Cannibalise the corpses!Publication date: 25 June 2011
Originally published 2010 in Atomic: Maximum Power Computing
Last modified 03-Dec-2011.
"Landfill miners" dig up the huge garbage-piles that modern society creates, and strip out the valuable stuff, like metals.
Landfill mining is becoming a big business. Just as biodiesel has made used frying oil a highly salable commodity, bulk garbage now has considerable value.
This makes perfect sense when you think about materials like aluminium, which is sometimes referred to as "solidified electricity" thanks to its incredibly energy-intensive smelting process.
We dig up bauxite, we process it into alumina at considerable expense, we convert the alumina into metallic aluminium at staggering expense, we super-purify the result and roll it out into amazingly even sheet... then we wrap some chicken in the foil for a couple of days, or drink some beer out of a container made of it, and then throw the aluminium away.
It takes a lot of energy to make a modern top-shelf video game, too. Teams of coders and animators and texture artists and 3D modelers and motion-capturers and motion-capturees and sound people and writers and who knows what else spend years on end making a big-ticket game.
And then their Great New Game is launched...
...and, sometimes, it's a miserable failure.
Several recent big-name game-deaths were pretty predictable, on account of how making a new MMO that competes with WoW is like making a new auction site that competes with eBay. Even some of those games, though, thoroughly deserved to fail - APB is the perfect recent example.
Or it was.
Here, this online version of this column diverges from the one that ran six months ago in Atomic magazine, because if a dead game is a corpse, then APB must be a zombie. APB launched as a pay-to-play massively-multiplayer game, died a well-deserved death, then a few months later was raised from the grave by another company for one and a half million quid. It will shortly, as I write this in mid-2011, be relaunched in a free-to-play incarnation. And perhaps it'll be good, now.
(See also Hellgate: London, another heavily-hyped massively-multiplayer rocket that plunged into the ocean almost immediately after launch. It, too, is now being necromantically reanimated as a free-to-play game.)
APB can still serve as an example if you ignore its re-invention, though. What did everybody say about APB when it first came out? "Excellent character creator, shame about the rest," right?
So, presuming the game had been left to rot and not sold on, why not take that character creator, and stick it onto the front of a better game?
Software isn't Lego, of course. If Rockstar decided to make all of the data for GTA IV's miniature NYC a free download, that wouldn't make it easy to use that data in another game. And voice acting is a major expense for a lot of modern games, but you're obviously not going to be able to reuse much of that without creating some abomination like Trail of the Pink Panther.
But there are plenty of components from failed new games and remaindered old games that could be given new life.
There may even be something salvageable in true legendary computer-game train-wrecks, the ones that somehow manage to get review scores below 50%.
Some of these disasters are so bad because they're the extremely pre-alpha bare skeleton of a game, like the infamous Big Rigs: Over the Road Racing; there whole problem with that sort of game is that there's no meat on its bones at all. Other failures are like the Terminator 3 and Terminator 4 games and contain, in the great tradition of crappy movie-tie-in games, little content worth bothering with except for things that are the prized intellectual property of a large corporation.
But then there's the more common kind of failed game, like the forgettable God-Of-War-in-Iran game Garshasp: The Monster Slayer. There's stuff in that game that looks and sounds pretty decent; regrettably there's also quite a lot of other stuff, but nothing in it belongs to Paramount Pictures.
There may even be rescuable material that no player has ever seen. It's normal for Development Hell game projects to have a lot of half-working stuff left on the cutting-room floor. The "cut content" will often still actually be physically present in the game files, like "Hot Coffee" or the numerous loose ends in the last two Fallout games. Cut content is usually just things like weapons you can only access via console commands, but sometimes you get confusing stuff like NPCs who look significant, but whose associated quest was cut.
Sometimes the game never makes it to market at all. Look at "Project LMNO", one of the three games being produced by Steven Spielberg. That project can reasonably be presumed to have absorbed a fair bit of human effort over the five years between inception and death; all gone, now.
There are two reasons why recycling of parts of failed games doesn't happen all over the place already.
One, the abovementioned compatibility problems. There's a reason why it's such a big deal when a game that's been in development for a couple of years changes to a new engine (coughDuke Nukemcough). Some resources are easy to convert, but even straightforward things like textures and sound effects can be surprisingly problematic, and AI and event scripting and model optimisation can be much worse.
(Sound effects are also, often, not a game-maker's property to sell. They're often bought from stock libraries, which is why you keep hearing strangely similar explosions and gunshots and sliding-metal-door noises in a wide range of games, movies and TV shows.)
All this aside, though, using resources from one game in an entirely different one is, often, perfectly practicable, and can be a huge time-saver.
It may not be financially sensible, though, because of the second problem: Intellectual-property death-grips.
But prying a 13-year-old game's awesome music loose from whoever now owns it may well be more difficult than licensing a score from a "production music" outfit, or licensing some classical music from an Eastern European orchestra, or indeed making a whole new score from scratch. Especially now that synthetic orchestras sound pretty much perfect.
(One notable example of music re-use, by the way, is the abovementioned "modern" Fallout games. They recycle various Inon Zur tracks from the earlier games, and it works beautifully.)
Music re-use is already common in the motion-picture world. Especially the motion-picture-trailer world. Most trailers come out before the movie they're promoting has a soundtrack at all, but even when the real soundtrack's already in the can, there's a surprisingly short list of Approved Trailer Soundtracks. Apparently it's acceptable for four dorks in the audience to say "Hey, that's the Stargate theme again!", if that music is suspected to make everyone else 7% more likely to buy tickets.
Failed games tend to turn into impassable legal mud-wallows. (After the Tabula Rasa debacle, Richard Garriott sued NCSoft for 24 million dollars. And won.) The whole developer and/or publisher may go broke, and then all of the intellectual property is in the care of an administrator, who probably isn't allowed to sell bits of dead game piecemeal, and may put an outrageous price on everything anyway.
I think the main problem may just be that game-makers aren't set up to do this sort of thing. They may give away their old games as free downloads, but dammit, Jim, we're programmers, not stock-photographers.
Given these obstacles, I wouldn't be surprised if copyright-free game resources turn out to be more workable than recycled content. I like the idea of there at some point being "open-source landscapes" for people to set games in.
The scale of the real world is wrong for most games. Real-world doors, for instance, are rather too small for many game avatars to get through cleanly, since the coarseness of your control over your character in a 3D game gives you more in common with a vision-impaired Dalek than an actual human. And the real New York is a lot bigger than its depiction in any game to date, because otherwise it becomes difficult to find things and/or get to them in a non-boring amount of time.
But I wouldn't be at all surprised if something like the "seam-carving" resizing technique that's now hit the mainstream in Photoshop could be applied to real-world terrain and/or city data. That'd make it relatively easy to whip up a detailed map that retains the flavour of a real-world location, but is small enough for, say, another Fallout game. (The Fallout: New Vegas map is only about three kilometres on a side.)
A crowdsourced city map constructed by a thousand Sketchup users could definitely work, though, and could be used in several kinds of game. I don't think people would be very concerned about a military shooter and a superhero game using the same map, as long as Third Platoon don't find themselves being tossed into Pennsylvania by the Incredible Hulk.
I still think failed games taking good content to the grave with them is a terrible waste, though.
Many games are bland for the same reason that many blockbuster movies are empty and predictable: They cost so much to make that the creators only make products that're similar to previous successful products. Reduce the price with a bit of recycling from a game that might be from a whole different genre, and you can afford to be more adventurous.
We should probably leave Daikatana in the landfill, though.