I've successfully broken the surface and have reached feature implementation to the point where I can truly claim I have returned to work on this game:

Feature-wise, I can remind folks of features I already revealed in previous entries that I'm pretty sure are going to make the cut:
  • An open ended universe that is capable of generating solar systems, planets, and even individual plots of land on those planets.
  • This is a persistent world style game.
  • Everything is dynamically generated or generated from map templates for potentially unlimited adventures.
That's no pipe dream: I know these features will make the cut because they are already in and functioning. (Well, I'm cheating a bit to say the "unlimited adventures" are in: what's in is the structure in which I could add content that presents unlimited adventures.)

There's more to the game than that, but I don't want to reveal any more features right now. Part of the reason why is because announcing things before they've been added tends to kill my motivation. The other part is because I do not want to paint myself into any corners. Therefore, I am in top secret development mode for a lot of cool stuff I'm developing right now.

Currently, I have a rather simple goal: get this game into a playable state using the quickest, easiest-coded interface possible before worrying about more elegant solutions. This is of sharp contrast to my earlier projects where I would get mired down in overriding BYOND's existing solutions with elaborate skin controls. A few examples:
  • In the above screenshot, the navigation pane in the upper right corner is a manually coded skin control, doing everything the info panel already does but with me having to manually code every little redundancy check. (It's a necessary evil in this case because I need a second, simultaneously active info control.)
  • I would go without using verbs unless they were hidden and being directly accessed by a skin control, keyboard, or click. Elegant-looking? Yes, but it would also quadruple the amount of work required to code what was needed to perform the average in-game action. If only verbs were capable of generating their own access icons.
Is it any wonder my earlier projects foundered? I made them about eight times harder to make than they needed to be. Isn't the purpose of a friendly IDE/API like BYOND if not to mitigate the coding busywork in order to allow its users to creatively realize their ideas? When I get distracted by cool code ideas, I'm inventing more coding busywork for myself. I'm trying to break that habit and instead refocus my mind on game mechanics.

The alternative is to use the standard info control and verb functionality for as much as possible. Later, I can come back and make the game much prettier and more accessible, but that really should wait until I have an already playable game.

I have another goal as well: never stop developing (with the exception of perhaps a couple hours in the morning and evening just so I don't burn out). I don't think any good can come from falling off the wagon again, it just causes my brain to fossilize from lack of use. So, if I'm not programming/learning to program or creating content, then it better be an emergency.

Before somebody points out that writing these things is procrastinating from these ends, I'm going to say that I'll restrict myself to one blog entry a week. If I succeed in both my goals I may have a playable alpha or beta to show for it by the time it comes to make the next entry.
What screenshot are you referring to? =)

I'm stuck in your rut now. I have procedural generation tools, but my designs keep floundering and my time has been spent in Steam games.
Is the screenshot working for you now? I had the file on "Hide" mode :P
Yup =)
Looks like I'll be doing a bit of revamping. This framework I built back in December was based on using obj. Trouble is, the universe has a potential to get too big! I need to use datums instead.

Still, it's not /too/ hard to do. Most of the existing code works, I just need to adjust it a bit. Probably won't take me more than a day or two to get back to where I was.
Halfway there. Making my own dynamic linked lists was a lot harder to wrap my head around than I expected it to be.

It was a necessary step to prevent the lists I'll be needing from counting against the user-defined list limit. If you have 625 systems of 5 planets each and each planet has 20 lots of land, and each keep a list of what they contain, that's 62,500 lists right there. That's just a 25x25 universe of systems, and I'm already at the user-defined list limit!

This transition away from using obj paid off in that I can now generate 1225 systems of roughly 5000 planets in under 30 seconds.

Of course, I'm not generating the actual *turfs* for those until the players request them, but all the data needed to do so when required is there. The goal is to make it so maps will be swapped in when players need them and swapped out when they don't. That way, hopefully, I'll be able to accommodate dozens of players per server.