Yet another one of Zachtronics Industries' so-called "games for engineers". This one is a full-on indie game that is to be sold and installed and stuff, as opposed to the cute little executables and flash games that preceded it.
Not that the Alchemical Codex and its relatives were in any way weak, or not worthy, by virtue of not being "full-fledged games": I myself have yet to finish the final level of Codex, not to mention begin tearing through The Magnum Opus Challenge. I just think that, being vidya with a price tag, the bar is set a little higher for SpaceChem than for something that costs no more than the bandwidth needed to download it.
The concept had looked interesting: it was pretty much Codex, except there's a completely different way of controlling the bonding agents. Waldos, they were here termed; instead of Codex's little programmable arms, the waldos followed prespecified tracks. It reminded me greatly of my esolang stint, specifically the "2D" languages.
So, I pounced upon a Humble Bundle (HB for Android #3), and promptly installed the game. I then proceeded to immerse myself in this little world—these little worlds?—of waldos and molecular reactors.
Before I begin the endless stream of praise I am so apt to spew about things I adore, a short description of what is going on and how you play. You are so-and-so, a budding Reaction Engineer, who, after persuading your parents to let you go, travel to the moon Something II to work for SpaceChem. You begin designing reactors, which are grids 10 long by 8 high, featuring two waldos, two input areas, two output areas, and various reaction engineering equipment like bonding stations, element sensors, and more. You direct the waldos along a user-defined track, with waldo commands appearing as nodes along it; they prompt the waldos to perform the specified action. Your goal is to transform the given input into the requested output, preferably as efficiently as possible. Not that anyone will cry about it if you spend thousands upon thousands of cycles (in-game time-steps) doing simple tasks, but there's a comparison chart at the end of every level and it feels just a little bad to find yourself short of the average. "There's nothing wrong with second place," my ass.
An interesting and absurdly useful waldo command is SYNC, which will stop the first waldo to reach it, and only lets it go when the second waldo reaches its own sync. This becomes especially useful when you figure out that two molecules can't occupy the same space (obviously), and that some transformations are sufficiently complex that the waldos need to "work together". (I have this one design requiring no less than six syncs in a row, because I need to turn a random feed of elements into a fixed one. It's beautiful in the most disgusting way.) Soon, you move from reactor design to making entire supply chains, with cute little windows into the reactors when the full simulation is running. And god forbid you have a subtle error somewhere in your mechanism, for it brings the beautiful cacophony of molecular rearrangement to a grinding halt, and you feel the sudden need to correct it as quickly as you can to view your creation in motion again.
As you progress through the planets and the challenges, the story unfolds, going from tame to foreboding to nearly scary. Tying in with this, we get the Defense Missions: you need to use your reactor-optimizing skills to fuel an attack on rogue elements. For some, the story could be the driving force behind completing each mission; for me, it's just another enjoyable factor in an intellectually stimulating game.
So, how well does it work? Exceedingly. The unique blend of micro- and macroplanning gives a real sense of accomplishment when you've finally turned your plan into a working supply chain, and the associated difficulty curve is dampened by more than a little, due to the way that the game eases you into the duties of a workaday reaction engineer. The fact that the waldos operate on tracks gives a very interesting flavour to the puzzle: you have to modify your strategies to work with the constantly moving nature of the waldos. The replay value isn't huge for anyone that isn't either a die-hard puzzle/strategy fan or an aforementioned engineer, as finding optimizations, especially the big ones, or ones in more difficult missions, is an almost Herculean task. Getting into that top tier on the performance charts will take literally every ounce of your ingenuity. You do get to pride yourself on being that much better than the average when you succeed, however, which makes it worthwhile.
I won't lie, I've only barely scratched the surface of SpaceChem; the lowly Reaction Engineer II tubular has many more molecules to synthesize, many more supply chains to complete. But SpaceChem has enthralled me the way few games do: enough that I don't even have to be good enough to finish, to realize it was definitely worth my time. And will continue to be, whenever I get ambitious enough to try again.
In short, on a scale from 'excruciating' to 'astonishing', SpaceChem ranks just a little above 'spellbinding'. Solid gameplay, fun, interesting, and definitely a challenge. If I somehow needed to goad my resigned, resenting parents into purchasing another copy of this game for me, I would be willing to lie profusely about how much homework I was assigned in school that morning.