Castle of the Red Prince

In C.E.J. Pacian's Castle of the Red Prince, a student of "the arcane" suffers from nightmares about the land of Amaranth and its tyrannical Red Prince, and sets out to end the Prince's oppression by killing him. It's unclear whether the whole game takes place in a dream, but there is something dreamlike about the setting – broadly fantasy, but with horror-like features common even outside the Prince's territory, and occasional intrusions from modern technology and real-world place names. Dreaming is also something of a theme, especially in an ingenious hint system that allows you to SLEEP and watch "your dream self" carrying out actions you can try after awakening.

The game's main innovation is its novel approach to location modelling. The world is still divided into locations (at least from the player's perspective – I can't comment on the programming), but all characters and objects that the player has discovered is in scope simultaneously. There's no compass, and no need to use movement commands in exploration; ENTER INN is equivalent to EXAMINE INN. Pacian appears to have avoided disambiguation problems by ensuring no two objects share a name, and it's impressive that this never leads to the prose becoming unnatural or obvious synonyms going unrecognised by the parser.

I think the experiment succeeds in at least two ways. First, it suggests that the protagonist's movements are extraneous to the real activity in much parser-based IF. (Several times in Castle I discovered an item and realised it could be used to solve a problem in a location I'd already visited; the fact that I could just use the item without retracing my steps brought my interaction with the program much closer to the process of solving the problem mentally.) Secondly, it demonstrates that making all of game-space available for interaction conflicts with the player's expectations about game-time. The time of day appears in the title bar and advances every two actions, regardless of how much movement those actions would realistically require from the protagonist. The effect works here because it's suggestive of the sudden changes of place we experience in dreams, but a naturalistic game with an internal clock couldn't follow the same approach.

Castle seems to me less successful as a game than as an experiment. It's short (it would probably have 15-20 rooms if implemented conventionally, but most are empty or contain one thing of importance), and the combination of brevity and eccentricity of setting kept me from feeling immersed. Puzzles tend to adhere to well-worn IF tropes ((Spoiler - click to show)lighting a dark area, attacking an enemy with the appropriate weapon), perhaps as a way of suggesting that the game's approach to location-modelling is applicable outside experimental works. I didn't find the Red Prince a dramatically effective antagonist, with his lack of concern at the protagonist's attempts to defeat him. He isn't even responsible for either of the two losing endings that I found.

Despite the monsters I encountered, my journey to Amaranth was more like a brief dream than a nightmare that would haunt me for days. However, a dream can be memorable for one unusual element, and Castle's success in dispensing with the usual IF approach to location is easily enough to make it worth playing.