K^2

It's not just filtering. Sun's spectrum is yellowish. But because peak of the Sun's output is right around the middle of our visible spectrum, which certainly isn't a coincidence, it's pretty close to white*. So if you were to look at the Sun directly through a filter that cuts out most of the light, but uniformly across spectrum, you'd see a disk that's white with a tinge of yellow. All G-type stars are yellowish to white, and F-type stars are white to blueish. So particularly hot G stars and particularly cool F stars can look almost perfectly white. I don't honestly know if there is a spectral type that's indistinguishable from white to human eye, or if it'd all look just a touch off, but at any rate, there are certainly stars closer to white than Sun. But there are also stars a lot colder than Sun and these that are way, way hotter. Cooler stars are often called red, but they're actually closer to orange, ranging from orange-red to orange-yellow. It's actually really easy to see what a light of red giant looks like, because light from an incandescent light bulb is very close in temperature to a typical red giant. And this is why talking about colors is a bit complicated. If you are sitting in a room lit by incandescent light bulb, you'd notice that it's a bit yellowish, but you'd think it's not far from white. While if you could actually compare it side-by side with daylight, you'd see how deeply orange incandescent light is. So if you ever want to picture what a world illuminated by a red giant looks like, you can just picture it illuminated by incandescent lights. If that's all you're getting, it actually wouldn't look all that alien. Likewise, there's a good example for what a blue giant light looks like down here on earth. A lightning generates plasma that's pretty close to the temperature of the typical blue giant's photosphere. The problem is that it's too bright and too brief for the color to register, and the gases involved typically shift the spectrum quite a bit. A bit easier to watch, perhaps, though certainly not directly, is an arc of an electric welder (always use eye protection!). Their temperatures are a bit cooler, but still, typically, in the ranges of blue stars. Unfortunately, most welding goggles distort colors quite a bit, but if you can find a set that doesn't, you could look at what the world illuminated by a blue giant would be like. Again, it doesn't register to our brain as particularly blue, and you'd call it bluish white, most likely. And if you're looking around a room illuminated by an arc welder (still through some sort of UV filter, always!), you might not even see it as particularly blue, because your eyes and your brain are going to compensate quite a bit. The blue tint of the arc will only really be visible in side-to-side comparison with a source of white light. And, of course, there are factors that play when light is too bright or too weak for our eyes to register correctly. Pretty much anything sufficiently dim will look yellowish and anything sufficiently bright can go from white to basically anything as it overwhelms your retina. So in all of the above, we're assuming that you're getting enough light to activate all the cones in the retina, but not so much as to overwhelm that, and that will necessarily involve looking through some sort of filter when looking at a star from up-close, or a telescope to amplify the light if you're looking from far away. Both of these can be achieved with minimal color distortion. * Ok, it's actually worth discussing what "white" means. Perception of color is a complicated thing. The most convenient definition of white is color you perceive when looking at so-called white noise. A spectrum that has equal power at all wavelengths. If you were to look at an artificially created white spectrum and compare it to sunlight, you'd perceive sunlight as slightly yellowish. Edit: I've actually found some code I used a while ago to try and display night sky from various places in our stellar neighborhood as precisely as possible. One of the things I tried to do is approximate the spectral color of each star. And the very first step was matching RGB output to be as close to what the eye would perceive. This involves figuring out the approximate spectrum of the star, and for this I just used black body radiation, ignoring chemical composition, then computing relative stimulus to the L, M, and S cones in the human eyes by that light, and finally, finding best RGB match that would produce the same stimulus. I found some papers that did most of the hard work, and was able to produce the following spectrum. It starts at 1700K, which is about as cold as a star can get and still be a star, and goes up to 25,000K, which isn't the hottest that stars get, but up there. To be clear, this is not what the stars at these temperatures actually would look like. It's the closest that the RGB display can produce, with pretty big error for the white-balance of your particular display, which I have no way of accounting for. (There are "correct" ways of tuning your screen to match this better, but they are not even possible on all displays, and pretty involved one ones that do support it.) So take this with a pretty big grain of salt. Nonetheless, it has the right features. The "red" it starts with is rather orange. The "white" area is pretty narrow, and not all that white. Clearly, most stars are going to be either yellowish or blueish. And the bluest it gets is kind of pale light blue.

Irrelevant chart, though. Imagine a world where this was backwards, and you were getting more energy from fission than fusion. And not just per atom, but per kg of fuel. You could still make a fusion bomb that's way more powerful, because there is a limit on how much fissile material you can put into a fission bomb due to criticality. There is no such limit for a fusion device.

If you just go by total yield, it's not even a competition. Even if we limit the discussion to devices that have actually been built, fusion bombs have over a hundred times the yield of fission-boosted ones, and theoretical is well into thousands times difference. There is just no reason to even test a device with several hundred megaton yield, which is why one hasn't been built, but all indication is that we could build one if needed. Fission-boosted devices are limited to sub-megaton range, and in practice, none have been built that are even close to a megaton. Now, if you're working with limited weight and size, that's where boosted fission can shine. Very often, you don't need more than 200kT per warhead, and boosted fission devices perform great in that range. Likewise, if you want to fit a nuke into an artillery shell or a backpack, you want to go with something small and light, which probably means a fission device.

This is actually another good example, yeah. The reason space-time warp for FTL is such an interesting topic for science fiction is because a) it absolutely works on paper, and b) while known limitations seem very restrictive, we don't know if they are real limitations. Or rather, exactly how far out the boundaries let us go. There's a lot of physics and math involved that I don't want to get into, but the main observation is that our universe is currently expanding at FTL speeds. So this whole "bend space to go faster than light" isn't working just on paper. It's happening. And there are a lot of reasons for why universal expansion is getting away with blatant violation of everything physicists claim is a strict law. The fact that universe is very big, filled with dark energy, whatever that is, and is currently in a process of rather violent explosion that we can't fully appreciate based on the time scales we're dealing with, see the being very big note, are all playing a major role here. Which is why we don't normally think of these loopholes as practical, but they're there, and even to a dogmatic scientist the speculation on what might be possible is too hard to resist. This is probably why hard sci-fi is so difficult to get right. Scientists are pants at writing and writers are pants at science. Go figure, right? And we kind of speak different languages in terms of what's plausible. It's very hard for a real scientist to say, "No, this is impossible." I mean, there are definitely cases where that's something you say, because it's close enough. But then overwhelming majority of things will be either called "likely", because there's existing principles and some lab work that indicates we can do this, or "unlikely", because we don't know how that'd work. And "likely" is great, because if you're working on near-future fiction, that's what you want to hold on to. Things that probably will come to be, unless we stumble on something better, but "unlikely" is a mixed bag. Because it's a very loose word and can apply to a wide range of plausibility. If you ever want advice on these things, you probably should be discussing it within context of a setting. Within a setting, anything is possible, but some things are still contrived. A flying wheel powering a starship is contrived, because if you can deal with materials of that super-strength and these levels of energy, a black hole reactor makes way more sense. But that's an extreme example. We like warp for FTL because it works. Hyperspace is worse, because while it's just as plausible, we see no evidence of it, so it's a bit like door to Narnia. Yeah, maybe we just haven't found it, but then if your story relies on that, then you just picked up a bit of "what if" fantasy feel. And maybe that's fine if it works with your setting! And then if your setting is sufficiently wild, then maybe you fold space through power of clairvoyance granted to you by a drug secreted by giant worms only found on one planet in a galaxy. And this would feel extremely contrived if that's all Melange did. But it's not. It's a central element to how much of society functions, whether they realize it or not, and drives the entire plot. It's hard for me to view sci-fi purely through eye of someone unfamiliar with technology and science behind it. But I think readers get the feel for contrived inventions in bad sci-fi purely through realization that something that powerful doesn't exist in isolation. This is particularly relevant now, when a lot of people remember what life used to be like before smart phones and even before internet. When you introduce a piece of technology, if something it relies on is almost magical, then it will change all of society. And make no mistake, modern processors in phones and laptops are absolute magic from perspective of actual computer manufacturers from 50 years ago. They would tell you that transistors smaller than 100nm are flat impossible, and at any rate, heat generation from a chip that runs at several gigahertz and does trillions of computations per second would have to have a building dedicated to cooling it. So I think, people now better than any time in the past understand, at least subconsciously, the consequences of "magic" technology. If you have to come up with impossible material or drive or energy source, and that's all it does, it will feel contrived, even if it's not that implausible outside the setting. And on the contrary, you can take something absolutely bananas when looked in isolation, stick it in the setting that acknowledges all of its impacts, and you have something that flows, and you still have the choice to go soft or hard on the science in the fiction. For a contrast, consider Dune vs Mass Efect. Dune's Melange and Mass Effect's Element Zero serve fairly similar purpose of explaining a lot of wizardry in the setting. At the same time, Dune is basically a space-opera, despite being pretty descriptive in places, and Mass Effect has elements of hard sci-fi in it. Anyways, I'm kind of starting to stray off on a tangent from what I was trying to say - setting matters when discussing how well an explanation for something works, and without knowing the setting, it's hard for me to say how contrived something is. Best I can do is point out what sort of "magic" changes to our understanding of technology it would take to make it workable. And sometimes, it's pretty clear to me that these changes would make the task itself obsolete, and I'll call that out. But sometimes, I just don't know if it works without knowing what else is different about your world.

Energizing magnets usually takes a bit of time. Again, you can kind of think of it as a pressure vessel. It's not easy to instantly fill a vessel with pressurized gas without risking blowing something up. There is a lot of energy and very strong forces involved in both cases. But it is possible. Even superconductor magnets can be "turned" on and off. Admittedly, "turning" them off is a lot easier. The process is known as quenching; the coil stops being superconducting, and pretty rapidly dumps its energy as heat. The hardest part is not letting magnet melt down in the process. In terms of technology allowing us to overcome these difficulties, well, that's a complicated topic. Lets use an example with something way simpler than a magnet. (If we gave out prizes for deceptively complex topics, magnets would take my first place nomination with no competition.) Think about using a flywheel to store energy. It's amazingly simple and efficient. Humans have used flywheels since, well, pretty much since they've known what a wheel is. By the time of ancient Greeks, they've done some nifty things with flywheels as temporary energy storage, and it kept finding uses in clocks, trains, gasoline engines, jets, rockets. Modern flywheels can store energy for long time with conversion factors far exceeding these of chemical batteries and they're still cheaper than a battery. So why don't you have one in your phone? Well, it's all about energy density. In order to increase amount of energy a given weight of flywheel can store, you need to spin it faster. And at some point, no known material can survive the centrifugal forces involved. Can we make stronger materials? A bit. But there are limits to how strong forces between atoms can be in principle, and that's all we get to build on. So lets say you're writing a sci-fi story in which humans are using flywheels for energy storage. If you write a story where we're using them to store energy from solar panels for use at night and to balance grid demand, you're writing near-future sci-fi with very plausible setup. It's probably one of the methods we'll use for storing renewable energy in the near future, unless we come up with something drastically better. If you will write about airliners powered by flywheels made out of spools of extremely strong fiber, you're writing something that's still hard science fiction, but much more speculative. We have fibers that should potentially allow for it, but we have no idea how to make them in sufficient lengths and sufficiently cheap to replace jet fuel. And odds are, we will come up with something better by the time we have this kind of tech. It's sort of a space-elevator kind of tech, and would, in fact, require similar materials. But it's not implausible. Finally, if you write about personal orbit-capable shuttle size of a car that are powered by a flywheel, you're in Star Wars territory. This would take materials stronger than maximum limits on chemical bonds, meaning, you're dealing with the kind of unobtainium where you might as well be making up whatever you want. Saying that this magical power source is based on a flywheel at this point is contrived, and unless you're writing a comedic space-opera, I wouldn't recommend it. When we're talking about building absurdly powerful magnets, we're dealing with similar kinds of limitations. We're not just talking about limits of technology, but of material properties. We can push these a lot further than what we can practically make now, but these limits aren't infinite. Past a certain point, it's not just that we don't know how to make these things. It's that if we figure out how to make materials tough enough to handle it, we'll have better ways of dealing with the problem you're trying to solve. Way, way better ways, which if you want to write into a sci-fi, you might as well treat as magic, because it goes outside of our core understanding of physics. At that point, anything you come up with might as well be magic, so there is no reason to try and explain it with magnets. Unless, of course, you're intentionally going for that Joules Verne style of sci-fi that's so naive, it actually becomes kind of charming.

Note, it's still magnetic field. Material is a diamagnetic, and the effect is called diamagnetic effect or diamagnetism. But nothing special happens to the magnetic field itself. It's all about how matter interacts with said field. And if you figure out how to generate and maintain said field, maybe? I haven't done the math. But like I said above, making exceptionally strong magnet that's not tearing itself apart is very hard. The stronger the field, the stronger the effective pressure that's trying to make your magnet explode. Even modern superconductor magnets require a support framework that doesn't contribute to the magnetic field, but just to the mechanical strength of the magnet, and what you're suggesting would require magnets several orders of magnitude stronger. The energy density, which is equivalent to mechanical pressure, of the magnetic field is B²/(2μ). So a 1.5 Tesla magnet of a medical MRI tomographer is already generating nearly 9 atmospheres of effective pressure inside the magnet that has to be supported by magnet's walls. Strongest magnets I've worked with were 14T, and despite having a bore barely large enough to fit your hand through, the magnet itself is too heavy to lift because of the amount of superconducting wire and the structural framework required to maintain such a field. Not to mention the cooling tanks of liquid helium and liquid nitrogen that actually house the magnet.

Your solution is equivalent to "put some padding in the front". The ship still has to accelerate all the air in front of it, which causes rapid deceleration of the ship itself, which destroys the ship, let alone killing the crew. How you push the air is absolutely irrelevant. Air doesn't have the time to get out of the way, so you're pushing, well, not quite everything you sweep up, but a significant fraction of it, which might be as much as a few tons of air per square meter of your craft's cross-section. Conservatively, you'll lose a few km/s of your speed in under a second as you're going through dense layers. Technically, you can still send cargo to space by shooting it out of a long cannon. There are components that will survive the deceleration of impact with air. It also happens so fast, you barely need shielding. For re-entry, you have to hit atmosphere at a fairly shallow angle to actually give the ship time to decelerate and not kill the crew with G-forces. Because of that, re-entry heat is generated close to the ship over considerable time, and you need significant shielding to prevent your ship from burning down. That profile is the opposite of what you want when firing something into space out of a cannon, though, so instead, you make the angle steep enough to punch through atmosphere in seconds. No crew to worry about, and most of the heat gets generated much further away from the craft due to much thicker shock. You still want something tough and heat-resistant out front, but not a whole lot of it compared to what you want for re-entry. So magnetic shielding for launch is pointless thrice. It doesn't help with the actual problem, it's not needed in cases where we could still use such launch system, and it would be entirely too heavy to launch even if we could build magnets that strong. Magnetic field has pressure, and maintaining that much pressure would require very thick "walls" indeed.

Art of War has lengthy passages on logistics. Also, deception around logistics.

Don't underestimate the power of the brute force methods. There are some wild trajectories out there, but you rarely find something good near these, and if you did have such a mission profile, one little error would mean a total loss. So all of the practical missions are found in fairly large "valleys" of reasonable solutions. Once you constraint where you're coming from, where you're going, and how many transfers you want to make, there are not that many initial conditions you have to try before you can just gradient-descent the rest of the way. Some general ideas, like Earth-Earth fly-by used by Rosetta sound silly until you do the math, and somebody must have had a stroke of genius to suggest them initially, but once you know the kinds of "points of interest" you might encounter, iterating through all the possible combinations doesn't really take all that much computing power. The hard part is actually coming up with a "map" of the Solar system that lets you integrate motion precisely enough that your answers aren't total nonsense. Computing optimization problem for one ship, even with all the iterations and false starts that any search will have to do, is nothing compared to building up the map of every single object of any significance, as well as reconstructing average contribution from all the other orbital junk. And the truly amazing thing about all of this is that NASA is sharing their data via convenient interface. JPL Horizons provides positions of the Sun, planets, over 200 moons and planetary satellites, thousands of comets, and nearly a million asteroids. With this data, you can actually plan missions on your home computer, provided you have solid enough grasp of numerical methods or have software that handles that, of course. Once you have a mission plan, you still have to execute on it, of course, and that's another matter entirely. Space telemetry is hard. I'm sure a lot of otherwise good mission plans end up on the cutting floor because they require too much precision in execution. A mission to "maybe make a crater on a moon of Jupiter, or maybe miss it entirely," is not going to get a lot of backing even if you promise to cut delta-V requirements in half.

Depends on what you call a stable orbit. Take a look at Saturn's moons Janus and Epimetheus. They essentially share an orbit without actually having a fixed orbital relationship to each other. This appears to be long-term dynamically stable, even though the orbital parameters of each moon constantly change. Specifically, one moon starts out in slightly lower orbit, until it catches up with another, at which point, gravitational interaction causes them to trade altitudes. And so the moons basically play tag with each other without colliding or causing instabilities.

Because eccentricity of Mars' orbit is quite a bit higher than Earth's, there are transfer options that give you significant time on the surface with almost no delta-V budget increase. But your transfer windows for these might be even fewer than once every two years. In practice, almost any convenient mission to Mars is going to be noticeably worse than optimal in terms of fuel requirement, but it's probably worth it if you're sending humans and not just cargo or experiments.

Pretty consistent. There is no differential heating due to continents, since all the solar energy is absorbed by opaque clouds, and the Coriolis effect is weaker than on Earth. So while there's considerable wind speed in the cloud layers, the shearing isn't as big of a problem. There's some weather, so some amount of steering might be necessary to avoid regions of turbulence, but with good forecast, it might be enough to change elevation. I'm sure there's a limit to how big these can get anyways, but you don't necessarily need these things to be enormous. A square kilometer at city population densities can be in tens of thousands of inhabitants and still have room for infrastructure. Furthermore, you definitely want these to be modular, so there is an option to decouple in case you can't avoid particularly turbulent area. And yeah, at relevant altitudes, there's effectively just the equatorial jetstream. Things get more exciting at the boundary with thermosphere, but that's too high for cloud cities, anyways. At the relevant altitudes, the atmosphere is super-rotating, which means the cloud city will be "orbiting" the planet. Location is dynamically unstable, and left to its own devices, the cloud city would drift towards polar vortices, which would be bad. However, you can utilize the thermals at the equator to actively stabilize the cities there, giving you most stable weather conditions and a good location for launching to orbit.

The main obstacle with colonizing surface of Venus is that without building extensive infrastructure at the cloud layer, trip to Venus is one way. Yeah, we have materials that, in sufficient quantity, will provide adequate protection and can be cooled relatively cheaply. Nothing heavy enough to withstand the pressures of lower layers is leaving that planet, however. In order to connect Venus to the rest of Sol, you have to have a launch complex at the cloud layer, and in that case, there is zero reason for your main population to be locked in tin-cans at the surface. If mining operations on the surface are important, it might be necessary to have some human presence at the surface, but travel between cloud complex and ground stations is actually pretty straight forward. Blimps with bathyscaphe-like crew compartments are entirely viable for surface-to-cloud transport on Venus. Without building cloud complex first, who's going to sign up to go to the surface, knowing that they have no way of coming back up, and they are forever locked away in tiny hab units surrounded by environment that will kill the occupants instantly if there's ever the tiniest of leaks. Sure, the later is also true of any orbital habitat, or something on an airless rock, but at least you go to these knowing you have at least a chance of leaving it at some point in the future.

Might be easier to build floating structures filled with breathable air, which is lighter than Venusian atmosphere, and put plants or artificial CO2 processing factories there. There's an altitude that should be quite comfy for long stay, and anything you can't get there can be acquired at lower altitudes by autonomous systems. You probably couldn't get any metals, so you'll have to get really good at recycling these. But everything you need for carbon-based life you should be able to get, and you can use most of that for a range of building materials. So long as you get occasional shipments of things you need for electronics, it should be possible to adapt the planet for long term stay. And sure, maybe over a long enough time, that would allow us to bind enough carbon to make surface survivable.

Every single orbital maneuver is more efficient if you perform it in less time. Landing and takeoff from airless body are prime examples, as their efficiency is directly governed by available TWR. It also opens up a lot of options in reentry. If your crew can survive a belly-fop against atmo, you don't need much of a heat shield, and the amount of time spent in reentry, where craft is particularly detectable, vulnerable, and often cut off from communications, is greatly reduced. But you shouldn't underestimate dodging a missile, either. Unless the kill vehicle heading for you is surrounded by thrusters, its ability to accelerate in transverse direction might be quite limited, allowing you to dodge. You also shouldn't underestimate the benefits of abandoning ship if you can't dodge. Being able to eject at really high accelerations in just a fraction of a second can make a difference between losing the ship and losing the ship with all hands. We're also not limited to submerging crew in fluids when we're talking about short spikes. By far the largest difference in densities, once you fill cavities, is between tissues and bones. If you don't mind additional supports surgically added to your skeleton that get attached to the ship, you can significantly increase the high-g tolerance, possibly well into hundreds of gravities. It'd break every bone in your body and skewer your flesh with shards of bone if you tried to move during a maneuver, so it's definitely not good for sustained, but might be just what you need for some of the maneuvers discussed above. We can also add mag fields to provide fairly uniform force across your body via diamagnetic levitation effect. What we can apply continuously will only give you a few Gs, and you couldn't move much because the field can only be made uniform for particular body configuration, but we have ways of generating spikes of magnetic fields that are much stronger, possibly giving you tens of Earth gravities with minimal setup. If you're prepared to make some surgical modifications, this can be greatly extended. Possibly also pushed into high double digits or low triple digits. Combining the two approaches above can push you into territories where structural integrity of the ship and equipment is starting to become as much of a problem. Finally, gravitomagnetics is not quite as sci-fi as it might sound once we get to sufficient scale. Brief spike of artificial gravity induced by a sudden change in gravitomagnetic flux, such as discharge of a capacitor in external magnetic field, can provide high force uniformly across every atom of crew and compartment with minimal tidal distortions. Now the key word here is brief, and the system needed to generate this will be rather large, as are energies involved. So I can't imagine this being useful for absolutely anything but an emergency escape, but that is a plausible way to generate accelerations that go way beyond ability of even structural materials of withstanding the stress.

Something that operates on longer wavelengths might use a pigment with metastable excited state using a double-photon absorption to get sufficient energy. There are known organic dies with fluorescence in near IR range that would be perfect starting point for such an energy-capturing molecule. Furthermore, if it has double captures in 800-1200nm range, it would imply another absorption band with single capture in the 500-600nm range, leading to a magenta appearance of the plant in the visible light. That's just one out of many, many possibilities. Most Earth plants have multiple pigments, which are effective at absorbing reds and blues. But far from all. We have pretty good variety even down here. It's hard to rule out any possibility for a plant living on another planet somewhere.

Unlike sustained high-G, the uses for spikes in high accelerations are going to be useful for absolutely everything. Without a limit, it even makes lithobraking a viable landing option. It's much easier to summarize when you don't need high-G, and that's pretty much covered by, "When it's sustained."

I'm surprised nobody mentioned asparagus staging in this thread yet. Also in before announcement that to make deadlines, this will be FH payload.

Even then, it's hard for me to imagine scenario where you'd need acceleration greater than 1g desperately enough to sacrifice mobility around the ship for the duration. A 1g torch gets you to Neptune in just over two weeks. At 100g, it'd still be almost a day. Two weeks on a cruise ship or a day in a vat of fluid? I don't think I'd opt for later, unless this is a time-critical mission. And on more common inner system routes, high acceleration would make even less sense. Likewise, if we go interstellar, relativity kicks in. A quirk in how proper acceleration works at relativistic speeds means that both the Earth time and ship time aren't impacted all that greatly by acceleration. A 1g tocrch is perfectly capable of running to the center of the galaxy and back within lifetime of the crew. Increasing acceleration by factors of 10 buys you back a few years from that. Exponential increases in acceleration give you linear improvements. Even by the time you get to 1000g, it doesn't make a huge difference. And in terms of time in Earth frame it doesn't make a difference at all. The round trip will be 50k years of Earth time no matter what. So I don't think sustained high g is practical unless we can make it work like in sci-fi, where you still get a comfortable deck to walk around the ship.

Which isn't bad, but not groundbreaking. It lets you bring more equipment on a similar type of mission, giving you either more mobility on the Moon or allowing for longer stay. If we were building a permanent outpost on the surface, that kind of boost could even be critical. But this doesn't instantly open up doors for completely new missions. Lighter engines and tanks on top of the ISP boost does. A similar sized rocket could be used for a direct Mars fly-by mission with room to spare. And we are talking about getting a sufficiently large hab unit to get the crew there with their mental faculties intact. With two launches, we might seriously discuss a landing mission. That's not something you could have done with original Saturn-V without multiple launches to assemble an interplanetary rocket in orbit. I know there were proposals for something simpler and more direct with Saturn-V, but they weren't really taking into account things we learned about extended stay in microgravity since then.

ISP difference isn't huge for modern LH2 rockets. There have been some improvements, but if that's all we had, it wouldn't be much. Materials that would let us squeeze all available ISP out LH2 without melting just don't exist and probably can't exist. We can do 10-15% better, and again, that doesn't make as much difference as it does for kerlox, because the g*ISP factor for LH2 is already a large fraction of delta-V. So you don't get these exponentially growing benefits of improving ISP like you do with kerlox. But when you couple that with better TWR and MUCH better weight to capacity ratio of tanks we can build with modern materials, the net performance can be significantly better. Switch to kerlox or some other hydrocarbon makes sense if you're trying to optimize for launch cost. Which, you know, is probably the sensible thing to do. But then, yeah, you really are building a very different rocket.

Yeah, the golden section search can be used to look for local minima or local maxima. You just have to flip the inequality used to decide whether you pick the left node or the right node inside the iteration. Likewise, the outer loop compares results from golden section for each arc segment to find the global optimum. You'll need to flip that inequality as well. And I don't recall how I initialized the initial best distance. Given that we were searching for maximum, it was probably set to zero. If you're looking for minimum, you should set value to something greater than theoretical maximum. Sum of semi-major axes is absolutely guaranteed to be greater than actual minimum separation, so it's a good choice.

Tank pressure is often required for liquid fuel rockets to maintain structural integrity. That's not really anything new. In fact, you can see one of the Starship prototypes collapsing due to loss of pressure in lower tank during testing. It has enough rigidity to support its weight while empty, but it can't support weight of the filled upper tank without pressure in the lower tank.

Unlike liquid fuel rockets, the most complex and expensive part of SRB is the fuel. There are parts worth recovering, but these can already be recovered by parachuting the boosters down. And because boosters are inherently non-structural to the rocket, and solid fuel helps support its own weight, there is absolutely no benefit to pressurizing them like the liquid fuel tanks. Anything you could gain by making the structure inflatable you simply don't need, so it'd only be making the booster more expensive.

It's not just NASA. I think this is a common trope of experimental physics. You should see what particle physicists and cosmologists come up with. It's sufficient to point out that two (arguably) best candidates for dark matter have been known as WIMPs and MACHOs. If you didn't know otherwise, you'd probably think these came out of Marvel comic books. It's almost as if nerds who went into hard sciences grew up reading these, or something.