Hmm this made me wonder why something like this wouldn’t melt the rock and then sink into the crust and then into the planet. Probably not hot enough.
And that made me think if we could build something like a big pellet of fissile material, encase it in tungsten or something so that it is hot enough to do so but remains stable, and then let it sink into the earth. Maybe that could be tracked? Then we could learn something about how it moves and where it ends up. But probably can’t be tracked since this isn’t star trek 🖖

Still interesting. zstandard using pretrained dictionaries (zstd --train). Previously they used zlib.

Does anybody know if you can pretrain dictionaries for 7zip for large ebook libraries? Or any better compression library for text?

Now make a square out of squiggly yarn

String theorists claim this is the true shape of spacetime!

Interesting. You could design a robot arm where most of the surface can transmit and radiate heat away from the motors.

PS: Oh I just found out the design I’m thinking of is “frameless dc motor” or servo so you can route cables internally and have aluminum tubes on the outside for heat transport (“When high temperatures are a concern, the machine structure can provide effective heatsinking”). And now that I know the name they are actually not that expensive on aliexpress! Something like this as a servo with CAN and 48V connection and flanges to directly attach lightweight ~15cm tubes would make building a robot arm pretty easy! EDIT: Ok basically you’d just need a lathe to make a housing with flange and to fit a bearing.

I think you need to compress the air and cool it and you get condensation. I think the advantage of baking the soil would be that the concentration of water vapor would be higher so less energy.

Hmm… maybe the smartest way would be to just build greenhouses and pump compressed air into it. Plants grow, absorb CO2, release oxygen and water. At night you lower insulation and condense excess water out of the air. What is left is a surplus of oxygen. Of course that requires a ton of glass.

I dabble a little (3D printers, composites). I’m thinking more about 30-100 years for actually living on moon or mars like a real self sufficient colony instead of missions.

I do believe mechanical engineering can be represented in a way that is accessible to AI. It doesn’t need general knowledge or sentience. Not just the generative stuff we have now in CAD, but more like the AI that can generate images. Ways to generate infinitely complex “nonsense” machines isn’t hard, but AI can then learn of how to work more intuitive.

At some point we’ll have an AI that can design, supervise manufacture / CNC and robots to assemble and then control all the machines we need for this. Of course the design would have lots of human input at start. But robotics are improving and deep AI is able to do remarkable things already. Eventually we would feed all the existing CAD designs and function and physical laws and material science into an AI as a lot of data and then suddenly you can bootstrap an entire industrial society fully automated. But every step and progress towards that goal would make it that much cheaper and easier to do.

Of course that’s sci-fi but so is the idea of living on mars :)

True, and you have more gravity. On the other hand you have perchlorides. Obviously you need water but I thought they did find some already near the south pole on the moon? (EDIT: Yes but apparently little and not very concentrated). I didn’t know there was so little carbon though.

Well yeah, then underground in lava tubes. But then the moon would be much easier for building habitats to live in.

There is a sort of paradox. To do any of this in any scale would be much more likely with automation. And automation, like robots being able to build robots and then factories and anything else is coming especially with the recent advances in AI and computer vision etc. And then such megaprojects won’t be unthinkeable any more. But then it would probably also be easier to build a mega space habitats mining the moon. Without automation, it might never be worth it to actually colonize there so it’s better to wait a few more decades.

Interesting video. The increase of efficiency of solar panels and the constant sunlight gives them like 100 million in cost savings over 10 years for a 5GW center. Pretty impressive.

It would also be cool to buy storage space for fileservers or seed boxes in space. Probably not very practical but cool haha. Would also not work well with sun-synchronous orbit.

Yeah we know whenever Musk opens his mouth he’s lying. So we need to expand regulatory power because people like him doesn’t give a shit.

There is Spacecraft_radiation_challenges.

I believe simply using enough water would be perfectly fine for shielding. The problem is using enough of it and the extra weight.

In Stowaway (2021) - IMDb they have a special room that is shielded using electromagnetism.

You’d probably need to grab ice balls rich in nitrogen ice from the kuiper belt and somehow hurl them on a curse to impact mars. And do that a lot. Or maybe build a railgun on Triton to shoot nitrogen bullets at mars

The wikipedia article mentions the idea of superconducting rings build around mars surface. They could also be used for energy transport. This is basically already possible with current technology.

Maybe it’s also possible to somehow extract energy from the solar wind but that is only speculation on my part.

Yeah The Expanse is awesome but they are mining rather big ice chunks. In the Starcarrier series they find a giant shipyard churning through protoplanetary disk. Which sounds kind of ideal for mining raw materials for e.g. space habitats. But I guess a low gravity moon or asteroid would work just as well.

You could probably just roll out big black tarp and connect to a vacuum. The black tarp would increase heating from sunlight and the vacuum would press the tarp to the ground and suck up the air. Then during the night you roll up and prepare for the next day.

Basically a rolling robot that rolls the tarp up, then continues rolling, then reverses and rolls the tarp back out.

But it would probably be better to just condense the water out of the air at night at a higher volume.

What if you had a giant vacuum nozzle at the front and shoot the ice particles out the back of your rocket? :D

Hmm that might actually not be so stupid. First there must be some gas pressure since water ice would sublimate otherwise, so that actually should help gather and guide particles.
Then if you would go slow enough the forces should be minimal and you could mine the water ice this way. Or extract solids and turn the water into big blocks of pykrete to reinforce the nozzle. Or use the blocks of ice to build a space habitat. Or accelerate waste out the back to accelerate again. You’d only need a reactor with enough power.

mining industry / [LTIFR of] 5.5 is middle of the road for manufacturing and construction, generally, but that includes all sorts of manufacturing, from building houses, to steel foundries , to making cars

They work in clean rooms and should have far more controlled conditions and a better safety culture than those industries. This is not good enough.

Also “middle of the road” is concerning, considering that this stuff is going up into space.

That photo is amazing, I haven’t seen a lava tube on the moon before! It’s great that China is pushing ahead with this.

I wonder how you could make lava tubes airtight. If it’s hardened stone or even iron, I imagine you could just leave most of the surface as is or seal it with some kind of resin. And for cracks and gaps you could “3D print” on the surface to create a seal. Not sure if or what 3D print plastics you could print in vacuum. All this could theoretically be done with robots.

They just need to find a way to transport and land enough materials on the moon. Maybe a railgun that can be used in reverse for cargo delivery.

What’s ABI? API compatible with linux kernel? EDIT: Never mind, Application binary interface

And what’s TCB?

I think it’s a problem with VPNs