We’ve passed Christmas and made our way through a Steam sale with our wallets mostly intact, and now most of us have a pile of games that we’ll probably never actually play.
Game programming is hard. Setting aside the “cultural” problems in the industry- endless crunches, compensation tied to review scores, conflicts between publishers and studios, and a veneer of glamour over unglamorous work- the actual
work of developing a game is a hard job.
Building a game engine is even harder. Not only do you have to build highly performant code, you have to build a system flexible enough so that game developers can build a game on top of it. You need to provide a set of high-level abstractions that make it easy for them to build a game, and this is where the problems come in.
For example, I went through a brief period of playing
Frozen Cortex, an interesting approach at a turn-based sports game. I was stunned at how
badly it performs, though. Weirdly, it’s not during
gameplay that performance stinks, but when staring at the
menus. I was puzzling over this for some time, when
Anonymous sent us a message.
You see, Frozen Cortex is build on the Torque engine, and
our anonymous submitter is working on a different game that also uses the Torque engine. And they’ve encountered a few… special warts.