I am unofficially (whatever that means in this context) considering November to be National Game Writing Month again, and am making an effort to get serious traction on Neon Galaxy. To wit, here's the stuff I've implemented since the beginning of the month:
Got the inventory fully working. When weapons are added to the player's inventory, they are displayed in the user interface and actually modify the player's current firepower. The system is flexible enough so any inventory items that need to do something to the player ship have the hooks they need for it, though the weapons are really going to be the vast majority of that.
Added a much less annoying way than SVG for getting vector graphics into the game. I wrote some code for loading vector graphics in a simple XML format, and... I am going to draw them on graph paper and then type in the results by hand. Believe it or not, that's much less of a pain in the ass than using Inkscape (uuuugh) or teasing out the actually useful information from an SVG bloatfile (aaaargh).
Got intra-level gameplay logic implemented. You spawn at an entrance, can destroy the reactor, then have 30 seconds to reach the exit safely.
Got the scoreboard working again.
Moved all the Neon Galaxy-specific physics classes into that project.
Removed the ridiculously complex and pointless collision filtering system, which never worked right.
Designed a giant list of AI-controlled enemies to add to the game. Believe it or not this never existed before.
And a crapton of cleanup.
My next course of action will probably be to implement enemies, one by one. As I get each one working I'll implement its specialized systems in the process, making the next one easier. After that will be collecting inventory items and loot, and then inter-level progression. I bet that'll carry me beyond the end of the month, though.