My first assignment at Top Dog, and my first project in the game industry, was as a programmer on Lunar Golf. This original title was the idea of Berkeley Systems (makers of You Don't Know Jack and various screen savers) and was to be developed in full by Top Dog Software.
The development team consisted of three programmers, four artists, and various management personnel. My primary responsibilities were the physics and terrain simulation (which was based on an altitude grid generated from a 3DStudio grey-scale elevation map). The game was to be unique in that it showed the golf experience from a first person perspective, using a huge number pre-rendered 360 degree snapshots of various places on the course (Berkeley's idea, not ours :). Aside from the playability problems with a first person golf game, this "quantized" snapshot approach caused tremendous difficulties. Trying to control your swing without an external view; attempting to get the golf ball to land exactly on a snapshot node, and matching up the terrain grid with the warped 360 degree panoramic view, all proved to be serious technical challenges.

The first version of this game ran over schedule due to various technical problems and was canceled when it become clear that it could not make a Christmas release date.
After a brief hiatus helping out with the Final Fantasy V port, we learned that Berkeley Systems had been bought out by CUC and that they wanted to finish up golf (probably for legal reasons). Regardless, we formed up a new team, and as the sole survivor of Lunar Golf one, I was made lead programmer for the second attempt. Luckily we were able to start again from scratch, keeping only those modules that were complete and well documented. We also made the decision to take a more traditional approach and show the game in third person, and with a real 3D model of the course (so that you could hit the ball anywhere and put the camera anywhere).

Our team of four programmers did pretty well on this one as we had well defined roles, a complete design before we started, and a great list of what exactly NOT to do when developing the game. Though golf is not the most exciting thing to work on, we were glad to have a doable project and a reasonable schedule. Unfortunately, this was not enough to guarantee our success.


The second incarnation of Lunar Golf was canceled when CUC realized that they were funding direct competition for their own titles (namely the Sierra golf series). At least it was on-schedule and certainly finishable this time around. Perhaps six months into the schedule, one hole was playable with sample art and all major modules in place.