You are right that the no moving parts thing is speculation, but it's what I'd do. Several people have worried about disk failures and such like as a concern with the idea, and noise would also be a concern.

Regarding cabling, yes, you are right. In densely populated areas of the Netherlands there is probably fibre to the apartment building already, but they might have a low-networking workload in mind.

Simulations suggest that it is very sensitive to exactly where the gas giants form and the density of different parts of the dust cloud. Small changes in initial conditions mean that they may head in and stay there -- hot Jupiters; never head in at all -- hot super-Earths; or do what ours did and dive in and then out.

1. Thermodynamics: if you need to convert electricity to heat for any purpose you can get computation out for free. Electricity is very low entropy, low-grade heat over a large area very high, you can have the difference as useful computation

2. The article makes clear these are compute servers, not data servers or web servers. They may well be bitcoin mining, or running large-scale compute jobs for universities or the local met office or rendering a movie or ... In any event you expect a proportion of the servers in any job to fail. When you think they may have failed you restart the tasks they were doing somewhere else. Most of these tasks do not need much security either. There is little to gain by stealing or changing the predicted air pressure in a 100x100x10km block of air over Belgium next Thursday.

3. They are surely custom servers, not standard racks -- no moving parts. SSD for boot, application data over the net and a fanless design. They can be totlally sealed units entirely immune to junior's orange juice. Use mainly nonstandard form factors and they become basically unsellable reducing the theft problem and getting round some more security issues.

3. The article says that the supplier supplies power. Whatever cable they use for that can easily have a fibre built in for data.

4. Since this is cloud compute, it doesn't matter much if it gets turned off on rare hot days in the Netherlands, but if you care, pay the owner to open a window instead.

Have we any reason to think the water was actually ever liquid on Mars's surface to any great extent.
The young sun would have been cooler than todays. Could the water not have been present as ice, or perhaps as an ocean covered by a thick layer of ice.

I think the idea is this:
You have a large volume of clathrates underneath ice or frozen soil.
As things warm, they start to break down and a reservoir of methane gas builds up
at high pressure.
Eventually the pressure reaches the point where it can push aside or lift up or whatever the ice at its weakest point
and it finds a route to the surface.

Now you have a LOT of gas rushing through some kind of hole, a little bit like an oil well blowout and the gas flow erodes the sides of the hole and throws soil or ice into the air and generally starts to make a crater.

Furthermore the escape of all this gas lowers the pressure down where the clathrates are quite suddenly, so the breakdown accelerated greatly, providing still more gas to ruch up through the hole.

So not really an explosion, perhaps more like a blowout, but still fairly violent simply because of the amount of gas and the pressure.

At no point does it combust.

You need to find out about methane clathrates. They are very roughly a chemical compound of methane and water which is solid and stable at low temperatures and moderately high pressures (as found under a few hundred meters of water or ice, for instance. When they get a bit too warm, or the pressure drops a bit they turn back into methane gas and water. One cubic meter of clathrate released almost 200 cubic meters of methane gas, which then has to go somewhere producing something like an explosion. At no point did the methane burn (it was nowhere near any free oxygen until it got to the surface, it was just a gas pressure explosion.

This approach to government canbe seen working well in Somalia.

OK. I stand corrected.

Know is a tricky word, but there is plenty of evidence that most of the dark matter is not baryonic. The proportions of light elements formed at the end of the big bang gives a contstaint on the baryon density of the universe at the time, as do the ripples in the cosmic microwave background (which reveal the balance between radiation pressure and gravity in the early universe and tell us that most of the mass did not interact with photons at all). The bullet cluster is another piece of evidence. The stars in the colliding galaxies interacted with one another and with dust and merged into one bigger galaxy, but something, detectable by its gravitational lensing of galaxies beyond it, went straight through. It's hard to see how brown dwarves would have done that.

As I recall the best choice for a low-mass sunshield is a grid of fine conductive wires about 100nm or so apart in both directions. This forms a Faraday cage at optical frequencies. There's a complicated tradeoff between what you make the wires out of, how much is reflected and how much absorbed and how fine you can make the wires without them melting. I'm not sure what the winner is for this application, but the area density of such a material can be less than that of a carbon monolayer, since it's mostly holes, just holes too small for light to get through.

You can probably keep it on station without rockets by opening and closing flaps in the sail to manage light and solar wind pressure, although the control processing might be pretty severe.

You can use similar techniques to terraform Venus and Mars -- for Mars you make the "shield" into a Fresnel lens that actually concentrates sunlight.

I seem to recall third, but the stars that make and scatter medium-weight elements are big bright short-lived ones, so the first generation might only have taken 10 million years. There is some uncertainty about where the heavier elements (gold, uranium, etc.) come from. It is possible they are produced by a much rarer process.

You don't get relatavistic velocities with fission or fusion propulsion at reasonable mass ratios. You need antimatter for that.

You talk about ""science" --- the one with hypothesis, testing, reproduction of results". These things do kind of apply to cosmology. Hypothese are about things like the statistical distribution of galaxy sizes and redshifts, or the exact spectrum of the cosmic microwave background or the proportions of elements in the oldest stars or ... The speculators are working out these prediction so that the observational astronomers can test them with their next set of instruments. Or in some of the other areas, about what we will see in the LHC when we reproduce on a very small scale certain conditions.

Reproduction of results is harder, because we only have one universe, but people only become convinced of an explanation when there are multiple chains of evidence supporting it. So dark matter is supported by galaxy rotation, features of the cosmic microwave background spectra, gravitational lensing AND siumulations of galaxy distribution.

Actually it lasts only about a millisecond, but the 1 MYears of solar output part is right. It's about the mass of the moon converted to RF energy in
1 ms.

Constant approaching velocity is special relativity again, and again the velocities don't add the way you expect.If the planets in your example are approaching at 2/3 c they each see the other approaching at 12/13 c and they will very definitely and messily interact. Each exists for the other.

In this case acceleration makes no essential difference though. In either planets frame of reference there is an event horizon behind it (in GR acceleration and gravity are equivalent) but none in front of it, so they can see each other and interact freely.