When we started our project in January 2004, all we knew was that we'd like to make "some kind of jump'n'run", and that it should be a small "quick and dirty" game. Both of us had already tried to program games on their own before that, but usually we didn't come very far (...a common ill with indie game programmers), so our idea was to "join forces" and, for a start, create a really simple game with the only goal of getting it finished, whatever it was.
So, without any real planing, we started coding like mad and, after a few days (and nights) of programming we had a simple tilemap with scrolling and a particle system. We were sure that this would take no longer than a month, maybe two. But somehow the whole thing got out of hand after a while, and when we saw that our code (we never dared to call it an "engine" due to a total lack of architectural planning) was actually working, we started stuffing feature after feature in it, even though the core of our game was never designed for it...
Now, almost a year later, Atomic Butcher is finally finished, and it's truly a miracle that it actually works. The code grew bloated and became ugly as hell, filled with heaps of hacks to circumvent our lack of planning in the first place. But it is done now (and didn't turn out THAT bad, we think), and we are actually quite proud of our first finished game, even though there are a lot of obvious flaws with it...and we learned a lot for our future releases, which will definitely come! (Even though it may take a while...)
Thanks for reading through all this drivel, and have fun with the ATOMIC BUTCHER! (And don't take anything in it too serious...).