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-_>
⟨egg⟩ There is a thing that would be interesting in the spirit of our solution to #2400, which is to draw only a segment of the history that doesn’t include the present (say you want to look at the launch of Voyager 1 in ECEF to see the trajectory of the launcher from the pad, but obviously ou don’t want decades of the Sun revolving around the Earth covering the screen in yellow noodles). If we added th
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ remembered a thing long ago
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ when was Principia's "disable Sun lens flare" gone
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ (in KSP features)
<queqiao-_>
⟨egg⟩ You can find that in the change log, but it had stopped working years ago anyway.
egg|laptop|egg has quit [Remote host closed the connection]
egg|laptop|egg has joined #principia
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ Oh, by the way
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ yesterday I found the feature of "warp to manoeurve" will overshoot for 2 min in flight planner
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ idk why
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ (it shows "all manoeurves are in the past", and I had to toggle manoeurve time back 2 days, warp, then toggle back, it shows 1d 23h 58m left, indicating there is a overshoot of time existing)
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ MJ warp helper+show on navball solves.
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ Another thing is
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ the vessel just shows the speed in HCI (like in stock), but not showing the velocity in the navball
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ the vessel just shows the speed in HCI (like in stock, it is already out of Venus SOI), but not showing the velocity in the navball (edited)
<queqiao-_>
⟨egg⟩ yeah, we are not overriding the vessel label; I guess we should.
<queqiao-_>
⟨Sol⟩ It worked at some point?
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ Another thing is that how's principia's prediction line (pink) drawn every frame?
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ At least not for xCyA.
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ And another thing is that analysis only works for inertial orbit?
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ (no analysis for L-points derived orbits)
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ At least not for xyO/xCyA. (edited)
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ At least not for xyO frames. (edited)
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ And another thing is, is principia prediction line re-drawn every frame?
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ And another thing is, is principia prediction line (pink) re-drawn every frame? (edited)
<queqiao-_>
⟨egg⟩ it’s recomputed asynchronously ; of course it is redrawn every frame, otherwise it would appear in random places as you move the camera about…
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ However it's surely quite performance-impacting (when switching to map view, my i7-9750h laptop just drop to about like 1/3 time warp)
<queqiao-_>
⟨egg⟩ We are getting back to a point where you do not understand what you are saying, nor what I am saying. If it shows up on the screen, it has to be redrawn. Otherwise it stops being shown, that’s how drawing things work in this kind of environment.
<queqiao-_>
⟨egg⟩ The numerical integration of the prediction is asynchronous, and therefore doesn’t really incur a big cost (asking for more prediction just means it gets updated less often).
<queqiao-_>
⟨egg⟩ Drawing less stuff is where one could improve performance (this is what #3035 is about), but the main culprit here isn’t generally the prediction (if your prediction is a tangled mess, you’re not going to get much use out of it anyway; but the celestials will happily orbit while you are getting somewhere).
<queqiao-_>
⟨egg⟩ There are some things that we could try to do in the middle (streaming APIs for reference frame change and plotting, caching part of the reference frame changes), but there probably isn’t very much to gain there, and at the end of the day, whatever actually is drawn still needs to be drawn.
<queqiao-_>
⟨DRVeyl⟩ Tangentially related, there can be a lot of code that only runs during the map view, besides what Principia draws. Taking a big frame rate hit in map view indicates something, but doesn't identify where. Really, this requires looking at the Unity profiler using dev tools.
<queqiao-_>
⟨egg⟩ (and if it is in Principia it requires profiling Principia itself, which is its own can of worms…)
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ so only solution being switching to flight view to avoid lags
<queqiao-_>
⟨egg⟩ or to a reference frame where the trajectories of the celestials are less noodly, or reduce the timespan plotted to reduce the amount of noodles
<queqiao-_>
⟨Jovietus Gravitas⟩ Wierd. anything wiwrder for l4 or 5
<queqiao-_>
⟨Jovietus Gravitas⟩ Wierd. anything wiwrder for l4 or l5? (edited)
<queqiao-_>
⟨!Kerbinator (UTC+8)!⟩ *imagine severe lags in ECEF
<egg|laptop|egg>
(test)
queqiao-_ has quit [Read error: Connection reset by peer]
queqiao- has joined #principia
<egg|laptop|egg>
testing intensifies
<egg|laptop|egg>
test
<queqiao->
⟨egg|matrix|egg⟩ the testing will continue until interoperability improves
<egg|laptop|egg>
This will either fix the issue of people thinking those who talk from IRC are bots, or it will make it worse.
<queqiao->
⟨egg|matrix|egg⟩ At least it will make those discussions a little more readable.
<raptop>
🤖🤖🤖
<queqiao->
⟨ZombieZilla⟩ We have many bots speaking now
<queqiao->
⟨Silavite⟩ The bots are taking over! Ready the ECM egg cannon!
<queqiao->
⟨Silavite⟩ /s
<queqiao->
⟨egg⟩ yeah, different people on the IRC side are now different bots with the correct nicks
<queqiao->
⟨Al2Me6⟩ Oh, it will certainly make it worse 😂
<queqiao->
⟨Vlood⟩ My god, they're multiplying
<queqiao->
⟨morzenmebs⟩ Multiple bots? That’s not necessary for a chat bridge afaik
<queqiao->
⟨morzenmebs⟩ Bots can change their name for each message
<queqiao->
⟨egg⟩ there’s a single bot, it’s just using webhooks to make the chat look a little better
<queqiao->
⟨egg⟩ the bot is still our good friend @鹊桥IRC
<queqiao->
⟨morzenmebs⟩ If someone figures out how to make replies work with discord bot bridges please lmk
egg|laptop|egg has quit [Remote host closed the connection]
_whitelogger has joined #principia
<queqiao->
⟨Butcher⟩ Hmm #3035 would be a nice fix. 🤔
egg|laptop|egg has joined #principia
egg|laptop|egg has quit [Remote host closed the connection]
<queqiao->
⟨!Kerbinator (UTC+8)!⟩ well #3035 can be easier done by simply disabling celestal orbit line drawing
<queqiao->
⟨!Kerbinator (UTC+8)!⟩ well #3035 can be easier done by simply disabling celestal orbit line drawing
<queqiao->
⟨!Kerbinator (UTC+8)!⟩ (add a button to manually disable it) (edited)
<queqiao->
⟨egg⟩ I have already told you multiple times to stop telling us how things should be done, and *especially* to stop saying that those solutions are simple.
<queqiao->
⟨Soviet Onion⟩ I would expect the creators to know best of their mods, tbh
<queqiao->
⟨Soviet Onion⟩ I would expect the creators and/or writers of the mod to know best of their mods, tbh (edited)
<queqiao->
⟨!Kerbinator (UTC+8)!⟩ Truth or disguise depends.
<queqiao->
⟨Denwey⟩ > disabling celestal orbit line drawing
<queqiao->
⟨Denwey⟩ (add a button to manually disable it)
<queqiao->
⟨Denwey⟩ Yes please
<queqiao->
⟨Denwey⟩ > disabling celestal orbit line drawing (add a button to manually disable it)
<queqiao->
⟨Denwey⟩ Yes please (edited)
Simat has joined #principia
Simat has left #principia [#principia]
<queqiao->
⟨egg⟩ Before we venture down that path, we should do some serious performance analysis to see which trajectories induce lag, and why.
<queqiao->
⟨egg⟩ Perhaps we can improve the computation-frame conversion-projection-rendering pipeline to the point where it doesn’t matter; perhaps we can identify which celestials get in the way, and which ones we can keep plotting (right now we depend on the stock logic for showing a specific celestial’s trajectory or not, which is inappropriate in many situations).
<queqiao->
⟨egg⟩ Perhaps a declutter button may be of some use (even without performance concerns, sometimes there is a bit too much noodliness for readability), but figuring out what it should really do requires understanding which things are slow.
<queqiao->
⟨Butcher⟩ Or you could make Principia a mess of hacks stacked on hacks like other ksp mods. 🤔
<queqiao->
⟨siimav⟩ Pedestrian programming FTW!
<queqiao->
⟨Charon_S.⟩ Perhaps this question has already been raised a couple of times, but I do not know the answer though. What's the difficulty in realizing persistent thrust under principia environment? Is it the coding field? Or the mechanism of Principia itself? Or some sort of mechanisms between stock KSP and Principia conflict each other? I'm noob in those area so I apologize if I said something incorrect.
<queqiao->
⟨DRVeyl⟩ Primarily implementing the resource checking/availability and consumption in unloaded vessels, on the KSP side, which is generally also not where Principia wants to or should focus.
<queqiao->
⟨Quadrupole⟩ Speaking of that, a feature I always want is to manually input thrust, isp, and initial mass in maneuver planner to simulate staging. What difficulty would that impose?
<queqiao->
⟨Kerbinator ("might" applied)⟩ flight planner or persistent thrust?
<queqiao->
⟨DRVeyl⟩ Same problem: what are the active engines after you stage, what will be the mass flow? These are not things Principia wants to assess
<queqiao->
⟨egg⟩ UI work is very hard, especially if it's about adding rarely used things to something which already like too much like a tax form. Rarely used pieces will also tend to be brittle, it's not just about implementing things, but also maintaining them as the mod evolves over the years.
<queqiao->
⟨Quadrupole⟩ That makes sense. Thanks.
<queqiao->
⟨Quadrupole⟩ At least the current state is better than stock
<queqiao->
⟨Kerbinator ("might" applied)⟩ better than stock, emm
egg|laptop|egg has joined #principia
<_whitenotifier>
[Principia] pleroy labeled pull request #3264: Make it link - https://git.io/Jya0d
<_whitenotifier>
[Principia] pleroy unlabeled pull request #3264: Make it link - https://git.io/Jya0d
<queqiao->
⟨Charon_S.⟩ Regardless of the whole community, so far, users around me and myself are eager about those "rarely used pieces". I do not feel like that it's a rarely used feature.
<queqiao->
⟨egg⟩ Manual entry of engine parameters is not a rarely used feature, it is a nonexistent one.
<queqiao->
⟨Kerbinator ("might" applied)⟩ according to what he said earlier here
<queqiao->
⟨Kerbinator ("might" applied)⟩ \*probably he's saying persistent thrust, utilizing the astonishing Isp of ions IRL is a special art and can be efficient (edited)
<queqiao->
⟨egg⟩ Ah right, I missed that message.
<queqiao->
⟨Charon_S.⟩ alright it's ok
<queqiao->
⟨egg⟩ So, to answer this. What @DRVeyl said, resource consumption involves interacting with KSP in ways that are very difficult (to my knowledge this has been done only once, by Kerbalism). Then there is obviously some supporting UI work etc.
<queqiao->
⟨siimav⟩ Isn't making Persistent Thrust *work* as *easy* as excluding a vessel from Principia processing (aka Hack Gravity but for unloaded vessels)? 😛
<queqiao->
⟨egg⟩ well, yes, even easier, it is supported by removing Principia altogether :-p
<queqiao->
⟨Quadrupole⟩ Just edit the save file
<queqiao->
⟨Kerbinator ("might" applied)⟩ Then I understand, this does makes sense.
<queqiao->
⟨Kerbinator ("might" applied)⟩ however there is a success in history, the non-principia PersistentThrust itself, proving interaction with KSP is doable
<queqiao->
⟨Kerbinator ("might" applied)⟩ Well honestly saying its actual implementation difficulty cannot be estimated, but idk about trying to do that or directly implement PT code (that UI only concerns a right-click menu, nothing externally shown). No estimation or guarantee of any kind, but trying out?
<queqiao->
⟨Kerbinator ("might" applied)⟩ however there is a success in history of KSP, the non-principia PersistentThrust itself, proving interaction with KSP is doable
<queqiao->
⟨Kerbinator ("might" applied)⟩ Well honestly saying its actual implementation difficulty cannot be estimated, but idk about trying to do that or directly implement PT code (that UI only concerns a right-click menu, nothing externally shown). No estimation or guarantee of any kind, but trying out? (edited)
<queqiao->
⟨Kerbinator ("might" applied)⟩ however there is a success in history of KSP, the non-principia PersistentThrust itself, proving interaction with KSP is doable
<queqiao->
⟨Kerbinator ("might" applied)⟩ Well honestly saying its actual implementation difficulty cannot be estimated, but idk about trying to do that or directly implement PT code (that UI only concerns a right-click menu, nothing externally shown). No estimation or guarantee of any kind, but might be trying out? (edited)
<queqiao->
⟨Kerbinator ("might" applied)⟩ however there was a success in history of KSP, the non-principia PersistentThrust itself, proving interaction with KSP is doable
<queqiao->
⟨Kerbinator ("might" applied)⟩ Well honestly saying its actual implementation difficulty cannot be estimated, but idk about trying to do that or directly implement PT code (that UI only concerns a right-click menu, nothing externally shown). No estimation or guarantee of any kind, but might be trying out? (edited)
<queqiao->
⟨Kerbinator ("might" applied)⟩ however there was a success in history of KSP, the non-principia PersistentThrust itself, proving interaction with KSP is kinda doable
<queqiao->
⟨Kerbinator ("might" applied)⟩ Well honestly saying its actual implementation difficulty cannot be estimated, but idk about trying to do that or directly implement PT code (that UI only concerns a right-click menu, nothing externally shown). No estimation or guarantee of any kind, but might be trying out? (edited)
<queqiao->
⟨egg⟩ Kerbinator, I told you already that this would not happen anytime soon. Nothing you say is going to make this state of affairs change, and yes, I am aware of the existence of that mod.
egg|laptop|egg has quit [Remote host closed the connection]
<queqiao->
⟨Kerbinator ("might" applied)⟩ Oh there is a difficulty of how to draw vessel prediction whose definition is unclear. About other difficulties, may be thinking about it (just jumped out one)
<_whitenotifier>
[Principia] Kerbinator-CN commented on issue #2347: Burn in time warp with RO ion engines (new feature suggestions) - https://git.io/Jy1Oq
<queqiao->
⟨egg⟩ come to think of it, now that we have rotation and thus that Principia has its own copy of the mass of every part, resource consumption is even more complicated than it already is, we would need to update the masses in our data (and then to do so in a way that is consistent with the laws of physics, so figuring out where the angular momentum goes, and whether jet damping is a thing), so resource consum
<queqiao->
⟨Kerbinator ("might" applied)⟩ Oh, stationary pointing is also a problem in Principia
<queqiao->
⟨Kerbinator ("might" applied)⟩ With only stock SAS stabilizing direction can only get you to minimum angular vel about 5° per hour or so
<queqiao->
⟨Kerbinator ("might" applied)⟩ with AA I can get to 1° per day
<queqiao->
⟨Kerbinator ("might" applied)⟩ but for year-long mission it my still cause problem
<queqiao->
⟨Al2Me6⟩ Have you considered spin-stabilization?
<queqiao->
⟨Kerbinator ("might" applied)⟩ With only stock SAS stabilizing direction can only get you to minimum angular vel about 1° per hour or so (edited)
<queqiao->
⟨Kerbinator ("might" applied)⟩ Solar panel pointing, etc.
<queqiao->
⟨egg⟩ Anyway, it’s not happening in the next couple of years. Further discussion of persistent thrust can serve no purpose anymore, let’s get back to something more constructive, like pictures of cats.
<_whitenotifier>
[Principia] Kerbinator-CN commented on issue #2347: Burn in time warp with RO ion engines (new feature suggestions) - https://git.io/Jy1Oa
<_whitenotifier>
[Principia] Kerbinator-CN deleted a comment on issue #2347: Burn in time warp with RO ion engines (new feature suggestions) - https://git.io/Jec2V
<queqiao->
⟨egg⟩ (or the test build above; if someone has anything interesting to report regarding save growth of lack thereof, that would be interesting)
<queqiao->
⟨Kerbinator ("might" applied)⟩ Well for me going to dreamland is the best option: it's 1:08 AM at my place already.
egg|laptop|egg has joined #principia
<queqiao->
⟨Al2Me6⟩ Hm I'm seeing a lot of `serialized_plugins` with huge chunks of repeat patterns scrolling through my test save (6MiB, 3 vessels + asteroids):
<queqiao->
⟨Al2Me6⟩ Hm I'm seeing a lot of `serialized_plugins` with huge chunks of patterns scrolling through my test save (6MiB, 3 vessels + asteroids): (edited)
<queqiao->
⟨Al2Me6⟩ Hm I'm seeing a lot of `serialized_plugins` with huge chunks of near-repeats scrolling through my test save (6MiB, 3 vessels + asteroids): (edited)
<queqiao->
⟨egg⟩ Compression works in mysterious ways…
<queqiao->
⟨egg⟩ What are your 3 vessels doing?
<queqiao->
⟨egg⟩ 6 MiB is not unexpected, I think we are aiming for a MiB or two per vessel
<queqiao->
⟨Al2Me6⟩ Earth-Moon L2, Earth-Sun L2, en route to Earth-Sun L4 (heliocentric orbit)
<queqiao->
⟨egg⟩ it would be interesting to see what happens with things in low orbits as well (those are the ones that would cost a lot with #2400).
<queqiao->
⟨von Kerman⟩ Test Principia with Russian localization is not showing celestial name correctly. I suppose it's because my RSS does not have Russian celestial names yet, but this is what people with old RSS version will see
<queqiao->
⟨egg⟩ Earta sure is some creative Russian grammar :-p
<queqiao->
⟨von Kerman⟩ Things like that actually cause very bad bugs sometimes, Eartа contains both Latin and Cyrillic letter which look the same, but are not. Imagine what happens when a program doesn't check such inputs...
<queqiao->
⟨von Kerman⟩ Things like that actually cause very bad bugs sometimes, Eartа contains both Latin and Cyrillic letters which look the same, but are not. Imagine what happens when a program doesn't check such inputs... (edited)
<queqiao->
⟨von Kerman⟩ Minor issue in orbit analyzer: "сидерических оборотов" is in the correct form, but "драконический оборот" is not (wrong case).
<queqiao->
⟨von Kerman⟩ There is actually a bigger issue in this screenshot: History length translated simply as "Длина" (Length), without the word History, maybe it didn't fit. It is not clear.
<queqiao->
⟨egg⟩ that’s not the history length, that’s the duration of the analysis ; in English it just says « duration ».
<queqiao->
⟨egg⟩ should be драконических оборотов I suppose?
<queqiao->
⟨von Kerman⟩ Yes
<queqiao->
⟨von Kerman⟩ Not sure how to translate "duration" better. "Длина" in this context sounds like "длина орбиты", "length of the orbit", which doesn't make sense. "Длительность" doesn't fit and is also not very clear. Maybe translate it differently: "Прогноз на" , which means "Forecast period is"?
<queqiao->
⟨sichelgaita⟩ Can I have that save? These patterns look odd, I don't see anything like that in our saves.
UmbralRaptop has quit [Remote host closed the connection]
<queqiao->
⟨Al2Me6⟩ Sent to egg already, will do
<queqiao->
⟨von Kerman⟩ Minor issue: case is incorrect, should be "Планируемое столкновение с ..."
<queqiao->
⟨von Kerman⟩ Maybe it will not happen with correct celestial names, which are hopefully shorter than distorted "The Moon" here, but it's a potential issue anyway.
egg|laptop|egg_ has joined #principia
<queqiao->
⟨egg⟩ Theого Mooа :D
<queqiao->
⟨egg⟩ please add the ru.cfg file to your RSS install next time you restart the game; while it is good to know that it looks weird, I don’t really want to do work to make that something that shouldn’t happen look good
egg|laptop|egg_ has quit [Remote host closed the connection]
<queqiao->
⟨von Kerman⟩ I added ru.cfg, celestial names are correct, the issue with text not fitting almost does not happen anymore. The only exception is Mercury.
<queqiao->
⟨von Kerman⟩ This tooltip is scary, not only in Russian but in English as well 🙂 "for operations involving the surface of The Sun: *landings*, remote sensing, ..."
<queqiao->
⟨von Kerman⟩ I added ru.cfg, celestial names are correct, the issue with text not fitting almost does not happen anymore. The only exception is Mercury. Correction: It's not the only exception, it also happens with Ganymedes, Callisto... (edited)
<queqiao->
⟨von Kerman⟩ I added ru.cfg, celestial names are correct, the issue with text not fitting almost does not happen anymore. The only exception is Mercury. Correction: It's not the only exception, it also happens with Ganymedes, Callisto and Enceladus. (edited)
<queqiao->
⟨Jovietus Gravitas⟩ Launched on a Saturn IB, headed to L2 where the calculations were found to be off, went to L1 for a correction burn, and now on my way back to L2 to finally insert into a halo orbit.
<queqiao->
⟨sichelgaita⟩ Screenshots are hard, let's take pictures of the screen 🤓
egg|laptop|egg has quit [Remote host closed the connection]