As the analogy goes of the sculptor who knew the statue was within the block of marble all the time and just had to remove the pieces, it seems this game concept I had in mind was within my head all the time and I need to remove the cognitive dissonance. However, at least hammer and chisel versus marble has a certain air of finality to it. Cognitive dissonance has a tendency to grow back.

This "epic space game" I'm making is not marble, but a an organic collection of ideas. As it grows, these ideas have a tendency to mutate in new and interesting directions. Inevitably, these mutations begin choking the life out of the game. Then I've got to take a scalpel to the cancer and hopefully have more usable mass left afterward than before I started.

Cryptic talk aside, I'm saying My Own Net Dream tends to shift with my mood. I don't know if it's because I get bored of my old ideas or if I'm learning more about what a truly good idea is. Probably a bit of both.

Current screenshot. No, the interface isn't done. Yes, the name of the character is randomly generated, and you can control more than one at a time.
Currently, it's shaping up to be perhaps closest related to games like Starflight or Star Command (the 1988 SSI game and not the 1998 RTS game), but also something uniquely itself:

• You have a space ship. One of them per player. You can poke around its interior and watch the goings on inside (see the screenshot above).
• You roll up a number of characters to serve as a crew. No longer a Roguelike so much because there's now multiple characters. They'll likely be commanded in a manner similar to Dwarf Fortress, with some forced commands but mostly going about their own business.
• You explore a procedurally generated galaxy. (Okay, so Star Command and Starflight's galaxies weren't random, but they did generate a lot of content from formulas.)
• There's quite a bit more to it, but it wouldn't be any fun if I ruined the surprise. Besides, any greater details than these I've an even greater likelihood of changing my mind about.

Of course, one thing my game has that neither of those did is that making it a persistent state shared multi-player environment, that's a major benefit of working in BYOND.

My continual focus right now is to not get too bogged down in the minor details; Make the game completely playable. It's easy to get distracted. For example, I spent a day or two teaching spawned mobs how to identify each other as threats and attempt to kill each other like civilized beings, then realized I had essentially stopped to put bugs in a jar and shake it. The finer details can wait, let's see if I can't just get the whole machine running and doing what it's supposed to do first.