egg changed the topic of #principia to: Logs: https://esper.irclog.whitequark.org/principia | <scott_manley> anyone that doubts the wisdom of retrograde bop needs to get the hell out | https://xkcd.com/323/ | <egg> calculating the influence of lamont on Pluto is a bit silly…
<queqiao-> ⟨Starcatcher⟩ ⟪PrinceZola ùwú⟫ can you tell me more about those two […] ⮪ ksp is very bad in handling memories.
<queqiao-> If you run KSP Allo fix and run it, the program will automatically do it for you.
<queqiao-> For mem reduct go to settings and tick the autoclean box. be sure to turn off notifications.
<queqiao-> ⟨Starscraper⟩ ⟪egg⟫ I think you are confusing multiple […] ⮪ Hmm. I'm sure there's some way to make a simplified form that's a lot more tractable, though, at the expense of precision - which is fine in moderation for a gaming application.
<queqiao-> I remember one textbook worked through the Schwarzchild solution without going beyond mathematics covered in Calc 2, and you did a fair bit of trajectory stuff in that - even touching on the Kerr Metric.
<queqiao-> ⟨lamont⟩ https://www.amazon.com/gp/product/020138423X probably
<queqiao-> ⟨lamont⟩ i took that as a course back in '96 at UW with Taylor
<queqiao-> ⟨stone⟩ bro my principia be bugging
<queqiao-> ⟨stone⟩ can't open flight plan, tolerance and steps are gone
<queqiao-> ⟨stone⟩ already tried reinstalling it, doesnt change anything
<queqiao-> ⟨stone⟩ nvm
<queqiao-> ⟨expanders are superior⟩ for GEO in prinipia, is it recomeneded to use the time to AN?DN to plan ap raise burn?
<queqiao-> ⟨面条参考系⟩ I'd look at the LAN in the flight plan analysis
<queqiao-> ⟨Starscraper⟩ ⟪lamont⟫ https://www.amazon.com/gp/product/020138[…] ⮪ Yep! That's it.
<queqiao-> ⟨itsRyan⟩ ⟪expanders are superior⟫ for GEO in prinipia, is it recomeneded […] ⮪ That works. Also, for the RP-1 GEO contract, mechjeb maneuvers are close enough.
<queqiao-> ⟨egg⟩ ⟪Starscraper⟫ Hmm. I'm sure there's some way to make […] ⮪ Please stop being sure of things.
<queqiao-> ⟨lamont⟩ lol
<queqiao-> ⟨egg⟩ Studying properties of a specific problem is an entirely different business from simulating what happens in the very same problem, let alone in a broader category of problems.
<queqiao-> ⟨egg⟩ And baseless speculation is not how you should approach numerical analysis.
<queqiao-> ⟨sichelgaita⟩ ⟪Starscraper⟫ Hmm. I'm sure there's some way to make […] ⮪ This discussion is becoming a bit pointless. The source code of Principia is out there on GitHub, why don't you clone the repository and try to change the force computation? I did that years ago to see if I could "fix" the Mercury perihelion. I failed miserably because the first thing you realize is that you have to be extremely cautious about the coordinate...
<queqiao-> ... system that you use. Slapping an 1/r³ term in the potential just doesn't lead you anywhere. The Schwarzchild solution is valid, in, well, the Schwarzchild coordinates. I quickly gave up when I realized how hard it would be to support GR coordinate systems in Principia.
queqiao- has quit [Ping timeout: 192 seconds]
queqiao- has joined #principia
<queqiao-> ⟨egg⟩ [testing if the bridge is up]
<queqiao-> ⟨egg⟩ Besides the reference frame questions, while we could in theory, with great effort (the equation becomes a DAE which is a lot more complicated), « fix » the force computation to account for small relativistic perturbations using Einstein–Infeld–Hoffmann, these approximations do not lead to the qualitatively different and interesting effects (lensing, frame dragging, gravitational waves, what have you).
<queqiao-> ⟨egg⟩ For those you would need a completely different approach: wave propagation isn’t something you can model with ODEs whose only state is the motion of a handful of point particles.
<queqiao-> ⟨egg⟩ I am not opposed to having a discussion of GR here, but please stop trying to pretend it is possible to shove that into Principia.
<queqiao-> ⟨Starscraper⟩ ⟪egg⟫ And baseless speculation is not how you […] ⮪ It's not baseless, but there's nothing that does not begin, at first, as speculation. And I must respectfully state that no, I will not stop being sure of things. Especially not when used as a figure of speech that is motivation for researching something more deeply.
<queqiao-> (It's not baseless because typically, you can exchange modelling accuracy for a model that's simpler to work with, and given the game application here, a fair bit of accuracy can be sacrificed. Newtonian gravity itself can be seen as a more tractable form of GR that is less precise - ie, GR reduces to newtonian gravity with certain assumptions)
<queqiao-> As to the Einstein–Infeld–Hoffmann equations - perhaps they don't lead to frame dragging, but frame dragging was first predicted in a weak-field approximation - ie, linearized GR, in the case of Lense–Thirring precession. https://en.wikipedia.org/wiki/Lense%E2%80%93Thirring_precession
<queqiao-> And Post-Newtonian expansions were _how_ the precession of Mercury's orbit was calculated. EIH was just the first of the whole field of Post-Newtonian Expansions that's matured a lot over the last century.
<queqiao-> ⟨Starscraper⟩ +a statement of
<queqiao-> ⟨egg⟩ If you keep saying things with an air of certainty that conflate very different numerical methods and approximations the nature of which you evidently fail to understand, you will achieve little more than annoying us.
<queqiao-> The #rules apply here, and in particular #5.
<queqiao-> ⟨egg⟩ The interesting relativistic effects, even in approximations, will necessarily involve thinking about fields and PDEs, which is fundamentally different from dealing with a small set of point particles and ODEs.
<queqiao-> That is true even if the words linear or low order are uttered in both contexts.
<queqiao-> ⟨Vlood⟩ addendum to #5: do not argue maths and orbital mechanics with 🥚, you _will_ lose
<queqiao-> ⟨vevladdd 🇺🇦⟩ ⟪Vlood⟫ addendum to #5: do not argue maths and […] ⮪ *do not argue with 🥚, you will lose
<queqiao-> ⟨vevladdd 🇺🇦⟩ I think nobody has ever won
<queqiao-> ⟨Standecco⟩ ⟪Starscraper⟫ It's not baseless, but there's nothing […] ⮪ I mean, you're not going to convince them that you're right by arguing as such. If you want to have that in game, try implementing a proof of concept and, if you are indeed right, I'm sure that they'll be interested. Having no expertise in this field, I tend to listen to the opinions of (and facts reported by) people with objective skills as impressive as them;...
<queqiao-> ... but hey, giants are wrong sometimes, so if you're certain that you're right, you're more than welcome to prove that to all of us
<queqiao-> ⟨egg⟩ The interesting relativistic effects, […] ⮪ To take an analogy, the patched conics 2-body approach of KSP is good enough nearly all the time when you play with Principia. Nevertheless, to get the interesting effects you get from third-body and extended-body effects requires an entirely different approach to the problem from a numerical standpoint.
<queqiao-> ⟨Starscraper⟩ ⟪Standecco⟫ I mean, you're not going to convince […] ⮪ I think there's a misunderstanding about what I'm saying. But I no longer feel I can safely discuss it without being told I'm violating rules, so I'll cease.
<queqiao-> ⟨Standecco⟩ What I can say is that there is no universal rule saying that the linearization of a model has to provide some useful results. Some featured are so tied with the complexity of the model that you're not going to see them in a simpler one
<queqiao-> ⟨Standecco⟩ * features
<queqiao-> ⟨egg⟩ That much is true, but also, just because it has linearized in the name doesn’t mean it is simple, especially when you are trying to simulate something rather than prove a specific property in a reference frame hand-picked for the purpose.
<queqiao-> ⟨egg⟩ ⟪Starscraper⟫ I think there's a misunderstanding […] ⮪ Part of the misunderstanding may be the discrepancy between studying a problem (such as the classic derivations of various effects in the Schwarzschild or Kerr metrics you will have seen in textbooks) and running a numerical simulation, even in an approximate model (the latter being what a game must do unless it can resort to a purely analytical method as stock KSP does).
<queqiao-> ⟨egg⟩ These are profoundly different exercises, and your knowledge of the former is largely inapplicable to the latter.
<queqiao-> ⟨egg⟩ To take a Newtonian example: you can look at the circular-restricted 3-body problem, put yourself in the barycentric rotating frame, derive the effective potential to find the Lagrange points, and compute the Ляпунов exponents there to see which ones are stable.
<queqiao-> ⟨egg⟩ To simulate what happens in that problem, you don’t care one bit about the existence of the Lagrange points, but you need good integrators (you mentioned being familiar with RK methods, you might want a symplectic method here, but this is basically the idea).
<queqiao-> ⟨egg⟩ These two approaches have nothing in common.
<queqiao-> ⟨egg⟩ To simulate what happens in that problem, you don’t care one bit about the existence of the Lagrange points, you don't want to touch a rotating reference frame, but you need good integrators (you mentioned being familiar with RK methods, you might want a symplectic method here, but this is basically the idea).
<queqiao-> ⟨egg⟩ To simulate what happens in that problem, you don’t care one bit about the existence of the Lagrange points, you don’t want to touch a rotating reference frame, but you need good integrators (you mentioned being familiar with RK methods, you might want a symplectic method here, but this is basically the idea).
<queqiao-> ⟨egg⟩ To take a Newtonian example: you can look at the circular restricted 3-body problem, put yourself in the barycentric rotating frame, derive the effective potential to find the Lagrange points, and compute the Ляпунов exponents there to see which ones are stable.
<queqiao-> ⟨egg⟩ ⟪Starscraper⟫ I think there's a misunderstanding […] ⮪ To your specific concern wrt the rules, you are welcome to ask questions about numerical methods; if you make statements about numerical methods, you should probably remain concrete and specific.
<queqiao-> Statements about the difficulty or feasibility of a technical approach (including numerical methods) have a tendency to be uninformed and thus unwelcome; this has been a recurring pattern here.
<queqiao-> ⟨Soviet Onion⟩ why are people underestimating relativity now ? there's a reason why most people aren't taught GR in schools or colleges, and some of the best solutions to certain situations we have are numerical solutions that require enormous computer power 🤦
<queqiao-> ⟨Soviet Onion⟩ numerical relativity is a field of study on its own
<queqiao-> ⟨Starscraper⟩ ⟪Soviet Onion⟫ numerical relativity is a field of […] ⮪ I know. I'm in it.
<queqiao-> ⟨Soviet Onion⟩ even linearizing the fields won't reduce the computational power drastically :D
<queqiao-> ⟨egg⟩ > Statements about the difficulty or feasibility of a technical approach (including numerical methods) have a tendency to be uninformedIn either direction, I should add. I remember the days when people said n-body gravitation in KSP would require a supercomputer.
<queqiao-> ⟨Soviet Onion⟩ ⟪Starscraper⟫ I know. I'm in it. ⮪ that's good to know 👍
<queqiao-> ⟨Soviet Onion⟩ why are people underestimating relativity now ? there's a reason why most people aren't taught GR in schools or colleges, and some of the best solutions to certain situations we have are numerical solutions that require enormous computaational power 🤦
<queqiao-> ⟨Soviet Onion⟩ why are people underestimating relativity now ? there's a reason why most people aren't taught GR in schools or colleges, and some of the best solutions to certain situations we have are numerical solutions that require enormous computational power 🤦
<queqiao-> ⟨Soviet Onion⟩ but I didn't know this discussion would be going on even till now 😳
<queqiao-> ⟨Soviet Onion⟩ indeed
<queqiao-> ⟨Starscraper⟩ +(But only barely, though)
<queqiao-> ⟨Butcher⟩ ⟪egg⟫ > Statements about the difficulty or […] ⮪ Principia running on a TI-92 when?
<queqiao-> ⟨egg⟩ Principia on an HP-67
<queqiao-> ⟨Soviet Onion⟩ ⟪Butcher⟫ Principia running on a TI-92 when? ⮪ KSP on TI-92 ? 😳
<queqiao-> ⟨Starscraper⟩ In any case, Principia shows it's possible to fundamentally alter how KSP models gravity. That's promising/interesting. Maybe one day if I have the time/energy I'll take a shot at seeing if I can understand it well enough to make something to approximate relativity in KSP. Doing so might be a distraction from my actual education - but on the other hand, it could also actually be extremely informative for it, as an exercise.
<queqiao-> Back when I took linear algebra, I don't think I really understood the basic vector operations - even after I took the class - until I started using vectors to make programs in kOS, lol. Nothing teaches quite like doing. It was only then that it all really _clicked_.
<queqiao-> ⟨SpaceMarine⟩ Happy to report i finally got a proper transfer to venus after the first failed attempt a couple years ago
<queqiao-> ⟨egg⟩ ⟪Starscraper⟫ In any case, Principia shows it's […] ⮪ > Nothing teaches quite like doing.That is very true; for instance we learned a lot about how to think about reference frames in classical mechanics by reworking successive broken abstractions in Principia (we eventually ended up with https://github.com/mockingbirdnest/Principia/blob/7e8200a620b19e29455c2df5baccbc5888cdd52b/physics/dynamic_frame.hpp#L33 which works quite...
<queqiao-> ... well).
_whitelogger has joined #principia
<queqiao-> ⟨Starscraper⟩ So does Principia use multiple frames to make its calculations in the actual numerical integration, or is the frame changing just for displaying things, or converting maneuvers into the calculation frame?
<queqiao-> ⟨sichelgaita⟩ The integration itself happens in the inertial barycentric frame of the solar system, but then motions are transformed into the appropriate frames for display and analysis.
<queqiao-> ⟨Butcher⟩ In the land of the blind...
UmbralRaptop has quit [Remote host closed the connection]
UmbralRaptop has joined #principia
UmbralRaptop has quit [Remote host closed the connection]
<raptop> Soviet Onion/Butcher: I'm pretty sure that a TI-92 is running a 16bit CPU, and even if I'm wrong it's at best 32bit. Principia has never been able to target either of those.