Lost Infocom Games Discovered

I Don't Believe in Imaginary Property writes "Archivists at Waxy.org have gotten a copy of the backup of Infocom's shared network drive from 1989 and are piecing together information about games that were never released. In particular, there is the sequel to The Hitchhiker's Guide to the Galaxy called Milliways: The Restaurant at the End of the Universe, and there are two playable prototypes of it. And yes, they have playable downloads available."

Re:Infocom was a damn good company

[Mr.Radar]

That's actually already happened, in a way. After Infocom went out of business the fan community reverse-engineered their VM (the Z-Machine [wikipedia.org]) and Graham Nelson designed a new language and compiler for it (Inform [inform-fiction.org]). That, along with other interactive fiction languages/toolkits that compile to their own VMs (TADS [tads.org], Hugo [generalcoffee.com], AGT, ALAN, and many more) and a small but dedicated community has ensured that interactive fiction hasn't died out.

Every year dozens of new games come out, usually for the two major annual competitions (the IF Comp [ifcomp.org] and the Spring Thing [springthing.net]). Most of them are shorter than "commercial-era" games, mainly because they're written by hobbyists who don't have the time and resources to commit to building large games. They run the gamut from puzzle-focused games in the style of Infocom to story-focused games that eschew large numbers of elaborate puzzles to focus on story, and there are also more experimental and artistic games that try to push the medium in new directions. The IF Archive [ifarchive.org] has an extensive collection of these games, and there are several [tads.org] review [wurb.com] sites [ifreviews.org] that attempt to catalog and organize the archive. The IF community has long had rec.arts.int-fiction [google.com] and rec.games.int-fiction [google.com] at their center, though with the rise of blogs and web forums it has started to fragment some.

Strange Description...

[wdr1]

"Archivists?"

Last I checked, Andy was just one guy.

-Bill

Re:Infocom was a damn good company

[wrook]

I just want to add a small detail to this. If you are *at all* interested in literate programming, you have to check out Inform 7. To say they've pushed some boundaries is an understatement. It's one of the most innovative things I've seen in years. So even if IF isn't your bag, take a look.

And now writing the games is a game...

[Mr2001]

After Infocom went out of business the fan community reverse-engineered their VM (the Z-Machine) and Graham Nelson designed a new language and compiler for it (Inform).

This is fascinating not just for Infocom fans, but also for programmers. For example:

The Deathbot Assembly Line is a room. "Here is the heart of the whole operation, where your opponents are assembled fresh from scrap metal and bits of old car." The dangerous robot is a thing in the Assembly Line. "One dangerous robot looks ready to take you on!" A robotic head, a drill arm, a needle arm, a crushing leg and a kicking leg are parts of the dangerous robot.

That's source code. Inform 7 has been out for a couple years, and I've been working intimately with it for most of that time, but I'm still impressed.

Re:And now writing the games is a game...

[Mr2001]

Now, obviously Infocom did some pretty amazing things with Inform (1-5?) and Inform 6, before Inform 7 came out.

Actually, Infocom used their own language called ZIL, which looked a lot like LISP with angle brackets. Inform was created in the 90s after Infocom's virtual machine had been reverse engineered.

But now I'm curious (as a programmer unfamiliar with either environment outside of this /. thread!), what would have been the "Inform 6" way of constructing the "source code"?

Something like this:

Object assembly_line "Deathbot Assembly Line"
    with description "Here is the heart of the whole operation, where your opponents are assembled fresh from scrap metal and bits of old car.",
    has light;

Object -> dangerous_robot "dangerous robot"
    with name 'dangerous' 'robot',
              description "One dangerous robot looks ready to take you on!";

Object -> -> robot_head "robotic head" with name 'robotic' 'head';
Object -> -> drill_arm "drill arm" with name 'drill' 'arm';
Object -> -> needle_arm "needle arm" with name 'needle' 'arm';
Object -> -> crushing_leg "crushing leg" with name 'crushing' 'leg';
Object -> -> kicking_leg "kicking leg" with name 'kicking' 'leg';

Re:And now writing the games is a game...

[Mr2001]

Think about it--Inform 7 code has just as much arbitrary grammar as Inform 6 (or any 'classic' programming language) did, and is only marginally faster to read if you're not familiar with Inform 6 code.

True, the syntax is arbitrary either way, but what really sets I7 apart is how you can express certain concepts, which tend to show up over and over again in interactive fiction, in a way that's more natural than writing loops and conditions. Those concepts are rules governing behavior, relationships between objects, and descriptions of sets of objects.

Here's an example of all three:

After printing the name of a person who attends an accredited university, say ", Ph.D".

This is a rule about what to do in a certain situation: aspect-oriented programming, essentially. Here the situation involves an activity (printing the name) and the object which is the subject of the activity (any person who matches the description).

"A person who attends an accredited university" is an object description, which can be used in various ways as a condition -- does object X match the description? -- or as an iterator: show me all the matching objects. Here, "person" and "university" are kinds of object (classes) and "accredited" is an either-or property (a boolean flag).

"Attends" is a relation that expresses the link between a student and his school. Here it's being used as part of a description, but it can also be used in a condition ("if the player attends Harvard") or changed at runtime ("now the player attends MIT;").

These concepts can all be expressed in Inform 6 or any other OOP language, using properties, methods, loops, etc. But making them fundamental parts of the language gives them a whole new life.

RTFA - you wont regret it!

[BlackSabbath]

Wow. What a blast from the past! Sometimes, good things can result from wrongs (which arguably Baio's publishing of the emails was). The historical value of this stuff is undisputed. However the truly brilliant bits are the responses to the blog itself, especially from those actually involved.

There has never been a Slashdot submission where reading TFA was a greater pleasure.

The Original Hitchhiker's Game Online

[boot_img]

... can be found here [bbc.co.uk]. I never did get through that sulky door. Now I can relive the aggravation all over again.

Not 64K games - full 128K virtual machine

[drerwk]

in 64K

The Zork interpreter was a full virtual memory machine running in a 128K address space. Even the 32K Apple was able to run full 128K games swapping in from disk. No data was written back to disk, other than game saves. In 1985 the X-ZIP was written - I implemented the Apple IIc version. It was a full 256K virtual machine which was needed for AMFV. I was even able to keep users from having to flip the disk by writting a custom RT (Read Track) as opposed to the standard RWTS. This let the 5.25 in Apple floppy hold 164K per side.

In many ways the Z-Machine was similar to the JVM - cross platform in the day when there were still 30 platforms. The same day the code ran on the Dec20 it compiled for all of the micros.

Re:ATT: Michael Bywater. Was that Trinity?

[mbywater]

Michael, if you or your former co-workers read this, was that email the seed that ultimately brought forth the genius that was Trinity?

I wish I could say it was, but I don't think so. I agree about Trinity; and the remarkable thing about Brian Moriarty was that he could do that, at (if you like) the top end of the genre, while also writing "Wishbringer" which was theoretically for youngsters but managed to be really captivating for adults, too. The opening scene of Trinity, in Kensington Gardens, is still for me one of the most perfectly realised of all IF episodes. (Then he went on to do Loom, genuinely a kids' game, and even that was atmospheric and memorable. A remarkable man, Moriarty.)

Re:Meet Steve Meretzky...

[StaticEngine]

I too met Meretzsky, although this was at the Game Developers Conference in 2006 (or '05, I can't remember). He was talking to a Valve employee, and I had been drinking. I slurringly interrupted, and thanked him for inspiring me to become a games developer, then gushed about Planetfall and asked if he'd gotten my email to him about Splashdown [staticengine.com], my IFComp entry that was a blatent Planetfall ripoff.

I learned three things from this encounter:
1) Don't talk to your idols when you're drunk.
2) People have generally heard enough about things they worked on two decades earlier, and don't want to hear about it anymore.
3) Steve is really, really tall.

Re:Nostalgia

[drerwk]

I worked at Infocom writing interpreters mostly for the 6502 platforms - Apple II, C-64, Acorn, Atari800. The early games were 128K games. They ran in a virtual machine, with 128K virtual address space; including on the 32K Apple. The disks were 144K per side on an Apple II and I think that was the smallest 5.25 in disk capacity. There were no games above 256K through 1989 at least.

Re:Infocom was a damn good company

[YourExperiment]

Ah, those were the days, when it was actually possible to make money selling text adventures! I made a few attempts to write games myself back then, in Sinclair Spectrum Basic.

Today's interactive fiction authoring systems are more like general purpose programming languages, but with specialised syntax for creating rooms, objects and so on. There's very little that can't be implemented in them, with a little effort, and none of the frustration of being limited to binary flags and the like. TADS 3 has a lot in common with C++.

Inform 7 is a special case, with its natural language type syntax. I7 source code has to be seen to be believed. For instance: -

The jail cell is a room. "A darkened dingy room with water dripping from the ceiling." In the jail cell is a crowbar. East of the jail cell is the corridor.

That paragraph is valid I7 source, and does exactly what it sounds like it would. Things get a little more convoluted once you start declaring complex logic, but it's all natural language. The jury's still out on whether this is actually a useful way to program, but it's certainly an interesting one, and well worth checking out for sheer novelty value!