The "Experience"

As this project has progressed, our overall goal has become clearer - this "out-of-the-box" experience is not some tech demo, nor is it some cool game that people can play and say "This is really awesome." Our project is simply a quick introduction to Project Darkstar, how the technologies help developers, and what it can be put to work on, namely powering the server side of massively-multiplayer games. Having a user create a full fledged massively-multiplayer game is way out of the scope of this tutorial - many large companies take a crack at it and find themselves failing epicly while trying to make one. So we have to scale the scope of the OOTBE to be compact, quick, and informative. Fortunately, there is this great open-source project at Sun called Project Snowman that is exactly what we're looking for - its a 3rd person capture-the-flag style shooter where players fight each other as snowmen on a snowy tundra. Fortunately, the code base is pretty solid - with a little bit of digging I was able to find where Entity updates were handled, how messages were parsed, and how players communicated with each other, among other things. By "tutorializing" Project Snowman, hopefully we can give developers a fun and information introduction to the Project Darkstar technologies so that they can go ahead and use Project Darkstar on their own personal endeavors.

While I did mention Project Snowman last week, we were not exactly sure if we were going to use it. We were told that the code was not good - from the looks of it, the only thing lacking is great documentation. The code itself is written in a fine manner, and while it might be good to tweak it, its not so bad that we will be spending our time doing that, since there are far more important matters at hand. The key thing here is that developers have a good first experience with Project Darkstar - first impressions are lasting impressions, and if we want to get more people using it then they have to have a good first impression of the technology. We believe that we can achieve this goal, but it's not going to be easy.

"Speaking Points"

The subject of this blog is about this thing called "Speaking Points," and in particular, speaking points in relation to Project Darkstar. So far, we've spent most of our time researching and learning, but we've also done our fair share of production in the past week and a half. So far, we've nailed down our out-of-the-box experience to be a culmination of 3 things: An installer, a game/game-framework, and a series of tutorials.

The idea of an installer was not obvious to us from the start. When we first made our way to Sun, we were expecting to sit down, load up Darkstar, and immediately crank out the greatest game framework in 7 weeks with the most in-depth tutorials possible. However, in life, the devil is in the details, and the details of how to get Darkstar up and running were not easy to track down. Our first experience with Darkstar was that it was cumbersome to use and information on how to intially get it up and running was either outdated or nonexistant. Fortunately, we work with the guys who are making it, and their expertise was invaluable to learning how to set it up initially. As well, we were able to track down some minor issues with the loading process that only affect a small number of PCs, but in the end we were able to understand how to use the system. Still, for a newcomer, it wouldn't be surprising for them to turn away based on how intimidating it may be. In reality, its not bad at all. But, for us to say that this was a successful OOBE, every second has to be streamlined. So, our team came up with the idea of an installer, which the Darkstar team was all for. In the end, there will be both a GUI and shell version of the installer. What Chris ended up doing was writing a front-end installer framework, and what I've done is written an installer back-end framework. The goal is that making changes to the backend will only require a minor update in both front-ends, so the change is reflected in both views. Fortunately, most of the that is complete. I predict about another day and a half of work to get both the GUI installer and shell installer complete. We still have to write test-cases though - those might be tricky in our system, since there is a lot of file movement, and simulating that can be difficult unless properly done.

The two other parts of the project are the actual game as well as tutorials that go along with the game. I will touch upon those in my next blog post.