How the Liberated Pixel Cup changed gaming on Linux
17th Feb 2013 | 08:00
Inside the competition for the best free Linux games
Humans may be the most creative species on the planet, but we spend a lot of time doing tedious things.
Look at the internet: it's a revolutionary and disruptive technology, with the potential to change education, governments and scientific research, yet most people use it to post comments on YouTube videos of mobile phones being unboxed.
Here in the free software world, we're familiar with the collaboration opportunities that the internet brings, and many great applications have been developed by teams of programmers around the globe.
But desktop and server applications aren't the most exciting things in the world. Wouldn't it be great if all the untapped creativity of humans on the internet could be funnelled into making artistic things, such as video games?
If you've followed the Hot Games section of HotPicks over the years, you'll have discovered a few notable projects involving multiple contributors; the vast majority, however, are one-man efforts. And part of the problem is this: a developer may come up with a brilliant idea for a game, but lack the artistic flair to make it look good.
Or, conversely, a budding artist may love designing graphics for his/her ideal RPG, yet doesn't possess the technical nous to make it a playable reality.
We've seen loads of promising games let down by inadequate artwork, and heard from designers who'd love to contribute to an open source game, but don't know where to start or who to speak to.
This is where the Liberated Pixel Cup comes in. The LPC project aims to build an extensive resource of free (as in freedom) video game artwork for programmers to use. But it's more ambitious than a generic sharing system: using development phases, targets, deadlines and ratings, the LPC team have ensured that games get made, and it doesn't become a hypothetical exercise.
There are now a bunch of great-looking, professional-level games for Linux - and they're totally free software.
An epic contest for gaming freedom
OpenGameArt is a busy website that has been running since 2009, serving as a repository of user-submitted game graphics and music, along with a forum where programmers can request specific types of artwork. Earlier this year, Bart Kelsey, who runs the site, teamed up with Chris Webber of the Creative Commons, and together they fleshed out the LPC concept.
They wanted to create a fast-paced competition, whereby artists around the internet would generate as much graphic content as possible in a one-month period, and then in the following month developers would write games based on that artwork.
A core principle of the LPC is that all content is free as in freedom - that is, the graphics must be submitted under the Creative Commons BY-SA licence, which lets anyone share and modify the resources, as long as they provide attribution to the original author and share only under equally permissive licences.
Code, meanwhile, had to be under the terms of the GPL. With these requirements, the Free Software Foundation was happy to get on board and lend its support as well.
Kelsey and Webber wanted to award cash prizes for the best entries, so they set up a fundraiser; at the time of writing, it had raked in an impressive $12,256, slightly exceeding the original goal.
Then, the next step was to set up a style guide for the artwork. What sort of games should the LPC focus on? It would have been easy to let anything go, and open up the contest to all sorts of genres and artistic styles, but then it would be incredibly difficult to fit content from different authors together.
Imagine if you were trying to write a game using multiple graphics sources, and they all had different sizes, perspectives and lighting effects - it'd be a horrible mess.
So the LPC settled on 2D pixel-art graphics from a top-down view, but with a bit of perspective (60 degrees), so that you can see the front of objects. Games will be built from 32x32 pixel tile sets, with an option of smaller 16x16 sub-tiles where necessary.
Lighting should come from above, and possibly slightly to the left, so that everything doesn't look too symmetrical and samey. For characters, the dimensions should be 48x64 pixels.
This might all seem extremely specific, and force developers into producing top-down Zelda or Final Fantasy-like RPGs, but the LPC rules don't limit developers to one particular genre - they're free-to-code, side-scrolling shoot-em ups; turn-based strategy games; or anything else that takes their fancy.
For those interested in the RPG genre, however, there's a set of base tiles and 'naked' characters (see the screenshot) to act as a starting point. This makes it easier for new artists to adopt colour palettes and visual styles. You can find the style guide, base tiles and a simple HTML5 demo of a point-and-click game here.
It's just a phase
With everything in place, the contest kicked off on 1 June with phase one: artwork. This ran until the end of the month, with artists uploading their work to the OpenGameArt website.
A total of 31 artists contributed to the competition, and you can find their work at http://opengameart.org/lpc-art-entries - as you can see, it's a mixed bag of static tiles, animated tiles, characters and collectable items.
Phase two kicked off on 1 July, again running for a month, and focused on coding. An important requirement for the code submissions was that they should run on completely free software (ie, a GNU/Linux system that has no proprietary bits installed). All of the free content and code in the world wouldn't mean much if the end result could only run on Windows…
At the start of August, the judging process began. Kelsey and Webber led the panel, asking them to rate the submissions based on their overall quality and adherence to the style guide. Unfortunately, due to various complications (see the interview over the page), the judging progress has been running rather slowly, and at the time of writing only the art submissions have been rated.
The top winner, who created an attractive set of farm-based tiles (fancy writing an open source version of Harvest Moon?), scooped up a $1,500 prize. By the time you read this, the code judging phase may be complete.
Our pick of the best
You can find the full list of code submissions for the Liberated Pixel Cup at http://opengameart.org/lpc-code-entries - there are 48 to try out, and most of them can be played from inside a web browser, so you don't need to fiddle around with compiling stuff and grabbing dependencies.
Due to the one-month coding-time limitation, obviously none of these games are vast Zelda-like adventures, but they are lots of fun nonetheless. Here, we've selected some of our personal favourites…
Volleyball meets zombie attacks in this bizarre shoot-em-up. You're challenged to play the sport, using the WASD key group to move your player around, and Space to jump up and knock the ball back with your head. But while this is going on, hoards of monsters approach you from all angles, so you have to shoot them using the arrow keys. If the monsters touch you, your health bar depletes. This game works especially well with two players sharing the keyboard: one focuses on the volleyball, and the other on wasting zombies.
You might be expecting something action-filled from the title, but Castle Defense is all about using your brain. Essentially, it's a maths-based game, whereby skeletons are trying to break through your castle's towers, and you have some archers at your disposal. By solving the maths questions that appear (just type in the answers), your archers fire arrows horizontally along the screen; you start off with two archers responding to addition exercises, and on the next level there are three with subtraction questions, and so forth. The music is great, too.
Green blobs of slime crawl out of a hole, meandering around the screen until they reach the princess on the other side. Using bits of fence, bait and magic spells, your job is to keep the blobs at bay. It's entertainingly silly and surprisingly tough - you're limited by the amount of fence sections you can build, and you can use magic spells only once every few seconds (due to a mana meter that depletes and restores itself). The developer has included a level design mode, adding considerable longevity to the game. It's a simple concept, executed well, and has that special 'one more go' touch.
We're into classic top-down action-RPG territory with this game. It's beautifully presented, with a touching intro story and gorgeous pixel art, although the music is painfully repetitive. It's quite small right now - there are just a handful of items to collect and things to swap with villagers, and clearly the plot needs fleshing out more. But you can talk to people, collect coins, fight snakes with a giant mallet and explore an island - if anyone wants to expand it with more content, that'd make us very happy bunnies.
Imagine Reversi (aka Othello), the classic disc-flipping board game, and then make it real-time. In Terramancers, you control a character walking around a grid, and each tile you step on becomes yours. If you step on a tile that has a direct line of sight to another you own, the whole line is yours. But the computer is also racing to dominate space; once all tiles are owned, whoever took the most wins. Run it with java -jar Terramancers.jar, and you'll have to rename resources/Characters/Professor.png to its lower-case equivalent as well.
This is our favourite of the lot. It's also a top-down action RPG, with a great little intro, lots of polish and plenty to see and explore. The battle system is somewhat unusual: when you come across an enemy, you use the arrow keys to switch into one of four positions around it, from which you can attack. Another neat feature is the Desolation, an alternative Dark World-like version of the main map. You can switch to it with the R key, which is useful for getting past things that are blocking your way. Really, really promising.
The man at the helm
Chris Webber started the LPC, with help from OpenGameArt and the FSF. We asked him about the ups and downs of running the contest, how the judging phase is going, and what next year's LPC could look like…
Linux Format: It looks as if the fundraising attempt has exceeded its target - how is the money going to be used?
Chris Webber: I posted a blog post on this, detailing how the finances worked out. But the short answer is the money went to artists to pay for the base assets and style guide that was the basis of the Liberated Pixel Cup; and to prizes, except for 10% that went to the Free Software Foundation for handling the overhead. None of the money went back to us. Bart and I both put in over $1,000 each. But it's a cause that we, like the many others who donated, believe in.
LXF: What takes place in the judging process, and when do you expect it to be finished?
CW: There's two phases to judging, just as there's two phases to the games: judging art, and judging code. The art judging just wrapped up very recently, but the code judging has yet to be announced.
Unfortunately, as the art judging was, it's really quite delayed. This is due to a couple of reasons… first and foremost, Bart Kelsey and I are the primary people driving things. Bart had some personal matters to deal with, and I simultaneously quit Creative Commons to focus on MediaGoblin, was travelling and technically homeless for a couple of weeks right after the submission part of the LPC wrapped up (we were moving between states), and starting to push forward MediaGoblin as a full-time thing. This lead to a lot of delays in even starting things, which was frustrating for everyone.
The other problem was the kind of problem you want to have. We didn't expect so many entries, and so many great entries at that! I know it's the kind of thing you can't complain about: we got so many amazing entries, that we're struggling to judge them all. But the sheer volume of good entries we got meant that a good number of the initial people who agreed to judge couldn't handle that level of load, and it took some time to find a group of people who could. The code side of judging should be along soon, but I can't say when for sure. I'm hoping within the next month (to early November).
LXF: What have you learnt in the process of creating and running the LPC? What went especially well, and what would you have done differently?
CW: Well, we learned that the premise of the project went really well - way better than we expected, anyway! We were hoping that people would respond to the idea of the LPC. We didn't anticipate this strong a response. We built the contest along the lines of proving that collaboration in this kind of space is possible if you put enough front-loaded work into it.
The style guide was really at the core of it, and the money we put into paying for the base assets, and for having artists help us develop the style guide, really paid off.
Going through the Free Software Foundation for finances also helped a lot. I actually made some mistakes, and initially we weren't going to do everything through them, but I learned the lesson that going through a non-profit that has experience in dealing with finances just simplified a lot of things. Working with the FSF was pretty great!
And I was really happy to bridge the cultural divide of Creative Commons, the Free Software Foundation, OpenGameArt and Mozilla all working together on a project like this. The main thing we learned about what went wrong was that we really needed to be more prepared around judging. Next time, we'll do more planning around that so it's more smooth.
LXF: Is there going to be another next year? And will you try a different topic, like graphics for side-scrolling platformers?
CW: Signs point to yes! It's not confirmed yet, but I know at least OpenGameArt and the Free Software Foundation are interested. I'm not sure what my involvement will be, but I'll definitely be involved at least in advising and some mild co-ordination… probably not quite as heavily involved as I was this year. But the success of LPC has also stirred up a number of volunteers who already want to help next year.
If it happens next year, we'll probably have an entirely new style for a new set of games. I can't say what that would be at this time… we've only had vague conversations. Sidescrollers are a popular suggestion, so is isometric-style games, and some people are interested in 3D games. I think the latter might be harder to do, but there's a strong community of 3D people in the Blender community that maybe we could tap into.