Postmortem news

Starting in 2007, this project was supposed to become a sprititual successor to the highly successful Movie Battles 2 based on Jedi Knight: Jedi Academy. Even ten years after the game's initial release, the mod continues to serve a loyal online community and enhanced Jedi Academy's lifetime beyond any expectations.

In the beginning

Our original core team descended right from Movie Battles 2 in an effort to make some level design / scenario ideas possible that we could no longer realize on the aged id v3 engine. The Lost Coast demo level for Half-Life 2 and the Source SDK that became more or less independent from a specific base game got the ball rolling. We merged with a team that initiated "Star Wars: Source", renamed it to "Moviebattles: Source" and began working on some demo assets and levels to attract further developers. Due to the merge process, the project initially had no clear focus, but two branches:

• One game mode for CTF and TDM skirmishes that kinda imitated JK:JA's fast-paced gameplay
• One game mode for class-/objective-based gameplay with more tactical depth

As you can imagine, this prerequisite wasn't very healthy to get started, since it did not allow one common codebase or design structure for all modes. Nontheless, the mod quickly built up pace and the developer team grew fast.

Critical team fluctuation

While this was a great motivator and the work delivered by the artists was thrilling us with a professional quality level, we soon suffered from management problems. We tried to maintain the gameplay vision that we brought along from MB2 at first and thus, installed a core team that would serve as a vision keeper and direct people at a common goal - pretty much like a creative or game director in a professional production.What we didn't realize quickly enough at that time was that the high fluctuation of team members also brought in many different expectations in where the game should head. We soon got into a situation where it was impossible to motivate the team to provide the work necessary to finish the mod - people simply couldn't identify themselves well enough with the existing creative direction. After some skilled people dropped off, we learned that we had to democratize the development process. However, this caused new problems. As it turned out, the Source developer community had quite different ideas than the people we brought along from Jedi Academy. As the democratic process turned into major design changes, some of the old project initiators left as they felt "betrayed" with the game's vision. As a result, we were forced to change the mod's name to distinguish it from Movie Battles 2.

Down the course of the project, over 60 developers of different disciplines contributed to the project. But at the same time, there was a lot of fluctuation. People dropped off due to creative differences or personal time constraints; new people with new ideas joined in. As a result of the democratized development process, the game's direction changed several times, making some of the already achieved work redundant. While the atmosphere within the team improved, the development process started to stall. Without a stable goal to work towards, no one could really see where the project would lead - and more importantly - when it would hit a release-ready state. It wasn't until very late in the project that several people with more development experience joined in that were able to streamline the process. We introduced an agile development process very similar to SCRUM: Using weekly sprints with regular test sessions and meetings, we were able to communicate a sense of continuity across the team that was missing up until that point. Week after week, the team could see and discuss the actual progress and get a sense of the project's pace. Working like this, we could preserve the democratic decision-making process, but at the same time work towards common goals. A very important method to achieve this were the weekly conferences: Instead of having endless discussions on the forums, the time-constrained meetings were the points were we made decisions and debated them much more efficiently. We noticed how the entire team became a lot more focused; previously there were many suggestions that didn't really contribute to the project and were considered little private projects - such as adding jiggle-bones for the female characters as opposed to rigging unfinished weapon models that were needed. This totally vanished after introducing the new project management method.

Technical frustration

Unfortunately, the project was already too delayed at that point and technology had progressed a lot. Low-barrier engines like Unity had arrived and many developers abandoned the sinking ship that was Source SDK. With the arrival of Windows 7, many components of the tool chain stopped working and some critical issues stalled development for months, e.g. a defective projectile system that took us a year to work around or defective demo tools right when we were finally able to shoot a gameplay trailer. And it seemed like we weren't the only ones frustrated with the now outdated development tools. There were no longer any Source developers interested in supporting the project and the remaining ones had to drop out one after another as they entered studies or found their way into the industry to develop commercial games.

The end

What followed in 2012 was an attempt to open-source the project and deliver its assets to the community under Creative Commons license. The few remaining core members all agreed on this step, so that the community could get at least get the latest development version and perhaps assemble a new team to continue. We actually just recently got a request from Gryffin Studios to do just that. However, we couldn't just simply assign a free license to the assets without asking all the original authors for their permissions. As you can imagine, tracing down over 60 former contributors is tough, especially when they are no longer reachable through the only communication channels known by the team. We could negotiate the rights to a few levels, assets and the source code. But the majority of assets such as many of the player models or weapons remain inaccessible after several unsuccessful months of trying to get in touch with their authors. With just few assets being available to provide under a free license, we decided to discontinue our efforts. The assets that are available are source-centric and with that many people having moved to other engines or professional work, there's no reasonable chance anyone is going to put in the effort to recreate all the assets for Source after all these years.

Conclusion

If we had started with the experience we have after the project, I think the project would have had great chances of actually reaching a solid, playable release. While we had incredible artists on board, who were a pleasure and a great experience to work with, we didn't have the appropriate project management methods for such a large project in place until it was too late to realitically reach a stable release. Following our experiences, the recommendations I can give to other large developer teams are the following:

• Involve your members, don't structure straight top-down.
• Try to establish some brand values that every suggestion should be measured against. Even if there are many different creative influences from several team members, this is a very useful instrument to keep a common vision.
• Establish short development cycles ("sprints") of 1-2 weeks with gameplay sessions for the teams to make progress perceivable and keep motivation up.
• Use these cycles to make decisions; don't span decision-making over endless forum threads.
• Don't allow the many ideas that come from your creative crew to lead to feature-creep. Try to measure your sprint pace to find a realistic project scope. Try to set deadlines for specific feature sets or even smaller releases. It's important to have a fixed and reachable goal to work towards.

Last but not least, I would like to thank everyone involved for their passion, their creativity and their effort they put into this project. Even if we couldn't make this dream come true, I think we all could learn a lot from the project and it certainly wasn't a waste of time.

So, what about Star Wars modding?

While Movie Battles 2 is still alive and working towards another release, the UDK-based successor has been canceled, as it never really managed to get into a stable development pace. If we want to see great, new Star Wars mods coming up, a lot depends on the new license owner Disney and its licensee Electronic Arts. LucasArts was known for letting fans do their thing to keep the fan community alive that was always so important for the franchise. Disney however appeared pretty protective of their brands in the past. A new, good Star Wars online title with mod support of its own could possibly revive the modding community to the extent that we knew from the Jedi Knight series. Looking at DICE's and EA's take on modding in the past Battlefield titles however, I have no real hopes that the next Battlefront will make an impact. It seems like the fear of hackers and the urge for data mining has become more important than creatively engaging the fan community. Perhaps the new J.J. Abrams movies will inspire people well enough for ambitious standalone projects and then we'll see if Disney will stick with LucasArts' tradition.