76
  66 77
  69 78
   4 82

These are the four offending lines. The first comes from section 3 (travel table) and the rest from section 9 (bit flags).

108     95556   43      45      46      47      48      49      50      29      30
0       1       2       3       4       5       6       7       8       9       10
7       42      43      44      45      46      47      48      49      50      51
7       52      53      54      55      56      80      81      82      86      87

If your version of Adventure has suffered truncation at the 80th column at any point in its pedigree, you’d expect to see (1) going DOWN from Witt’s End is impossible; (2) the Cobble Crawl does not have light; (3) entering maze room 51 or 87 resets the maze hint counter.

I first became aware of this possibility in December 2025, when Mike Willegal sent me a fan-fold printout of HORV0350 (a SEL32/RTM port of WOOD0350 most recently touched by Ned Horvath) that he’d kept since March 1979. In that fan-fold printout, all four of the lines above are truncated and thus missing the last number.

Now, I have no evidence that HORV0350’s own data file was actually truncated. But anyone retyping the game from that printout could easily have assumed that the Cobble Crawl was meant to be dark, and that Witt’s End was meant to have no DOWN exit.

I see evidence that such truncation really did happen between LONG0501 and ANON0501/OSKA0501.

• ANON0501 reflows line (1), preserving the exits from Witt’s End, but truncates (2) and (3).
• OSKA0501 reflows (1) but truncates (2) and (3).
• MCDO0551 reflows (1) but truncates (2) and (3).
• ROBE0665 reflows all three.

My decompilation of the recovered LONG0751’s data file indicates that it did not truncate any of these lines; and my playtesting of the recovered LONG0501 indicates that it did not truncate (1) or (2) either. (It’s inconvenient to test (3) by playtesting.)

These observations are consistent with the hypothesis that first LONG0501 reflowed (1) but left (2) and (3) untouched; then ANON0501/OSKA0501 and MCDO0551 descended from truncated paper printouts of LONG0501 (or via a now-lost common ancestor that was itself descended from LONG0501 by truncation). Meanwhile LONG0751 and ROBE0665 descended, independently, from non-truncated copies of LONG0501.