Digital Analogue

Curses

I recently replayed Curses, an interactive fiction offering from 1993 which I'd played back in the mid-90s (although I'm not sure how far I got then—I didn't finish it).

"Interactive fiction" is the name of a whole genre of... works... that was inaugurated in the mid-70s as text adventure games. ADVENT was perhaps not the first, but certainly the first to gain notoriety. Zork became even better known, especially after being split into three parts (to fit micro-PCs of the time) and released by the new company Infocom.

The thing about one of these games is that there are no graphics. Thus, much like reading a book, you are free to establish your own images for how things look, and sound, and everything else. Unlike a book, if there is a piece of the world that catches your interest, you can investigate it further, and there's at least some chance that the program will respond. Other than the very earliest versions of ADVENT, which focussed a great deal on describing one corner of the Mammoth Cave complex in Kentucky, the early games were very much puzzle-driven. Whatever the cover story, the real reason you were there was to solve a whole bunch of puzzles, one after the other, and when you solved the last one you were done. As time went on, the Infocom offerings started becoming much more like a new form of literature, with clear plots and characters, although they never threw off the puzzle paradigm entirely.

Infocom had basically faded from the scene by the end of the 1980s, so when Graham Nelson wrote Curses in 1993 (well, he released it then—he'd been writing it for a while), it was very much following in their tradition. Indeed, he wrote his game in a format that could be run by the same bytecode runtimes that ran the old Infocom games (and to do that, in a rather Knuthian move he first had to write a compiler to compile to that bytecode format, and a programming language for it to compile—known as Inform, it's still in use today).

So, enough history. What about the game?

Well, for starters, it's really really hard. Funny! And extremely erudite. But hard. Which is more than a little frustrating, although of course it makes it all the more rewarding when you can solve the puzzles yourself. I won't lie, I needed quite a bit of help from the walkthrough—I wish it had been in an Invisiclues format, because a few puzzles got spoiled by accident that way—but I also got rather a lot of it on my own. There was a mix of answers that I whacked my head over—I should've gotten that; answers that I was glad I eventually looked up, because the puzzle was one I just never would've gotten; and a few that I felt a little betrayed over because even after seeing the solution I didn't think it was a fair puzzle.

But honestly, I think the game's biggest fault, using the author's own terminology, is that it gives the appearance of being extremely "wide" while actually being relatively "narrow". At one point in the middle of the game (when I'd acquired 195 of 550 points), I identified 28 different puzzles that I had found out about and hadn't solved yet. Of those, two were red herrings, one was a plot device, two pairs solved each other (what is X for?/how do I Y?), but the remainder were really puzzles to figure out. The problem was, most of them simply couldn't be solved at that point, because they were waiting on some other thing to happen first. Some of the things they were waiting on were things I hadn't even identified as puzzles. If there really were 28 different things I could be working on, that'd just be a very "wide" game. But while there seemed like that many, there were actually only five or six things I had the tools to solve then; I just didn't know which ones they were.

Related to that complaint is that the game was unpredictably unforgiving. Better not enter area X yet, because you'll only be able to enter it once, and you don't yet have all the keys to all the puzzles inside. How do you know when you're ready to tackle X? Hell if I know. You just go through, try to solve everything, see if there are any obvious loose ends—and hope you're seeing any loose ends that are there—and if there are, try to solve them or restore to before you entered the area, planning to try again later. There is one in-game mechanism that makes it somewhat forgiving (by letting you return to some areas later), but in order to discover it you have to pay attention to one paragraph very very early in the game, make note of an "action" (if you can call it that) done to one object, and then a few hundred turns later perform the same action on a not-obviously-related object. Or, just read the walkthrough. Even aside from that, though, knowing that there are three areas you can cross into, each irreversible, and they have to be done in a certain order but you can't know what it is... that's just discouraging.

And yet, and yet. You just can't put this game down, because the clever prose and the cleverer responses are worth it. The other has a very dry, very English sense of humour, and throws in classical references like parade candy. There is one throwaway item that you can use that has no real significance but let him throw in the phrase "alea iacta est" in context. And some of the puzzles just need the right background: on at least three occasions, I'd solved a puzzle fairly quickly and easily (though not always without work), only to run across its solution later on in the walkthrough with some apologetic remark about how hard it was and explaining the "tortured" reasoning one might use to get there. Not so! Of course, requiring significant outside knowledge is a bit of a sin for these games usually, and I feel that the very very last puzzle in the game is either dead easy or totally unfair (for me, the former). But solving one of those? It's like being in on a very elabourate in-joke, which always gives a bit of a buzz.

And of course we need to place the work in context. The early 90s were looking to be a dark age for interactive fiction, with Infocom out of the picture and very few people producing any new IF, with much of the output leaning heavily on the puzzle-puzzle-puzzle model. Curses broke that, reviving the best of the Infocom style, and boasting an improved parser and framework that—if imperfectly—reduced the amount of time you'd have to spend playing "guess the verb" or otherwise trying to express your idea in a way the game could understand. It, and Inform, set the stage for an IF renaissance that continues today.

I can't leave this without listing out some of the bugs I found in the game (more or less spoiler-free in case anyone plans to play it):

• On the ship, there is a flag hanging from a flagpole. In the description, it is described as an ensign; but ENSIGN is taken as a synonym for the flagpole, rather than the flag. This makes the puzzle basically insoluble.
• In the dream about the octagonal room, if you type WAKE the response is the standard "The dreadful truth is, this is not a dream." Of course, it is a dream, and solving this puzzle requires you to realise this.
• Perhaps not a bug, but the map of Hamburg claims to have a marked grid. The thing to look up? Looks nothing like grid coordinates.
• After you have safely sprung the oubliette trap, you can take the stone and the "wedged" message remains.
• The Rod of Infinity answers to the names FEATURELESS and MAHOGANY even after being identified. This requires a bit of creativity in identifying future rods.
• In the Lawn Ornaments area, if you go some direction other than east or west, it responds that "there are paths only west and northwest". Oops!

"I can walk into a Christian church and evoke a plethora of symbols that help lead people to compassion. The vast majority of people might suck at loving their enemies and forgiving those who tresspass against them, but at least the words are there to plant the seed of the idea. Whereas I walk into a UU church, and their secular symbols focus entirely on "justice," rewarding the innocent and condemning the guilty." --Jonathan Prykop