Jouni

The launch conditions for Soyuz TMA-22 were not really that bad. There was a blizzard, but blizzards in continental deserts are not the kind of "there will be 3 m of snow and the civilization will collapse" disasters you might expect near a major body of water. It's more like "it's uncomfortable outside, and you might get lost if you lose the sight of landmarks". First, commercial airlines routinely deal with precipitation under freezing temperatures. If it's cheap enough and safe enough for them, it should also be cheap enough and safe enough for orbital launches. Second, the wind conditions were described as "high winds". That sounds like Beaufort force 6-7 or 11-17 m/s, which is borderline for most rocket launches.

This is not a real entry, just a proof-of-concept. Never underestimate the power of the Twin-Boar. With a couple of drop tanks, a Twin-Boar can give the single-kerbal capsule and some other payload over 6500 m/s of delta-v. After a bit too steep ascent, I had around 3250 m/s left at LKO. Kerbin's orbital period is 1 year. If you launch into interplanetary space with an integral number of years as the orbital period, you will encounter Kerbin after a single orbit. I chose an easy orbit, where the encounter is at the periapsis and the apoapsis is 1 year after the launch. This required a bit over 1280 m/s for the launch, and around 2 m/s for fine-tuning the encounter at the apoapsis. I was left with almost 2000 m/s for managing the reentry and landing. That was necessary, as there is a lot of dry mass for the cross-section. On the other hand, because the Twin-Boar has a high impact tolerance, you don't have to carry enough parachutes for a safe landing. Especially if you manage to hit the ocean.

Tylo is my favorite destination. You need plenty of delta-v, a high-TWR lander, interplanetary transfers, some piloting skills, and so on. But in the end, Tylo is just like the Mun, but bigger. Unlike Eve, Tylo is a simple mission, where you don't have to refine your design with trial and error. This week I recreated the ship from my 0.90 Tylo mission, with some changes. While the old mothership used a Skipper and relied on aerobraking, the new design was supposed to complete the entire mission with nuclear engines. Symmetry is a funny thing. If you drop one booster above you and the other one below you, the forces acting on the booster are no longer symmetrical. The transfer burn started with the ship pointing directly at the KSC. The external fuel tanks lasted almost all the way to a low Tylo orbit. The large cargo bay is just the right size for a two-kerbal Tylo lander. The best time for staging is just before landing. Went there, planted the flag, didn't do anything useful. Tylo is no Kerbin. You only need vertical speed for avoiding the obstacles. It's perfectly fine to accelerate almost horizontally a few hundred metres above the surface. On the way home. If you launch directly from a low Tylo orbit, the return to Kerbin only takes a bit over 1000 m/s. Because I had plenty of delta-v left, I didn't wait for a launch window, but launched as soon as Tylo was properly positioned. A combined 1700 m/s plane change / retrograde burn / radial correction gave me an encounter with Kerbin one orbit later, and the insertion burn into LKO was no worse than when returning from Duna. In the end, the mothership had almost 2200 m/s left.

I consider MechJeb essential for information displays, maneuver node editing, and other quality-of-life improvements. Perhaps I could replace it with a bunch of other mods, but I have never bothered to try. Docking Port Alignment Indicator does only one thing but does it well. Without it, docking with multiple docking ports or in cramped spaces would be too tedious. Everything else is optional these days.

One more stock mission in 1.3 before reinstalling the mods. This time I went to Duna. Duna may be the best-looking planet in the solar system. A shallow approach, an inflatable shield, and plenty of parachutes. Somebody made a mistake in the assembly, and everything below the capsule was 90 degrees off. Including the ladder. The usual tourist shot. Bill discovered evidence of life on Duna. Almost home.

While waiting for serious missions to resume, I built another silly rocket. Rockets within rockets within rockets... I'm too lazy to calculate delta-v manually these days, so I wasn't sure whether Jeb had enough fuel to get back. Fortunately he did have more than enough.

While I was waiting for the mods to upgrade for 1.3, I decided to try KSP on my gaming PC. It ran much faster than on my old Mac, but the key bindings were wrong and graphics looked a bit weird with too much ambient light and/or brightness. For the first flight in 1.3, Jeb and the other usual suspects stole Baba Yaga's ship and landed it on the Mun.

Back in the day when air was made of pea soup and rockets were made of asparagus: We had no ISRU either, so the biggest challenge was landing something with 11 km/s of delta-v and high enough TWR in one piece. And not forgetting the ladders. Launch preparations used to involve fireworks: 18 boosters later, the optimal trajectory was still almost vertical: This was also an achievement in its own way: It was supposed to be a Tylo lander in the 6.4x scale system. The transfer stage / lander combination was launched in one piece, the Kerbin escape stage followed, and then I docked them using 7 large docking ports. Fueling the ship took 30+ tanker launches, and finally the crew module was launched on a tiny rocket. Unfortunately I grew bored with the project after the first test flight, and I never had the chance to try whether the nuclear asparagus lander would have actually worked. The project also involved my favorite launch vehicle: And my favorite launch picture:

A few things to consider: Payload size is more important than payload mass. If you can lift 50 tonnes in a single launch, you can lift 200 tonnes in four launches and assemble the payload in orbit. We already have decades of experience with that. On the other hand, if the payload is 10 m wide, assembling it will be complicated, if you can only lift 5 m wide payloads. This suggests that the core stage should be wide enough, perhaps 8 m or more. Using additional first stages as boosters seems to be a cost-effective way to increase payload capacity. While the Russians have a 4-booster design, most others have concluded that using two boosters is the best trade-off between capabilities and complexity. Hence a conservative design should only use two boosters. SpaceX has now reached the same point NASA reached a few decades ago. They have shown that partial reuse of launch vehicles is technically feasible. They do not know yet whether it is also reliable and cost-effective. While reusability should be a design goal, it should not be a critical element. The launch vehicle should still be viable, even if reusability turns out to be much harder than expected. So: Aim for cost-effectiveness and avoid unnecessary complexity; design a wide rocket to launch big payloads; increase payload capacity with two optional boosters; and plan for reusability but do not rely on it.

The docking port in my crew recovery ship always looked like it belongs in a space station, so I started building one. The required precision makes docking without RCS quite annoying.

I sent a mission to Dres. Solar panels are no longer useful that far away, so I switched from ions to nukes. Now that nuclear engines use only LF, I'm forced to use airplane parts, and all my designs tend to look the same. The usual stuff: deploying the lander and a tourist shot on the surface. Bob went to investigate the canyon. He had an accident. Bob fell all the way to the bottom, about 3.4 km below the landing site. As Bob was running out of propellant, Val had to take the lander down to the bottom. Launching from a canyon to the correct inclination can be tricky. Some serious aerobraking on the way home. And the usual crew recovery.

Yesterday my electric shuttle returned from Moho. Because the shuttle had no parachutes, I had to leave it in LKO. Today I built a ship to retrieve the crew. The ship may have been a bit overkill for 4 kerbals. I finally found good use for the cargo ramp. The ship behaved surprisinly well for its shape during reentry, and soon it splashed safely down in the ocean.

A couple of days ago I returned to KSP after a long break and built a simple electric shuttle for a Munar flyby. Today I decided to try where the shuttle can go with enough propellant. I also needed bigger boosters. I decided to depart immediately without waiting for a proper window. The shuttle was so heavy that I had to do the escape burn over three orbits. After a single course correction, I had a decent encounter with Moho. I did a quick orbital insertion and then departed immediately. Moho was ahead of Kerbin when I departed, so there was no reasonable way to return to home quickly. The escape burn put the apoapsis a bit below Eve, the first course correction raised it a bit above Eve, and the second course correction gave an encounter with Kerbin 1.5 orbits after leaving Moho. The insertion burn took a really long time, as the solar panels did not supply enough power that far away from the Sun. There was enough delta-v left to insert the shuttle to LKO without aerobraking. The mission took 1.1 Kerbin years, and total delta-v was about 17250 m/s. The ion engines are quite powerful now, as we have large enough xenon containers.

I decided to try KSP again after a long while. The first thing I built was a shuttle. Kind of. Destination: Mun Because I hadn't really played since 1.0.5, I wasn't sure how dangerous reentry would be in the current version. I did two aerobraking passes just to be safe. Unfortunately the parachutes were not properly balanced, and I lost some engines in the landing. After writing this, I remembered that the change of forum software was a major reason why I stopped playing KSP. I really don't like rich text editors.

Ion engines are perfectly fine for Mun landers. You use batteries as fuel tanks, and recharge them on surface. Six engines are enough for a lander based on the Mk2 lander can. Low-TWR landers can be a bit difficult to handle in the stock game, but it's definitely doable.

There is a simple test for determining whether the rocket engines are systematically too weak or too powerful: see whether they can lift a rocket that looks like a rocket. Real single-stack rockets are typically 10-15x taller than they're wide, with the newest models of Falcon 9 approaching 20x. To see whether the KSP first stage engines can lift such rockets, we can build as tall stack of fuel tanks of the appropriate size (2.5 m for the Vector) as possible without the sea-level TWR falling below 1.20. Here are the engines sorted from the weakest to the most powerful according to the height:diameter ratio of the stack: Rhino: 3.6 Skipper: 4.7 1x Vector: 6.6 Mammoth: 8.6 Dart: 8.8 Swivel: 9.8 Mainsail: 10.2 Reliant: 12.0 Twin-Boar: 12.9 2x Vector: 13.0 The Rhino, the Skipper, and the single Vector are quite weak, but the others seem ok.

The gap is caused by the lack of 1.875 m parts. There is roughly a 3x gap in thrust and a 4x gap in stack size between the Reliant and the Skipper, meaning that the Reliant can lift a taller rocket. A Mainsail-equivalent 350-400 kN engine might be useful, but it would make an 1.25 m rocket so tall that there would be problems with the joints.

That looks like 6.5 tonnes, so it's even heavier than the Mainsail. What I needed was something with more thrust than the Skipper and less mass than the Mainsail. The Skipper used to be my favorite engine, but 1.0 made it too weak for many purposes. 1.0.5 solved that problem with the Vector.

I don't understand this discussion. I'm a rocket guy, and while I haven't played 1.0.5 that much, it's obvious to me that the Vector fills a major gap in the 2.5 m engine range. For all intents and purposes, it's a 2/3 Mainsail. It's the main 2.5 m lifter engine, and it also looks better than the other 2.5 m engines. Before 1.0, the Skipper was that engine. A two-stage Skipper/Poodle rocket could lift a decent payload to LKO, and the same rocket with boosters could lift almost all 2.5 m payloads. Then 1.0 nerfed both Isp and sea-level thrust, making the Skipper too weak to lift almost anything on its own. For most 2.5 m payloads, you had to add boosters or use the Mainsail, which is too heavy and too powerful for most payloads. Now we have a 2/3 Mainsail. A simple Vector/Poodle combo lifts 15-18 tonnes, depending on the target orbit. As most of my payloads are about that size, I'm happy.

Somebody did it a couple of years ago, probably in version 0.23.5. I can't find it on the new forums, but the Imgur album is easy to find. Basically there is a huge flat area that can be used as a runway, and VTOL engines are necessary for both landing and takeoff.

Before 1.0, the LV-N was the best engine for reusable Tylo landers, especially considering that you could use the same engines in the transfer stage and the lander. With 1 tonne of payload and 5 tonnes of propellant per engine, you got a lander with around 6500 m/s of delta-v, initial TWR 0.85, and landing TWR 1.35. The payload fraction was 11.3%. 1.0 increased the mass of the LV-N from 2.25 tonnes to 3 tonnes. The same design still works with the new heavier engines, but the payload is only 0.25 tonnes/engine.

I fail to see the relevance of that point. Something complex may randomly evolve in an environment where such complexity did not exist before. The important word is "randomly". There is a categorical difference between a random process producing a complex result and a random process producing a particular complex result. Please think again. Do you really think that the following two scenarios are equivalent: A von Neumann machine arrives to a solar system and sets up a primitive ecosystem. Eventually, something complex (e.g. dinosaurs) evolves in the ecosystem and is able to reproduce reliably within the ecosystem. A von Neumann machine arrives to a solar system and sets up a primitive ecosystem. Eventually, the ecosystem is able to produce new von Neumann machines, which can be launched to other solar systems, where they can continue the work reliably. In the first scenario, evolution produces something complex that can reproduce within the ecosystem. In the second scenario, evolution produces a particular kind of complex system that can reproduce outside the ecosystem. Then define the laws of physics. In particular, is computational complexity a law of physics? There is a differerence between something being logically impossible, improbable (statistically impossible), and infeasible (computationally impossible). Something that is said to be physically impossible may be logically possible but extremely improbable. For example, there is nothing that makes perpetual motion logically impossible within the laws of physics – the laws that forbid it are statistical in nature. Physicists are currently debating whether some ideas from computational complexity should be considered fundamental laws of the nature – comparable to the laws of thermodynamics – or whether they are just practical issues.

That argument depends on a technically correct but rather useless definition of "ecosystem", at least for this context. I've seen a number of speculations on how life originally arose. In all of them, there was a complex self-sustaining system of chemical reactions (with an external source of energy), from which the first living organisms evolved. Technically it was not an ecosystem, but it was still a complex system necessary for supporting the primitive organisms. There is an important qualitative difference between those two cases. Evolution may eventually create something complex, but the results are essentially random. A von Neumann machine, on the other hand, must be able to produce a specific kind of complex systems reliably. In particular, each machine must create at least one replicate in the expected case. Just because something complex may evolve randomly doesn't mean that it's possible to build a self-replicating complex system that can move from solar system to solar system. That depends on your definition of "physical". By a simple counting argument, there is an unlimited number of things that are technically possible but extremely unlikely to develop, because the universe is too small and too short-lived. That may include many classes of objects that are conceptually simple but technically complex.

The primary system depends on the context. When the question in hand is whether a self-contained system can possibly self-replicate in a nontrivial fraction of all solar systems, the ecosystem is obviously the primary system. It's self-contained, while a living organism is not. And by "ecosystem", I mean something like "the transitive closure of the ecosystem", because we're talking about self-contained systems. Under this definition, the Earth probably has only one ecosystem. By "survival", I meant being able to reproduce and to establish a stable/growing population. Remember that the context is von Neumann machines, and the meaning of every word must be understood against that context. This is again context-dependent. A living organism is self-contained within the context of its ecosystem. Take an organism and place it elsewhere in the (local, not the transitive closure one) ecosystem, and it will probably survive, thrive, and reproduce. In this case, the context is the galaxy. We're talking about self-contained systems capable of moving from one solar system into another and self-replicating upon reaching the destination. I've essentially claimed that the galactic "ecosystem" doesn't look complex enough to support such self-replication.

Ecosystems developed as a result of random processes. Ecosystems may mix, expand, and become more complex, but they don't self-replicate. At least we don't have any evidence for their replication. Individual organisms replicate within ecosystems, but because they're not self-contained systems, they're not self-replicating in the sense of Von Neumann machines. So, essentially, ecosystems are the primary systems, while living organisms are just subsystems that can't survive without the complete system. When we're debating whether Von Neumann machines are possible, the relevant question is whether ecosystems can replicate themselves.