A lot of programmers (and non-technical friends) wonder how I start a game.  For me, it usually starts with a few hand-drawn pictures of how I imagine the game looking.  Next, I play the game in my head.  I imagine pieces moving and the buttons I have to click.

The first important step, though, is the "feature prototype".  I pick a feature in the game and code as quickly as I can to get to that point.  In Mythicum Goetia, that was movement and combat.

I took a detour getting there, though.  I went ahead and built a lobby system and a way to add artificial players.  To the new coders out there, DON'T DO THIS!  I shouldn't have done this, but I trust that detours won't prevent me from finishing the game.

So what would I have done without A.I. players?  I would have made the game temporarily a hot-seat game.  You can play for everybody while testing the first few features.

N.R. made 3D dice for the game.  I had some bouncy-ball code for another project and I adapted it for dice bouncing against each other.  It's mostly convincing and it adds a lot to the board-game aspects.  Feature #1: Implemented and worth the time!

The next feature was to test combat.  This is a really rough prototype.  It even requires that you use a little imagination because we have none of the weapons/spells available right now.  Using your imagination is okay.  In fact, seeing 75% of your idea implemented helps to motivate you and focuses your efforts on what will be more fun.  Feature #2: Almost there and looks to be fun!  It might need some tweaking, but it's better to know after a few weeks of coding then after months of writing the game.

Next is a more practical feature... loading and saving.  We want players to be able to re-start a board game as though you left it on the table.  There will be a few issues: connecting a player to his character's slot and having A.I.s take over for people who can't continue.  I'll talk about this once I've got some progress.

Later!