XNA Then and Now Part One

Hello everybody and welcome to the XNA Team Blog. This is an exciting time for myself and everyone else on the team as we finally get to go public tomorrow morning at Gamefest with the “Big Thing” we’ve been working on.  I’m pleasantly surprised that it hasn’t been leaked and expect there to be excitement out there as the implications of what we are doing become apparent.

I’ll talk more about the big announcement after it happens tomorrow. Until then, let me tell you a little about the strange/fun journey which as been XNA.

XNA was announced with much fanfare at GDC in March 2004. For many reasons the first XNA announcement was louder than it should have been.  Everyone wanted to talk about the Xbox 360, which wasn’t going to be announced, or even confirmed, until E3 in May. Next generation graphics were the thing and we knew we had to show some of it.

The real XNA message was supposed to be about solving the deeper problems in the game industry. Most people don’t realize this but games are a tough business. As the graphics quality bar rises, so does the art costs. Many games are better looking than they are fun and their sales suffer while the costs soar. Most games lose money while a few make lots. Hopefully, the ones that make money make enough to cover all of the ones that lose.

This creates a situation where the industry is afraid to try anything really innovative. Certain game play styles and genres sell in a “predictable” sort of way and, thus, appear to have less financial risk associated with them. This is why we see so many sequels on the market, so many copycat games, and so few real innovations.

To throw more fuel on the fire: making games is hard (we hope to fix that). In addition to the technology being hard, game studios have been fast and loose in their development process creating poor practices and controls around how things get done. This leads to delays, overruns, and games that are unplayable when they ship and require a big patch to finish them. It isn’t fun for the people making the games who work long hours, and the turnover in the industry is crippling. There just aren’t enough educated and talented new people coming in to meet demand.

The big XNA announce at GDC in 2004 was really about a project to fix some of the deep issues in the industry, although most people thought it was about the Xbox 360 which we couldn’t talk about yet. Some people thought we were going to make a game engine, and some thought it was all a big joke as then things went quiet for a few years. None of those were true.

Just over two years ago, I was the Development Manager for Xbox Live. I must have done a good enough job helping to define and build that service that they asked me to run the new XNA team. The offer felt a lot like the beginning of XBL as there was a lot of opportunity, big meaty problems, and very little clear direction. The task was: given the outline I just described above, go figure out, build and ship whatever XNA should be. And don’t take too long either.
I quickly made a few key hires and got a small, but talented team together. We then talked to a lot of people in the industry. The idea was to not only listen to their complaints, but actually watch what they do and see what bothers them the most.

Here is a sample of issues we found:

Most of the people on game teams are artists and story tellers, yet all of the “telling” is owned by the coders, who would rather work on engines.

There is very little innovation because of financial risk averseness.

Burnout on teams is very high, and there aren’t enough new people coming out of the schools to backfill.

Even if you do get a relevant degree, it is hard to get into the industry because of a certain amount of elitism.

It is too hard to build games – in the technical code build sense. Art assets are complicated and intertwined and your C++ linker doesn’t do anything to help.

Creating a game is hard. The APIs, tools and such have all been written with the expert coder in mind. Unfortunately, the best game designers are usually not the best coders.

And so on.

So we started multiple investigations to see what we could do to help the industry. The first two, which I started talking publicly about at the CEDEC conference in Tokyo last year were a build system for art and a related set of asset management tools. These were on the CTP we released at GDC this year.

The other project was secret for a long time. We were working with Mike Zintel’s .NET Compact Framework team to see if we could bring their technology to the Xbox 360. Rob Unoki and others on that team did a stellar job and we were able to show managed running, and being debugged, on a Xbox 360 Development kit to our execs in September 2005. After this we zeroed in on what we wanted to ship first, finished hiring a team that can ship code, and have been running like mad ever since.

GDC 2006
The GDC 2006 show was an important milestone for us. We showed working managed code running on an Xbox 360 development kit and talked about how it is going to help designers and others get games written more quickly and cost effectively. Just as importantly, we shipped a CTP of some of our tools and gave away all the art and source code to the game MechCommander 2.

Let me repeat that. We got permission from the company to give away all the art and source code for the game MechCommander 2 (minus the networking code). A big part of the team philosophy is to give back to the community.

We want people to see how games are built. We want people to learn how to build new games. We want people to do innovate things that nobody is expecting. We want people to surprise the industry with new genres and new gameplay styles. We want to see the equivalent of the Sundance for games. We want to see more creativity. We want to make games, and game development fun. We want to lower the barriers to entry and we want more people to participate.

I’ll talk more about how we intend to achieve this tomorrow.

Boyd Multerer
Product Unit Manager - XNA