Monday, September 16, 2013


So, I talked about noncombat starbases and starships a lot for the past few weeks, but one thing I didn't cover were the visual characteristics of these games.

It may sound unimportant, but it can be very important. A big part of the draw of science fiction games is the feel of the world. That grandeur and mystery - sensawunda, if you prefer - is very important. The very first thing that helps to establish it are the core visuals of the ships and stations. Especially if a player can custom-build things, they'll want to custom-build something wonderful.

There are a lot of details about this that really change how everything feels. For example, how customizable are the ships? Are we talking about swapping out the nacelles on a Star Trek vessel, or about sticking together modules, or about arbitrary mesh placement and deformation? Each has its own strengths and weaknesses both in terms of visuals and gameplay restrictions, and each also has constraints in terms of the resulting visual design.

For example, freeform node placement allows you to put ship components like nacelles, hab bubbles, and so on anywhere you like, overlapping with other components and tilted and scaled in whatever way you think looks cool. However, if the components actually have functions (like weapons, engines, etc) this arbitrary placement can make it feel very awkward when someone places them in a manner that should logically prevent them from working properly. It also makes it difficult to internally connect the components, so it's not really possible to have module A clearly connected to module B - they're all just part of the same mishmash.

On the other hand, a hardpoint component system allows you to place components on the hardpoints of other components. This system is extremely good at relating modules to other modules, and therefore you can create resource chains, interiors, and so on that make sense. Visually, however, the locked-in-place modules are quite restrictive, and you have to be careful to design your module components such that they result in a visually interesting result. It's a lot more limited, and typically the result is a very boxy, militaristic style, because that goes well with the aggressively locked-in orientations and edges.

Those are far from the only options, though.

In my case, I really got to thinking about more organic-looking designs. I got interested in the joins between the modules. In most games, the modules just bump into each other or arbitrarily overlap, but there is value in having a very clear joint or clamp. It looks cool and you can make it highly functional. I was examining some examples when I realized that there was no reason for a space station to have to be ship-shaped.

A space station has no real thrust requirements, so it doesn't have to be shaped like a brick. In fact, it makes a lot more sense for a space station to be sprawled over a vast amount of space: more room for solar panels, less collateral damage if one part is destroyed, a wider separation between living areas and loud, possibly radioactive industrial areas...

So my design today is actually inspired by a sprig of flowers. That's why it's named after flower arranging.

The idea is that a space station "seed" is flung from earth out into space at something like 0.3c. It then spends the next few decades slowly decelerating into orbit around a nearby star.

Over the course of the journey, the seed blooms, growing out a branching, bud-heavy space station. When the distant star is reached, everything begins to unfold and bloom with the energy of that star.

This approach has a few strong points.

1) It's very visually distinct. As far as I know, the "unfolding flower" visual is still very rare in science fiction, and I've never seen it as part of a long 'branch' of flowers.

2) It works well with multiple identical modules. Most science fiction visuals start to look bad if you have dozens of the same module attached - for example, dozens of living areas. That's why they have different sizes of living area modules for different sizes of ship. But with my visuals, having dozens of identical modules just makes your space station look more beautiful and lively, because each one is a bud that unfolds into a flower.

3) It creates a very distinct resource chain. Resources flow up or down branches. This is a bit different from having connected modules, because you can have dozens of modules connected to the same branch. But it's also not the overly-simplistic "global resources" approach many designs take, which means you can have pretty complex localization if you want to start designing trickily.

For example, you might want to do a heavy industrial starbase that squeaks by with minimal actual humans. Therefore, to optimize performance, you might have two industrial branches. One is on, one is off - and you can toggle it. So one branch gathers resources, using up all available workers. When your resources are full, you switch to the other branch and turn off this one, temporarily stopping the resource gathering so you can fire mass packets home. Then turn that off and go back to gathering resources.

4) It creates a very easy modularization: you can save branches, and later on if you want to create those branches again, just poke them into the same seed base. Like Japanese flower arrangement - hence the language chosen for our title.

5) Every flower withers, every journey ends. The visual is a good way to communicate to the player that whatever they've built is mission-specific and is not terribly permanent. This impermanence pushes the player to keep launching, and therefore also pushes the player to keep refining their designs. The network of relays, resource-gathering, and seed-launching is ever-changing.

It also gives us a lot of opportunities to design for longevity. For example, if you wanted a really long-lasting signal relay, the problem isn't the signal dish. It lasts hundreds of years. The problem is the humans. Human habs are medium-duration blossoms and therefore they'll die out long before the relay does. Nobody dies due to the the blossoms wilting, but they do go back into cryo sleep and stop working. Which means you have no workers to work the relay dishes.

You could use an expensive multi-generation mega-blossom, but that's incredibly pricey. Instead, maybe you build dozens of normal habs, but you only activate one at a time. You set up your station to automatically activate the next blossom when the current one is fading using a set of resource barricades on the branch itself.

This has the side effect of also forming a very nice "thermometer". The easiest and most robust design for resource blocking steadily works along the branch, so whenever you pop over for a look at your space station, you can quickly count how long it has left by looking at how many colorful buds remain unopened at the end of the branch.

Similarly, you can use it for rolling out in phases. Mining ships require a fair number of resources to build, and if there are several of them on a branch they will share the incoming resources evenly. This can lead to extreme delays in actually getting ships out: it may be better to use resource blocking to grow just one or two at a time and roll the fleet out in waves - otherwise your other blossoms will age pointlessly until you roll some ships out.

Anyway, I rather like this idea.


Verzor said...

Interesting idea which reminds me of solar/light sail ship designs. See for an example of real work being done on this idea.

Craig Perko said...

Yeah, the idea is somewhat related.

Chris said...

The most interesting part of this idea (for me) is 0.3c. I love the flower design, I dig the branches as a good mechanism for localising and compartmentalising complexity, but 0.3c is such an interesting idea that it sticks in my mind.

I did a quick check and found, which lists just under 150 celestial objects within a 20 light year (or 60 years of elapsed earth time) radius. Relativistic effects are reasonably easy to discount, as I found here:, which means cryogenic sleep and generally one-way trips. If we assume mass-accelerator technology (gravity slingshot, mass-driver, etc), then it suggests we can probably fling *some* resources back home (or relay them), but we'd need to set up a fairly hefty permanent emplacement to support it. Based on the transport delay, the design doesn't rule out interstellar mining expeditions but they would be more likely furthering future seed exploration rather than providing primary return-trip supply lines. For particularly rare resources an interstellar supply line would work (in theory) but there would be a major (generational) lag before the investment paid off.

This gives you a game timeline of around 500 years, to properly colonise a few dozen stars.

From here, the question becomes how the march of technology and progress effect the game over such a significant timeline. As a (bit of a) singularist, Moore's law and nanotechnology are mostly a given, rather than speculation. Fortunately, the laws of physics provide some hard-constants (energy requirements, transport time), but efficiency and computation appear relatively unbounded. Additionally, quantum entanglement could solve the FTL communication barrier, which enables greater synergy than simply resource sharing.

This all works in favour of your Ikebana design and theory, where most of the constraints you suggest are resource (and energy) bound.

Speculating wildly, part of me wonders if then the solution to human propagation is artificial body construction and conciousness transfer over entangled-link, but by that point we are all technically AI and the body part becomes kind of moot.

Anyway, just some interesting extrapolation from your idea.

Craig Perko said...

Well, in this design the technology is considered stagnant (at peak?), so the driving force is assembling resources and launch bases.

Some resources are very common: carbon, light, water. Others are very rare science-fiction materials that are rare enough that shipping them over a 200-year transit would make sense.