I hate memory errors above all others, mainly because they're so damned hard to trace back to a particular system. I'm just glad the game code is still basic, allowing me to root out these sort of problems without destroying too much code in the process. I'm thinking I'm going to add a generic particle system and use my old controller code I think, at least then it'd allow me to unit test the thing.
On another note, I saw "I, Robot" last night. Slate me if you want, but I really enjoyed it. It's standard Sci-fi fodder, "man makes robot, robot gets intelligent, robots take over humanity, man gets control back and pwns the robots again". I'd watch it again, it's not one to watch over and over (like Army of Darkness and Fear & Loathing in Las Vagas), but it's definitely watchable a second time round.
Today I was walking through town and had an awesome idea for a mini game, damn me and my ideas - I have too many. At least I try to keep mine small and reasonable. It's a shame a lot of beginners to programming have such grand ideas, I feel sorry for them once they come crashing to the ground. I suppose that's how I am, I rarely try to think beyond my capability, using each project to further that capability for the next one. The new idea will be kept under wraps, if at least to stop me thinking about it, getitng too excited about it and dropping my other million projects in favour of it.
Tonight I think I'll fix up my errors in Manta-X and throw a couple of enemies into the equation for fun, or perhaps finish up the script wrapper for jsI. Decisions, decisions.