The Stack

I’ve made it through level 3 — it turns out this isn’t the level with the trampoline where I got stuck previously after all. In fact, it only has one challenge room at all, and no shop. I’m assuming that this is a sort of special “Oops! All Maze!” level and that I won’t see the game go to the same extreme again.

Level 3 also introduces something I wasn’t sure I’d see: optional items. There’s a dead end in the maze where you can find a key, guarded by an arm from a window like the toll-taker, except he demands a watch. You can find a watch further along in the maze, where you probably haven’t been yet when you meet the watch-taker for the first time. If you don’t have a watch, he instead takes all your gold and lets you through. The familiar toll-taker is also found near the exit, as usual.

Now, I said before that the toll-taker requires all the gold in the maze. I don’t really know if that’s strictly true. I just know that whenever I haven’t given him all the gold in the maze, he’s sent me away for more. Maybe he has a minimum acceptable amount, but takes as much as you have. Maybe he’ll accept whatever’s left over after the watch-taker takes his cut, which would make the actual watch downright useless. But even assuming that you can’t let the watch-taker have any gold and still win the level, there’s an alternative to using the watch: Just go the watch-taker’s dead end before collecting any gold.

On top of that, if I’m not mistaken, you don’t even really need the key he’s guarding anyway. Past that point there are two locked doors, and two keys you can find. Making you look for the watch at all is pure trollery. Maybe I should keep the watch — unlike gold and keys, inventory is preserved across levels, and I could see this game suddenly requiring you to have something you gave away two levels ago.

I finally found the save room in maze 3, and close by, a single challenge room. On the other side of the challenge lies more maze. I have not yet found any more rooms. This maze is a lot bigger than I remember them being. And on top of that, the one challenge room I found? It’s another maze. A smaller one, with different rules and different presentation: multi-tiered, isometric, and with a sort of node that you disappear into at every bend or intersection. Key to solving it is realizing that pressing the primary action button while in a node switches you to a view of the inside of the node, where items and elevator platforms are hidden. That’s the kind of puzzle you find in the rooms. There’s always a bit of guesswork before you can start solving them for real.

Some other notable puzzles I’ve seen so far:

The very first puzzle you encounter on level 1 involves a spider-demon that sits in place and spawns spiders, which follow you erratically and attack you. It doesn’t take long to figure out that the action buttons make you step on spiders, or occasionally pick them up and eat them, but it’s a losing battle: the demon just spawns more to replace any that you kill, and you can’t do anything to stop it. Until, that is, you buy some bug spray from the shop. Then the two action buttons do two different things, one stomping and one spraying.

A sort of wizard’s laboratory, with a row of colored flasks, a bubbling cauldron, and four wizards playing poker — although before long, one of them gets irritated at another and turns him into a frog. Get close enough to the flasks, and you enter a special mode where you select stuff to throw into the cauldron. The correct combination is found elsewhere in the maze. Even once I had that, it took me a while to find the place I had to stand to take a drink out of the cauldron. Drinking the correct concoction temporarily turns you into a mouse, letting you escape through a hole in the wainscotting on the opposite side of the room — but only if you can evade the cat that’s been sleeping by the poker table until that moment.

A sleeping dragon on a treasure hoard, albeit apparently not the sort of treasure you collect in the maze. There’s one item in the hoard you need, but it’s on the other side of the gouts of flame the dragon is breathing out of its nose in its sleep. You can, however, reach a heap of large gemstones, and pick them up and throw them (with no ability to alter distance or direction). I suspected early on that the solution was to throw the gems to plug the dragon’s nostrils, but it still took me multiple visits to find the exact spot you have to stand on to get them in place. And even then, you have limited time before the gems burn up.

So there’s action elements in most puzzles, just as there’s a puzzle element in most of the game’s action scenes.

I’ve made it to the third maze. Despite a considerable time spent exploring, I have yet to find a single room — not even the save room, with the result that I have to restart from the save room in level 2 every time I run out of health. (In case I have not yet made it clear: This is not a friendly game.) Let’s look at the patterns established in the first two mazes and assume they hold through for the rest.

Each maze starts and ends with a fight against a monster on a bridge. Coupled with the fact that you can’t save between mazes, this means you wind up doing two bridge fights in a row. The maze-start monster and the maze-end monster are distinct, but there seems to be only one maze-start monster (a troll with a club) and one maze-end monster (a squat, grinning, vaguely lizardish demon with a flail), just palette-swapped and possibly with increasing health. The maze-start monster can’t actually defeat you, because the price of defeat is always being ejected from the room with slightly less health, and before you’ve entered the maze, there’s nothing to eject you to. All it can do is block your way indefinitely.

Every maze I’ve seen has the following rooms: the save room, where you can save the game; the shop, where you can exchange the gold you find scattered through the maze for items needed to solve puzzles; the exit with its aforementioned bridge fight; and seven challenges, for a total of ten rooms per maze, matching the Steam blurb’s claim of “five hellish levels of treacherous tunnels with more than 50 raucous rooms of gameplay”. Three of the challenges yield items you’ll need to exit the maze. The other four are just obstacles, blocking the path to sections of the maze. One of the obstacle rooms seems to always be a special arena fight that you need an item from the store to win. All challenges disappear after completion, turning into just more corridor.

There are occasional locked doors in the maze, not leading to rooms, just blocking the way until you collect a key lying in a dead end somewhere. One locked door shortly before the exit is special: the key is provided by a sort of toll taker who reaches a very long arm from a nearby barred window, demanding all the gold in the maze (less the price of all the items in the shop). In this way, the game motivates you to explore the maze thoroughly, as coins could be anywhere. Then it trolls you by including long, winding, trap-filled sections that don’t hold anything worthwhile at all.

Like many other maze-themed games, Litil Divil makes it basically essential to draw a map. Oh, it has an automap that fills in as you explore, but it only shows a meager 8×8 section of the maze at a time (in maps that can be in the neighborhood of 40 tiles on a side), and doesn’t communicate essential information like “How do I get back to that one puzzle room that I think I have the resources to solve now?” Back when I played this game for the first time, I drew crude schematics of the maze, just lines showing how things connected, bendy in places because they weren’t drawn to scale. That was okay in practical terms, and acceptable when I was just playing a game I had picked from a bargain bin. But when I’m looking at closing out a twenty-year pretend obligation? That feels momentous enough to warrant something tile-accurate, drawn on an actual grid.

The game makes some attempt at thwarting this: one spot in maze 2 pulls the old Infinite Hallway trick, periodically teleporting you backward inconspicuously until you get suspicious and turn around to see if you’ve actually gotten anywhere. But for the most part, the only thing getting in the way of accurate maps is the mere fact that it’s hard to tell how long a long, featureless corridor is. Your movement isn’t bound to the grid, so the only good way to tell when you’ve advanced to the next tile is to look at the automap — and you can’t even tell then, if the map view looks the same centered on the next tile as on the current one.

The game also discourages simply taking the time to draw more than rudimentary maps, by having Mutt’s health bar steadily drain all the time when you’re in the corridors. It should be understood that health is purely a corridor thing; it’s not reflected in the challenge rooms at all, not even the combat challenges, which effectively have their own per-room health. Challenges do affect health, though: completing one restores a large measure of it, and leaving a room without completing it takes some away. To minimize loss, you should always go to a challenge you can win by the most direct route available, avoiding any traps along the way, and win it in one try. In practice, it makes more sense to visit the maze’s save room whenever you beat a room that you don’t want to repeat. But directness is important regardless. You know what helps with that? Maps.

The overall effect is that I always leave the save room with specific intent. Either I’m exploring, and expect to die in the halls, or I’m trying to do something specific and concrete, like beating a room or looping over a known sequence of hallways to collect the treasure, and then make it back to the save room alive.

This blog has seen a request that I finish up Deus Ex next. I want you to know that I’m not ignoring it, but I have reasons to want to put it off for a little bit. In the meantime, let’s pull out something from the depths of the Stack.

Litil Divil is a 1993 game by Gremlin Graphics, a studio otherwise unfamiliar to me, but which apparently had the resources to advertise this game noticeably in the videogame magazines of the day. It’s a game of cartoony pixel art and cartoony sensibilities. You play as a lesser devil called Mutt, possibly because of his enormous bulldog-like jowls, as he traverses the Labyrinth of Chaos to reach the overworld and retrieve a pizza. Please understand that most of this information, including the protagonist’s name, comes out of the manual rather than the game itself. The game has no intro other than Mutt dancing on the title screen. (Apparently some later ports add an animated FMV intro cutscene, which you can see on the game’s Steam page.)

You could call it a “variety game”. The labyrinth proper is a series of largish grid mazes with occasional underpasses to make it harder to navigate. Some tunnels have doors to unlock or traps to dodge, but the real challenges come in the rooms scattered through the maze. Each room is a mini-game, which could be a puzzle, or a side-view fighting game, or an isometric platformer, or a shell game, or whatever else the designers came up with. All such challenges come without instructions; figuring out what you’re supposed to be doing, and what the controls do in that room, is just part of the challenge. Some challenges require inventory items obtained in other rooms, and this too is something you have to figure out on your own, usually by failing the challenge a few times.

Back in the mid-90s, this game was in rotation as something I’d pull out and try to finish from time to time, and I played it enough for a snippet of the background music — FM-synthesized, almost offensively jolly and lightly discordant — to get stuck in my head occasionally even to this day. (It’s always a little weird to hear music played aloud that you’ve only heard in your head for a long time. It never completely matches what you remember.) If I recall correctly, I got as far as the third maze, which could be the last one for all I know. There, I got severely stuck on a challenge involving a trampoline and could progress no further. I’m given to understand that it’s basically a timing thing, but I couldn’t seem to get the right timing no matter how many times I tried. The game’s low framerate definitely hurts timing-based puzzles.

I’m told that the MS-DOS version of this game is inferior to the Amiga CD32 version, but the MS-DOS version is what I have (and is what’s available through Steam and GOG). The CD32 has one feature that would make a significant difference to gameplay: the ability to rotate the camera 180° in the maze. In the version I’m playing, you can only change the camera facing at bends or intersections in the tunnel, which means that if you want to turn around, you have to go as far as the first such point and do a K-turn. Until then, you’re stuck doing the Crash Bandicoot thing, walking towards the camera with diminished visibility for any traps you’re approaching. This is an annoyance, but possibly a deliberate one, as annoyance is something of a theme. Mutt reacts with exasperation toward the player whenever he falls in a pit, and the manual cover features the tag line “And you thought you had a bad day!”

Okay, ParserComp 2022 has been over for a good few days now. I said at the beginning that the number of games seemed very comfortable for the deadline, but I wound up scrambling at the end nonetheless, due to interference from life for a couple of weeks in the middle. But most of the games were pretty short (and I gave up on a couple of the longer ones before reaching an ending), and I wound up giving a brief writeup to all of them but one, The Euripides Enigma. I could write it up now — there’s no rule saying you can’t play or discuss the games after the Comp’s deadline! — but I just don’t feel like it.

After the first few games, I fretted a bit about whether this entire event was a bit retrograde, more concerned with rehashing the past than with exploring new possibilities. I suppose that’s how a lot of people see IF in general: aren’t we all just still trying to be Infocom? But to me, amateur IF has always been more about taking paleo-IF as a starting point and striking out in new directions. But one of the main new directions people have struck out in over the last decade is the Twine Revolution and the shift to choice-based interfaces, and ParserComp is kind of set up to attract people who’d like to roll that back. Gladly, though, I found a good amount of experimentation here: new engines and interfaces, new forms of interaction. And if not all of it is entirely successful, well, that’s just in the nature of experiment.

Which is part of why I find the scoring system so unsatisfactory. Unlike the IFComp, which simply asks you to rate every game on a 1-10 scale, ParserComp imposes a rubric. Players are asked to rate the games in eight categories: Writing, Story, Characters, Implementation, and Puzzles each get a weight of 17% in the final score, with 5% going to Use of Multimedia, Help and Hints, and Supplemental Materials. Not only does this reflect the values of the comp organizer rather than the players, it’s clearly set up to favor the conventional over the extreme. Works are effectively penalized for doing a puzzle-free narrative, or a pure puzzle game without characters. Heck, one of the entries arguably didn’t feature writing per se. I hope the authors pay less attention to their numeric score and more to the reviews, where more nuanced reactions are possible.

Here’s another one it looks like I’m not finishing. The whole idea is that you’re spending the night in a spooky house full of Halloween monsters — at the very least, there’s a werewolf, a skeleton, and a mummy, and on the basis of the garlic and wooden stake in my inventory I’m going to go ahead and say there’s probably a vampire as well. Monsters are killed or escaped in time-limited chase sequences that can come without warning, so saving frequently is crucial. Other than that, it’s a largish exploration game with locked doors and secret passages. It has a bit of a problem with recognizing alternate phrasings, but nothing that an old hand like myself can’t power through with the aid of the in-game hint menu — very often, all I needed from the hints was confirmation that what I had attempted was the right thing, and all I needed to was to rephrase it until it worked. There’s my advice to the author: Accept more phrasings.

On the other hand, phrasing isn’t always the problem. Some of the puzzles are a bit too read-the-author’s-mind-ish, and I found myself playing mostly from the hints after a while. The reason I’m giving up on it at a mere 40% completion is that I finally hit a point that the hints don’t adequately cover — they advise recovering a ring stolen by a mouse by trading some cheese for it, but the mouse won’t take the cheese and I don’t know why.

Seems like it wouldn’t take a lot of reworking to turn this into a decent puzzle-based adventure game, though. It’s written in Adrift, but avoids the most common pitfalls of Adrift games, like overdescribing rooms.

I’ll have to state right off that I didn’t finish this game, as it soft-locked before anything very interesting happened, and it’s entirely possible that I didn’t even begin to see the real plot. If so, it takes a while to reach it. You spend a lot of time navigating the corridors of an arctic research station, looking for breakfast and a shower, getting some offhand world-building in the background, learning things about the player character and their coworkers. But maybe that’s all there is. The author says that this game is basically a demo for a new engine.

What I really want to comment on is the UI. First of all, the game has the output text trickle in character by character, like in a console videogame. This is just about tolerable in a game not made mostly of text, but it’s absolutely a bad idea for a game where you spend so much time walking back and forth through multi-room corridors. It just slows you down when you don’t want it, and if there was any button to skip the text animation, I never found it.

Secondly, this game takes autocompletion to an extreme. At every point, the command line is accompanied by a list of every word you could type as part of an acceptable command. The effect is that interaction feels a bit like a Monkey-Island-style point-and-click adventure, choosing words out of a menu. (You don’t actually pick them with a mouse, but you usually don’t need to type anything more than the first letter of each word plus tab to complete it.) Where one of the strengths of parser interfaces is the sense of boundless generality, this UI makes the player acutely aware of exactly how limited your input is.

Curiously, the output text imitates the form of traditional adventure-game room descriptions, with its lists of objects, a style that’s a consequence of a world model that the UI makes it clear isn’t actually present here. Like CGI lens flares, it’s technological artifacts reinterpreted as a style. This isn’t the first time I’ve observed this, and it will doubtless not be the last.

Here’s an experimental one. It’s a two-layered narrative: layer one is a short story by classic sci-fi writer Robert Sheckley, a critique of technological consumerism and consumer debt, and layer two, where all of the interactivity takes place, is a discussion of the story by a couple of audience members, breaking in periodically in a different font. Their conversation contains occasional blanks for the player to fill in with interpretive words: “Don’t you get the feeling that Carrin is ______ about Miller?”

Now, the system makes it clear that it’s paying attention to how you fill in the blanks. An introductory section is very clearly responsive, asking yes/no questions, and later parts bring up words that you typed in previously. Nonetheless, it felt mostly inconsequential. Obviously the course of the pre-existing Sheckley story isn’t going to vary with your choices, but even the discussion seemed like it was just producing the same output regardless of what I typed a lot of the time, just swapping in the words I typed. I suspect that it really was varying the output, but not being very obvious about it. I could accept this as what Emily Short calls “reflective choices”, prompting the player for a reaction just to provoke one, but a lot of the prompts seemed to be angling for specific responses, like a middle school English test. Consider the passage:

Vesper: He gossips about everyone in town. Company’s code. Yeah right!
You know he uses that line at every house on that block.

Harris: You don’t think Pathis is being ______?

How do you fill that with anything other than “honest”? And if you’re giving me a purely reflective choice and making it clear what you want me to choose, I start to feel like my interaction isn’t serving any purpose at all.

Also, the ending is less than satisfying — so much so that I thought at first that I had hit a bug and the game had ended prematurely. Admittedly, the inner story’s ending is unsatisfying by design — it’s depicting an unsatisfying world! — but the game gives the last word to Sheckley, not the audience, and I would have at least expected the commentary track to have a summing-up, giving the fictional audience’s thoughts after seeing the whole thing.

But I can kind of see a thematic justification. Two-layered stories always implicitly ask “What is the relationship between the layers? Why is this particular story told in conjunction with that particular story, and how do they resonate?”, and once you’ve posed that question outright, the obvious answer is that the Sheckley story is about a society that’s technologically advanced but constraining, pressuring people to conform while plying them luxuries that don’t really satisfy them, and then the interactivity is similarly technological but constrained, unsatisfying, and pressuring. I don’t really buy this, though, because you have to ignore so much about the story to make it work. The inner story’s central ideas are luxury and debt, and the outer story doesn’t reflect that at all.

Still, I give it kudos just for experimenting with form. That’s always interesting to see, no matter what the result.

I recently learned the term “database fiction”. It refers to works like Her Story or Portal (Brad Fregger, 1986) where the player’s main activity is querying a database for more story.

Kondiac is the smallest work of database fiction I’ve ever seen, consisting of “about 9 different pages” according to the author. Each page is an image of a document, mostly containing text with names or other notable keywords you can enter into a search bar to pull up more documents. (I’d be very surprised if it’s actually parsing the input at all.) There’s no definitive ending; you just stop looking for keywords when you’ve satisfied yourself that you’ve learned an Alaskan town’s grisly little secret, which, if you’re genre-savvy enough, could happen the moment you enter the game’s title and see the words “butcher shop”.

And that’s honestly a bit of a problem. The game starts with just a prompt and a photo of a building, with no instruction or orientation to let you know what kind of game it is or what you’re supposed to be doing. By the time you know what you’re looking for, the game is over. Can we really call it a mystery when answers precede questions in the audience’s mind?