Microsoft

Project Spark is simple, powerful, and creative, but Microsoft failed to show it well at E3

Project Spark is simple, powerful, and creative, but Microsoft failed to show it well at E3

Sooner or later when you attend enough E3 game demonstrations you end up in a group with a guy who just refuses to stop asking questions. Will the player be able to do X? Is Y possible? Will I be allowed to Z?

At a Project Spark demonstration inside Microsoft's booth this week, I was in a group with one of those guys, but for once it was more fascinating than it was obnoxious. See, every time he asked one of the developers of Project Spark a question about what could be done in the game, the developer sort of looked up in the air and escaped into thought for a brief moment.

He wasn't trying to remember the marketing lines like so many other presenters, it was more like he was flipping the beads around on an imaginary abacus in his head. Rather than spouting a stream of facts about what was in the game, he was forced to dream up a new scenario to see if he could figure out a way to achieve what was asked in each question.

When: do

That's because Project Spark is different from most games on display at E3 this year. Rather than a traditional game, this is a game that helps you create other games. It's a programming language and artistic tool made so simple that the average player could accomplish something without a computer science background.

“We heard about this great Microsoft research project called Kodu,” said Eddie Parker, a producer on Project Spark. “Kodu was a project that was trying to bridge the gap between kids who wanted to create stuff and adults that knew software programming. So Kodu is a simple programming language based on cause and effect. The way you code is based on 'when' and 'do'.”

Those two simple prompts can create elaborate results. You tell an object what to do and when to do it, and that's it. You've got a simple command up and running. For instance, “when” I get close to an animal, that animal will “do” a pre-determined action like running away or attacking.

“But just because it's simple doesn't mean it's not powerful,” he said. “So we took the ideas of the Kodu project and married it to this great engine for building and sculpting terrain and it ended up being a winner. We work for three weeks at a time on Spark, and then at the end of that, if you're finished with all of your official work then you can participate in our game jam, and we judge the entries over some beer and have a good time.”

In one of these early game jams, Parker says that one of their testers built a simple, working synthesizer. In some ways it's a very simple “when” and “do” scenario. But this tester went a lot further and built in fully functioning buttons, including a record button. So you could press record on his creation and it would be able to play it back for you.

The catch of this is that according to Parker, saving and loading within creations didn't yet exist in the game. So this tester was able to get around the limitations of the system's code and build something into the game that didn't yet exist.

When he tells me this, I instantly think back to some of the amazing creations the Minecraft community ended up building. Not just castles and monuments, but working circuit boards and calculators.

Tip of the iceberg

In a way, Project Spark seems to have drastically undersold itself at the show this year. On stage at Microsoft's press conference, the developers showed a lot of the basic functionality. They showed how you can sculpt the land with the touch of a finger. They showed how you can add a lush landscape with a very simple paintbrush tool, and how you can give creatures and objects simple behaviors with the easy to understand “when” and “do” programming system.

It was only when I stole a private moment with one of the game's developers on the show floor that I learned about some of the most interesting parts of the game. Their goal seems to have been to inspire people to want to create with Project Spark, but the main allure may turn out to be playing the games that other people create and upload. Even Parker acknowledges that only a very small percentage of players will end up being serious creators.

Several times this week I watched the game's designers create simple environments within Spark, but they only paid lipservice to the actual games that have been (or will be) created with this game creation toolset. They briefly showed a recreation of Limbo along with a turn-based RPG.

This is what I want to see more of. I want to get a glimpse at what the game's community will be able to do once the game releases. As with Little Big Planet before it, the true allure of Project Spark is going to be downloading all of the free content that gets dreamed up by the community.

Project Spark is a game creator and a world creator, but it also has another mode that is a choose-your-own-adventure generator where the world and the quests are created dynamically, along with your input. And beyond all of that it may also turn into a multiplayer game as well.

The developers aren't talking about multiplayer details, but our presenter made allusions to a mode where you could invite a friend to play in your world with you. He wouldn't say specifically what a mode like that would entail, but seemed to suggest that you could have a game of Project Spark running in creator mode with multiple friends in the level building, not unlike a Minecraft server. This is interesting mind candy, even if it may not make it into the final game.

Ultimately, Project Spark was one of the more interesting games I saw at E3 this year. It's just a shame that we weren't able to get a better look at the full potential of this neat piece of software.