Wander: The Surviving Games

Wander: The Surviving Games

So, last time I discussed Wander as a whole; this time I’m going to talk about the surviving Wander games. (And then next time I’ll go into the Wander language itself.) I’m not going to do a play-by-play of the games; if you want that you can go to Renga in Blue or CRPG Adventures, where Jason Dyer and Nathan P. Mahney, respectively, provide playthroughs. I’m just going to briefly summarize the games to provide a frame of reference in case I want to refer to them later.


Yes, that’s the name of the game files— “a3”. It’s short, as it turns out, for Aldebaran III; in this game the player takes the part of Jame Retief, hero of the eponymous series of science fiction books by Keith Laumer, as he visits the world of Aldebaran III to “avert an uprising of Terran nationals expected at the end of April”. Yes, the game does keep track of the date and time—each turn takes one hour (which seems a bit excessive for an action like picking up a credit card), and you can type “date” or “time” to see the current, well, date and time (which also takes an hour). However, it does this to little real effect… you’re told that “Tensions seem at the breaking point” after the 27th, but then after the 30th it goes right on to April 31, and then April 32, and then April 33, and so on… perhaps Aldebaran III has very long months.

In any case, this is apparently the oldest of the surviving Wander games… or at least the first one started, although the extant version shows clear signs of having gone through multiple stages of expansion. This is most obvious from the location numbers in the code, which are often not consecutive; there are also many locations in the code that were never implemented beyond a name, and one location that explicitly says “Under Construction”. These aren’t the only indications of a3’s antecedence; it’s this game that has code snippets from it used as examples in the Wander documentation (though they’re slightly but significantly different from the code in the game itself, suggesting that they came from an earlier version), and it’s a3 that gets homaged by the later Wander game Library (about which more below).

Peter Langston told me in an e-mail that he did indeed add code and locations to a3 over time, using it as a testbed for new language features, and that he didn’t consider it to really compose a coherent game. The former more or less confirms what I had already surmised from the code, but as to the latter I think he undersells the game somewhat. As it currently stands, no, maybe a3 isn’t really much of a coherent game; I’m not sure what a stereotypical vampire (complete with accent) is doing on Aldebaran III, and the weird reskinning of the old tiger, goat, and cabbage chestnut seems a bit strained. But I think it may have been more of a coherent (albeit small) game in its earlier stages, before he started using it as his testing sandbox. There’s one point (at location 33) where Retief has a confrontation with a nefarious native and then meets an important contact.  I suspect this was originally the end of the game: this seems like it would have been a decent ending; everything up until that point is more or less coherent; and everything afterward seems a bit tacked on. (Well, some of the material before it was tacked on, too; there is, for instance, a customs office very near the beginning of the game that appears from the location numbering to have been a much later addition; it’s a rather code-heavy location, so it makes sense that it could have been used to test some language features.)

In any case, as it stands the game has some odd parts, but it’s not actually that bad. One or two of the puzzles are rather unfair (it’s necessary to ask a bartender for something to receive a crucial clue, but there’s nothing to hint beforehand that this is the case), but maybe no more so than puzzles in other old (or even many new) adventure games… and in other ways it’s actually surprisingly forgiving; there are several vital objects that can be acquired in more than one way.

Peter Langston had hoped to one day approach Keith Laumer himself about writing a script for a Wander game. Of course, this never happened, and now never can happen, unless Laumer can somehow be coaxed to create from beyond the grave. This provides an interesting glimpse at a road not taken… but I’ll have more to say about that in the next post.

You can read playthroughs of this game at Renga in Blue and at CRPG Adventures.


Also written by Peter Langston, this is less a game than it is a sort of a proof of concept in using Wander as an educational tool—in this case, to teach binary. (The name, presumably, is short for “Tutorial”.)  While this is a very short and simple implementation, I think it’s interesting that Langston was ahead of his time in anticipating using text adventure games (not of course that they were called that then) for educational purposes.

Neither Jason Dyer nor Nathan P. Mahney posted a playthrough of this “game”, but that’s not surprising; there really isn’t much to it.


The only one of the known surviving Wander games not to be written by Peter Langston, this game was written by Nat Howard using Langston’s Wander language. It’s mostly a mishmash of in-jokes and science fiction literary references, and contains callbacks not just to a3 (the player can visit Aldebaran III’s spaceport bar), but also Adventure itself—which, of course, shows that this game must have been written after Adventure. (Amusingly, the Adventure reference is a jab at the game’s dragon puzzle, which I guess even back then must have been somewhat infamous.) Other than the example it provides of a Wander game by another author, maybe the most interesting thing about this game is the hostile gnome in the spiral staircase chamber, who is I’d say the most complex NPC in any surviving Wander game, though that’s not saying a lot (and actually, I suppose the a3 bartender comes close). Unfortunately, the gnome is seriously bugged (I’d guess the code for the gnome was added to and made more complex over time, and that the surviving version hadn’t been tested), which is a pity, because if it worked as intended (or as I surmise it was intended looking over the code), it would actually be kind of a neat puzzle. You can kill the gnome the obvious way with a weapon, but this isn’t guaranteed to work; there’s some randomization in the code, and the gnome could kill you first (not unlike the troll and thief in Zork, which I’d guess the gnome was probably inspired by… again, Wander itself was written in 1974, but Library apparently came much later, and could very well postdate the mainframe Zork/Dungeon). But there’s another, less obvious way to kill the gnome that is guaranteed to work, and lets you get rid of him at no risk to yourself. Or, again, there would be such a way, if the code worked as intended…

You can read a playthrough of this game at Renga in Blue. (Nathan P. Mahney said he was going to play Library next at the end of his review of King Tut’s Tomb, but as of the time of this writing he hasn’t yet done so, or at least he hasn’t blogged about it.)


With its fantasy setting, its conventional puzzles, and its more or less coherent plot, Castle is the Wander game that may look the most familiar to modern eyes, and that may most resemble what we’d consider a typical adventure game. But that’s probably because it took inspiration from Adventure and other games that went on to define the genre. As I mentioned in the last post, Jason Dyer dates Castle to 1974, but that’s clearly not correct. For one thing, Castle contains references to Adventure, most notably the implementation of the word “xyzzy” (which the game calls “an old, worn-out magic word.”). If that were all, then one might argue that Castle was originally written before Adventure but this word was added later, but, aside from the whole feel of the game seeming to be Adventure-inspired (which I admit may be somewhat subjective), there’s also an Adventure-style maze in the game—it’s not impossible Langston came up with the idea for the maze independently, but it doesn’t seem likely. Library and a3 also have mazes, but, well, it’s already obvious from other indications that Library postdated Adventure, and in the case of a3 the maze appears in one of the parts that’s apparently a late addition to the game, so all it suggests is that that section of the game was added after Adventure. (There’s a sort of a maze earlier in a3, a series of streets in which you can only go left or right, but it’s not a maze in the Adventure sense, and in fact it’s kind of a clever little puzzle… or it would be clever if it weren’t for the fact that you don’t get the necessary hint for the puzzle until you’ve made seven wrong turns, and the maze is small enough that you’re virtually certain to solve it by trial and error before that happens (as in fact Nathan P. Mahney did in his playthrough).) In the case of Castle, though, it really can’t be convincingly argued that the maze could have been added later: the location numbers are consecutive and there’s a clear ending at the highest location number, so it seems quite unlikely that locations were added in the middle.

But there’s an even bigger giveaway of Castle’s being a relatively late Wander game, and it has to do with a bug that Jason Dyer experienced in his playthrough… but I’ll explain that in a later post.

[EDIT (about 45 minutes after posting): Hm… you know what?  I just noticed something that makes me rethink that… I’m not sure that bug shows Castle to be such a late game after all.  (I still think it was a post-Adventure game and certainly post-a3, but maybe not as late as I thought.)  There are some things I need to try out with the code…  I’ll still explain this in a later post, but now it may be an explanation about what I had been thinking and why I was wrong about it.]

You can read playthroughs of this game at Renga in Blue and at CRPG Adventures.

So, anyway, those are the four surviving Wander games. Or at least the four surviving Wander games that have been found so far… if there are any other Wander games still out there in an archive somewhere, they have yet to resurface. Still, they’re enough to get a feel for the system. But of course we can get an even better feel for the system by diving into the documentation and the code… and that’s what we’re going to do next time. And then, of course, we’ll start writing a new game with it…

Spread the love

Leave a Reply

Your email address will not be published. Required fields are marked *