Keywords: design, motivation
The brain found some satisfying stimulation when I bought Dungeons of Dredmor for the ridiculously low price of $5 and found it entertained me for 3 days straight. Then I beat it on medium difficulty, permadeath with a juggernaut character who is basically am ambidextrous, sword-wielding, man-of-arms, artful dodger, who forges his own equipment. (Dredmor himself probably would have been the end of me if I didn't have a potion of invisibility and he wandered off and died somehow - probably from a Mass Destruction DoT.)

It's a pretty good refutation for my earlier belief that I'm sick of games, all games. Apparently I still like Roguelikes. This puts me in the mood to do some development.

How I've been screwing up for years.

Willingly, even, thanks to being overambitious.
Long version follows.

This latest space game I've been working on? I've had this idea I wanted to do a game involving open-ended space exploration. Like, a Star Wars RPG, but nothing copyright infringing, just a space opera adventure.

Unfortunately, I started the wrong side first. I built the outside of the game: a structure of supporting solar systems, planets, spots of land on those planets. It sucks eggs for a rather important fundamental reason: space is mostly only empty space, boringly so, and there's literally nothing to do in it in my current implementation.

The above video is of the Get Something Done Challenge for 2010. It's heavily based on the Future Shock code I've given away over here.

But, while I have given you a lot of cool icons in there, I would not recommend actually trying to finish the game. This is because I heavily mired myself down into this whole idea of an advanced interface where I need to code three objects for every single action. The entire game uses no verbs at all (other than to take input from the mouse and keyboard). This all came from wanting to have a visible action icon for all actions on the skin control, and probably could have been avoided if verbs inherently had icons.

Here we have a video that ultimately became Planetbreakers. What did I decide to do with that game? How about have three completely different player factions that play completely differently. One plays like a RTS game, another plays like a vehicular combat game, and a third plays like SimAnt.

What you should be taking away from this.

Are you seeing a pattern here? In every single case, I overcomplicated things before making the game playable.

It's the curse of an overly-creative mind, really. I get these really cool ideas that I want to put in the game. However, the game never really becomes playable and, instead of completing the project, I wrote a ton of code for features that simply slowed down the development process with sheer bloat to the point where it exceeded my pain threshold to continue. I end up wracking my brain against a sort of writer's block: why isn't this game fun yet? Try because it's not actually playable, me.

Okay, so I'm going to have to start over again. I can probably reuse some code from this latest Space Game I wrote later, but not Future Shock or Planetbreakers for reasons I explained above. My goal is now this: make a simple, fully playable Roguelike game. Only after it is fully playable will I start inserting cool features.

It does not get much simpler than this. BYOND is actually about 90% of the way to a Roguelike before you write the first line of DM code. About the only complicating factor is a map generation routine, which I already know how to do, and there's libraries to do it for me if I don't want to. It doesn't get much simpler than that!
I really like Dungeons of Dredmor, been playing it a lot lately myself. It has a lot of personality I think. It also lets you choose if you want permadeath or not if you're into that sort of thing.
Yeah, it's a great game, it's really hard not to recommend because it will undoubtedly give almost anyone more than $5 worth of entertainment.