Sponsored By

Postmortem: Torrent

Postmortem for Torrent, a 3D FPS built in the Unreal Engine

Navin Supphapholsiri, Blogger

October 7, 2013

3 Min Read

Enter a world of high stakes digital combat – cyber warfare is the way of the future, and in Torrent, players are on the front lines. Two teams of hackers fight over critical data points in this fast-paced action game. Each player controls a cyberspace avatar, an emulated combatant in the digital world. These avatars represent the hackers’ code in lethal combat through a simple interface: that of a first-person shooter.

Torrent was developed in UDK by ten developers at the Guildhall at SMU. The project lasted 12 weeks, with fifteen man hours per person per week. As the Producer on this project, I present this postmortem to outline the tragedies and triumphs we faced as a team starting from the pre-production phase until the final release date.

What Went Well

Strong Processes

Pipeline processes established by each department helped streamline productivity for our team. Whether it was quality control, asset approval, or game design changes, the process to deal with these obstacles consumed minimal production time allowing the developers to spend more time on development.

Morale Boost After Vertical Slice

It is very hard for developers to see the big picture during the early phases of development. Even though we have documentation and references to refresh our vision, it was difficult to gauge our progress. As we arrived at the Vertical Slice milestone and presented our deliverable to the Greenlight committee, we started to see our hard work brought to fruition. This served as a huge morale boost to the team and allowed us to drive the project forward.

Plethora of User Input

During each week of development, we were able to perform a dozen playtest and usability sessions that provided us with adequate feedback. Since development mostly took place in our studio room where other groups working as well, we were able to gather feedback from upper and lower classmen as well as the seasoned industry professionals (the faculty). The feedback gave our team the ability to iterate on our features, which improve the overall quality of Torrent.

What Went Wrong

Feature Creep

During the early phases of production and after each milestone, the user input received inspired many ideas to the design team. This resulted in minor aesthetic and gameplay changes that required our artists and programmers to work additional hours to implement these new and exciting ideas. Although these additional features added quality to Torrent, better planning should have been made to prepare for these changes.

Communication Issues Between Disciplines

Communication between the design and art team should have been better. Designers on the team were using the assets created by the artists in creative ways. Sometimes a pillar or crate would be scaled five to ten times bigger than it should have been to serve as trim or decoration. This reason for this was because assets were not being created as rapidly as the designers wanted. In my experience, this seems to be a common problem where the artists generally want more time to spend on their asset creation process while the designers always want more assets use in their levels. In the future, we will have our leads communicate with each other more often and run cross discipline approval processes between the art and design team to make sure everything is moving forward as intended.

Hardware/Software Issues During Mid-Development

We had all sorts of hardware/software issues with our Alienware laptops. Some issues that occurred were as follows: UDK engine bugs, Visual Studio malfunctioning, Perforce connection issues, Autodesk 3Ds Max errors, etc. We did the best that we could in dealing with these setbacks, but in the future, we will make sure that the software/hardware issues we have are tested before we start development.

Conclusion

Overall, we felt that this project was very successful and everyone enjoyed working with each other during and outside of core hours. This whole 12 week process gave us valuable learning lessons as developers of the gaming industry and we would definitely continue working on this project if the Guildhall curriculum allowed it.

 

Read more about:

Blogs
Daily news, dev blogs, and stories from Game Developer straight to your inbox

You May Also Like