Old Software Filing System Score! & Text Recog. question

Sharing a Filing system score! Husband sent me a link to code on GitHub that purports to be the original PDP-10 FORTRAN source code for the old Adventure game.

https://github.com/historicalsource/colossalcaveadventure

We got into a discussion whether this version is actually the oldest. He mentioned he used to have both a map and the source code of the version he played. I said I think I know where that is and within 30 seconds handed him the paper file folder with the FORTRAN IV+ source code for the DECSystem 10 under the IAS operating system and his original Adventure Map from 15 July 1977.

The code does not match the code on the link but is very, very close. It states it was modified by Kent Blackett, Engineering Systems Group DEC. Looking for differences manually is painful, we may end up entering it into a system so we can compare.

I’m wondering, does anyone have any clue how well text recognition will work on computer source code?

The original paper is still continuous so it’s going to be hard to scan without separating the pages. It’s also over 1/4 inch thick so a major project to retype it all and not something I really have time to tackle now.

2 Likes

Damn! This is a temptation to again join the cast of those who do not get enough sleep from chasing themselves through twisty mazes!

I doubt I will find my map. I was playing it while working at a lab in Munich at the time.

Would you consider this important enough to hand to a company that can do the job at a professional level?


JJW

All alike or all different? :slight_smile: :laughing:

First approximation is that I’ll try to take medium res images with a digital camera on the copy stand I am using for my Historical Society digitizing project and we’ll go from there.

We’ll also scan the original map that my husband made by hand as he played the game it after he got the code running on his HP 21MX minicomputer and later on his homebuilt Z-80 system running CPM.

Unfortunately, as far as we know we do not have any of that code left, it was on 8 inch floppies we got rid of in 1997/1998 when the drives died and even he could not resurrect them.

We do however have a garage full of stuff, which may or may not contain a printout or 3…

It was the early-to-mid 80’s when I first ran into Adventure. We called them “trash-80’s”. The lab where I worked had one person who was a technical expert on hardware interfacing the Z-80s as control systems and one person who was a technical expert on the software. One of my German colleagues had rather robust nicknames for the two (hardy and softy). Yes, I remember those 8 inch floppies. I think that I may have some stored somewhere in my lab in a box high up on the cabinet (not from the game but from data collected at that time). Hard to believe that I carted those things around between two different continents.

(sigh)

Memories, memories. This makes me also want to dig out my copy of Myst, fire it up on my old OS7 Mac computer from storage, and see whether I can finally get through the water maze. I should have my log book somewhere for this game.

I will be curious to hear how this approach turns out. Please post back with updates as you go along. And above all remember to “look” and seek “help” when things seem to be too deep.


JJW

My introduction was 1978-1979 ish. The version I ran on was on a CDC machine. Written in FORTRAN

Anyway, on the digitization, will get too it as soon as I can, I spent a lot of today working on scanning a bunch of old 35mm Data General marketing slides from 1981-1986. As soon as I have digital copies they are headed to the Computer History museum.

I looked briefly at the code linked in your original posting. It looks, to my memory, rather different from the code I recall going through way back when, when I played Advent on a VAX 11/780 (VMS 1.0) around 1979. I distinctly recall seeing a comment regarding authorship by Crowther and Wood, which is absent from the code you found - suggesting you have an earlier version than I played. That would be consistent since the DECSystem 10 predated the VAX architecture.

I don’t have any Advent maps around anymore, but I do think I have my original Zork maps (from the original version, written in MDL at MIT’s Lab for Computer Science, but later ported to FORTRAN for Digital’s hardware.

I don’t know how well text recognition will work on FORTRAN code. My guess would be quite poorly since the software is built with libraries for various languages that it can interpret.

The code is quite interesting as a classic example of “spaghetti code.” I haven’t seen stuff like that since the 70’s and early 80’s.