One of the authors talks about optimizing .kkrieger's size here: https://fgiesen.wordpress.com/2012/04/08/metaprogramming-for-madmen/ https://fgiesen.wordpress.

One of the authors talks about optimizing .kkrieger's size here: https://fgiesen.wordpress.com/2012/04/08/metaprogramming-for-madmen/ https://fgiesen.wordpress.com/2012/04/08/metaprogramming-for... So for our shared entertainment value, here’s the full list of things (so far as I know) he didn’t trigger during his trial run, which subsequently got compiled out of the released version: - We used shadow volumes; there was a shadowing path for one-sided stencil cards (two passes) or two-sided stencil cards (one pass). The card we were recording on (a GeForce4 Ti) had two-sided stencil, so the one-sided stencil code didn’t make it into the released version. Oops. (This one had nothing to do with “user error” during the trial run). - In the menu at the start, cursor-down works, but cursor-up doesn’t (he never hit cursor-up in menus during the test run). - The small enemies at the start can hit you, but he didn’t get hit by any enemy shots, so in the released version of .kkrieger enemy shots deal no damage. - Part of the collision resolution code disappeared, since it was never used in the trial run.