Playing Unholy Grail puts me in mind of the old saw about the glass being half-full or half-empty. For each positive I can think of, a counterbalancing negative also comes to mind. While the prose creates sharp, clear, atmospheric images, it is also burdened with numerous grammar and spelling errors. While the game had an inventive plot, this same plot was punctuated with moments of tediousness, implausibility, and pure frustration. And while Grail is orders of magnitude better than Allen’s 1996 entry The Curse of Eldor, it still fails to realize both its own potential and that of its author.
Allen has accomplished a noteworthy programming achievement: he has written his own IF engine, one which mimics much of the important functionality of the current front-runners Inform and TADS. Unfortunately, it still doesn’t perform at the levels of either of these popular “standard” IF engines, and suffers greatly by comparison. Again, it’s a yin and yang situation: a quality engine is written from scratch, but it’s still a poor competitor to the dominant systems, marred by problems ranging from the complex (tortured disambiguation) to the amazingly simple (an inexplicably arbitrary pathname in the CONFIG file.)
Still, Unholy Grail is the first competition game I’ve played, and it’s not an altogether inauspicious start. For one thing, it represents remarkable progress on the part of the author. [No, the comparison to non-competition games is not affecting my overall rating — I simply mention it in this review for the sake of completeness.] Unholy Grail is not the fulfillment of Stuart Allen’s promise, but it marks him as one to watch. With the improvements he’s already made to his JACL engine it seems entirely plausible that it could one day match the quality of the current state of the art. Also, this game is one of the few conceptually complete pieces of IF I’ve seen in the “thriller” genre, a field which is in many ways well-suited to IF, but whose only significant representative has been Border Zone, a quality game but one whose gimmick of real-time has often overshadowed discussion of its generic groundbreaking. Unholy Grail was uneven; some things were really very good, other things really not very good at all. I hope it’s a marker of better things to come.
Prose: I found the prose in Unholy Grail fairly difficult to read. Sentences seemed to string endlessly, clause following clause until I thought perhaps the author had asked Henry James to ghost-write. However, I also think that the lack of a status line and room name threw me out of my ingrained IF reading habits, the disorientation of which probably contributed to my difficulty in following the author’s long narrative strands. Or it could just be my own denseness — that’s always a possibility. Despite the game’s verbosity, though, strong images floated up to me out of the sea of words. I have a very distinct picture in my mind of the swivel chair and radar screen in the control room, of the battered hut whose floorboards parted to show the ground below, and of the elegant, elaborate hotel. The author clearly had done his homework, and was able to create a very convincing picture of the character’s environment. I just had to read some of the sentences a few times before I felt sure I knew what they were saying.
Plot: The most ringing endorsement of the plot I can give is this: after the two-hour judging period had expired, and I was only 75% through with the game, I spent another half-hour on it because I needed to know how it ended. I found the plot difficult to get into at first (see Puzzles), and needed to refer often to the science encyclopedia so I could have a basic clue of what the hell the game was talking about, but once I understood, I was inexorably drawn in by the skillfully dropped hints and slowly unfolding drama. On the other hand (and there’s always another hand when it comes to Unholy Grail), I found some things in the plot pretty difficult to believe. Small points like the layout of the complex were jarring: would the military really set up a female officer’s room so that her only access to it is through a civilian male’s room, and have her share a bathroom with him as well? Also, some larger points (such as the Rotenone) seemed only to serve as red herrings, but created major implausibilities in the plot: if I’ve determined that Rotenone is causing the fish deaths, how can it be true that they’re being caused by something which in fact behaves entirely differently? For that matter, if my basic science encyclopedia tells me that Rotenone causes fish to drown, why do I blame it for cancer?
Puzzles: For the first hour I played the game, I was absolutely stumped. Finally, I resorted to the hint system and learned that because an extra-long sentence in the room description of the lab, I had neglected to examine the lab bench as closely as I ought. Once I found the global positioner, I was off and running. Consequently, I struggled with this game a lot more than its puzzles may have merited. Most of the puzzles were fairly easy, when they didn’t involve guessing the verb (Can’t turn the drum. Can’t move the drum. Can’t push the drum. Can’t pull the drum. Can’t look under the drum. Oh, look behind the drum!), and some were quite satisfying (especially the filing cabinet.) However, one puzzle was amazingly tedious — it basically involved typing “n” 20 times and “w” 20 times, then doing the opposite. Here’s where a “swim to” verb would have been much appreciated!
Technical (writing): In addition to the stylistic factors I mentioned in “Prose”, Unholy Grail was also plagued with grammar and spelling errors. Certainly there was some attention to proofreading, but one or two more passes were needed.
Technical (coding): Unfortunately this is where Grail stumbles the most. JACL does a good job of imitating mainstream systems (especially Inform) in many ways, but in other crucial areas it falls critically short. For example, the system allows only one saved game at a time, and it lacks an “oops” verb. Also, its disambiguation is weak, a fact which caused a great deal of frustration for me as my reasonable answers to its reasonable questions kept getting the response “The sentence you typed was incomplete.” The system also overuses Graham Nelson’s famous “You can’t see any such thing,” applying it to sentences whose nouns are examinable and manipulable in other contexts. In addition to these general systemic problems, Grail itself had a number of particular bugs which I’ve reported to the author in a separate email.
OVERALL — A 7.O