Construction is a skill, and that's why construction games are usually very interesting to me. Whether you're building a bridge or a space ship or a castle, you are relating these objects in a complex space and then testing their fitness.
Something like Kerbal has more staying power because it has a lot of different kinds of tests of fitness, especially with mods. If you're doing a bridge-building game, the only test is to have a bridge that things can go over. You can build a lot of random fun stuff, but the game doesn't particularly care. On the other hand, in Kerbal the primary fitness test is launch survival... but then the world opens up and you can choose what kind of challenges you want to take on, and it offers you feedback based on those specific challenges. Successful mission to Eve? There's about a dozen possible approaches to that, and your design can therefore be rated in a dozen different ways. Lots of open-ended design.
That in mind, there are pretty basic restrictions on how you build things - in Kerbal and in everything else. There's a particular primary concern that everything arises from. In a bridge-building game, it's the bridge's ability to resist weight via supports. In Kerbal it's the ship's ability to survive thrust. The specifics of the vehicles/thrust change the fitness test, but it's still the same basic category.
Changing Kerbal or a bridge game to subvert that test weakens the game dramatically. For example, in Kerbal there are mods to help you set up base on the moon and launch from there. This sounds really cool - and it is - but the stresses of launching from the moon are microscopic. The challenge is slack as hell - as slack as if you set all the cars and trains to weigh 0 kilos in a bridge game. This isn't necessarily deadly - if there's another fitness test waiting to be applied, it can still be fun. But there generally isn't.
I was thinking: what other kinds of core fitness challenges could you create for a constructive space game?
Well, I've talked about aging: taking into account how durable things are and how they will break down. Then it's more about balancing time instead of balancing rigid mass. There's also the ever-popular option of combat as a fitness test, but that's boring.
I was watching a few videos about people putting together rockets and the CMS, light techno playing as they painstakingly assemble all the pieces. Here's one short example: CMS install.
And I thought, hey, there's an idea. What if the final structure of the device/station/ship isn't the main fitness test, but instead it's the actual construction?
For example, in the last essay I talked about using chemicals in an alien atmosphere to create drinking water. You'd just slot in the various conversion nodes - air intake, CO2 filter, CO2->O2 converter, O2->H2O converter... while it might be weighty or power-hungry, it's not really part of the core fitness test, because there's nothing particularly interesting about the physical layout of the nodes.
On the other hand, what if you had to ship them into space all disassembled, and then assemble the unit in space? These aren't structural modules you're slotting in. You've got some kind of structural frame, and you're embedding racks, wiring up connectors - all in a confined space, in zero gravity.
Now it's a topological puzzle. You're in space. Can you get the chemical processor into the area you're assembling it in? Can you rotate it to the right orientation? Can you fit it into the right spot, past whatever other things might be in the way? Can you get to the connectors and screws and tubes to actually connect it?
You might think "Oh, just assemble everything in an easy-to-get-to tower..." but you can say the same things about all construction games. "Just connect every part of the bridge to every anchor" "just surround every part of the castle with garrisoned walls".
The question is: what pressures the player to try and cram more and more stuff into less and less space? In most construction games, they limit the number of things you can have by putting price tags on everything. In our game, we would need to find a way to limit the space you are allowed to fill. The obvious answer is to have the containment structures/frameworks have specific sizes and shapes. You certainly can just put one thing in each framework and leave yourself a lot of space... but the framework is the heavy part of the space vessel, so that's going to make it hard to accelerate and turn.
I think it'd be a fun challenge to make the assembly process feel right - feel controllable and not too pat. The player might make a kind of "record" of how to do it while designing the ship in the first place, then NPCs could assemble any number of them in game days. You could even watch them assemble in the background of whatever else you're doing, perhaps with light techno playing in the background.
There are a few key areas to balance:
1) How the shape of the object fits into the chassis is obviously critical, and you might need to make it adaptive or have a variety of each part to support different fundamental shapes of chassis.
2) The connectors between objects are just as important, because you have to connect whatever you put in to whatever else relies on it. This makes the order of construction really matter, because a clever designer will put connectors to object A and B behind object C for space concerns, and then just be sure to add object C after adding and connecting A and B.
3) Mechanical structures matter, because racks which slide out, lids which open, spaces which expand and contract, legs that swing away and then back... these let you get at spaces that will be locked back down, or provide access in ways you would have a hard time with. This means that the chassis variety can be extraordinary: not just spatial configuration.
4) Secondary effects such as vibration and heat matter. If you have something that vibrates butting up against something that's taking pictures, you'll get crappy pictures.
Sort of "techtris", if you will.