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…
<discord-_>
DRVeyl. — Egg draws very pretty pictures to demonstrate and diagnose the Earth observation system.
<discord-_>
DRVeyl. — I scribble with crayons on a canvas trying to validate some of the basic points in the raycaster. (Unsuccessfully, at the moment).
<discord-_>
lpg. — I never drew lines that straight with crayons
egg|anbo|egg_ has quit [Remote host closed the connection]
Mike` has quit [Ping timeout: 189 seconds]
Mike` has joined #principia
egg|cell|egg has joined #principia
<discord-_>
Ashnoom. — On the subject of principia. I think I managed to find a way that works for me to get to the 'correct LAN':
<discord-_>
Ashnoom. —
<discord-_>
Ashnoom. — Non-cheaty:
<discord-_>
Ashnoom. — 1. Launch a transfer-window-planner probe.
<discord-_>
Ashnoom. — 2. Create a node roughly in the direction you want to go to (basically prograde burn only)
<discord-_>
Ashnoom. — 3. Create a node before, to change the LAN by changing normal/antinormal, keeping the inclination to whatever the launchsite is
<discord-_>
Ashnoom. — 4. Remove the 'last node' and use the orbit analyser to see the expected LAN.
<discord-_>
Ashnoom. — 5. Use that LAN
<discord-_>
Ashnoom. —
<discord-_>
Ashnoom. — Cheaty:
<discord-_>
Ashnoom. — 1. Launch a transfer-window-planner-probe//hyperedit
<discord-_>
Ashnoom. — 2. Once in space use hyperedit to change the LAN
<discord-_>
Ashnoom. — 3. use MJ's maneouvere planner to find a transfer
<discord-_>
Ashnoom. — 4. observe dV requirements and go back to 2 if the dV is larger than previous dV
<discord-_>
Ashnoom. — 4.b ensure the node created by MJ is mostly prograde
<discord-_>
Ashnoom. — 5. use that LAN
egg|cell|egg has quit [Read error: Connection reset by peer]
egg|cell|egg has joined #principia
egg|anbo|egg has joined #principia
<discord-_>
Ashnoom. — @egg I pressed rebase then the 'zero' next to plan-length and this message then pops up:
<discord-_>
Ashnoom. — Why is the 0 there if it doesn't work?
<discord-_>
egg. — yes that is fine, it means you asked for an overly long flight plan, and it timed out
<discord-_>
Ashnoom. — or, asked differently, What is the usecase of the 0 button?
<discord-_>
egg. — because nobody did the work to remove the 0
<discord-_>
Ashnoom. — Perfect answer 😉
<discord-_>
Damien. — the 0 is there because it looks like an egg
<discord-_>
Ashnoom. — Alright, I'll then do manual input when I have a 'long' plan to delete 🙂
egg|anbo|egg has quit [Remote host closed the connection]
egg|anbo|egg has joined #principia
<discord-_>
Zeusbeer. — It would be pleasant to have, especially now that the game freezes after deleting a long flight plan
egg|anbo|egg has quit [Remote host closed the connection]
<discord-_>
egg. — or you know, we could fix that bug
<discord-_>
egg. — at which point it becomes pointless again, because when do you ever want a flight plan of length 0
<discord-_>
lpg. — essentially never, but one might on occasion want a flight plan that's N days and everything else zero. the 0 button would be a convenient shortcut. (hardly important, though)
<discord-_>
Ashnoom. — you can use the button to set 'Minimum'
<discord-_>
Ashnoom. — ?
<discord-_>
Ashnoom. — you can use the button to set 'Minimum'? (edited)
egg|cell|egg has quit [Ping timeout: 198 seconds]
<discord-_>
Zeusbeer. — Not 0, but sometimes having it close to zero can reduce a lot of clutter noodles on the screen
<discord-_>
Zeusbeer. — Or sometimes your flight plan is super long, and it's quicker to set it to zero than using the slider to go to 0 days
<discord-_>
Zeusbeer. — Or sometimes your flight plan is super long, and it's quicker to set it to zero than using the slider to go to back in time (edited)
<discord-_>
Zeusbeer. — Also with the clutter problem: the slider works slower if the game has to calculate the longer flight path, making it way quicker, or at least smoother to go from zero back to a desired length
<discord-_>
Zeusbeer. — In short, don't disrespect the 0 option :snf:
<discord-_>
Zeusbeer. — In some way I'd prefer buttons for manipulating the length of a flight plan
<discord-_>
egg. — you can also type in 1 d or whatever length you like
<discord-_>
Zeusbeer. — True yes, but clicking is more satisfying than clicking, having to remove existing numbers, replacing the numbers and then pressing enter
<discord-_>
Damien. — the progressive slider is good but agony to use at low fps
<discord-_>
egg. — I guess the main question here is whether you ever want fine control over the length of a flight plan
<discord-_>
egg. — that is what the sliders are for: fine control regardless of the current value
<discord-_>
Zeusbeer. — I just want the 0 button 😅
<discord-_>
Zeusbeer. — Or maybe a 1 day button, or even better: a single orbit button
<discord-_>
Damien. — I personally like the precisenode approach
<discord-_>
egg. — No you don’t want the 0 button, because it is entirely pointless.
<discord-_>
Damien. — directional arrows for back and forth, with buttons to set the increment
<discord-_>
Damien. — but I also like sliders too so *shrug*
<discord-_>
Zeusbeer. — Only for flight path length however
<discord-_>
egg. — I cannot stand this separate increment size/increment approach, it is horribly fiddly
<discord-_>
Damien. — true but is fps agnostic
<discord-_>
Damien. — I hate trying to tweak things when the game gets laggy
<discord-_>
egg. — well, clearly the answer is to make the game less laggy
<discord-_>
egg. — not change the controls to be fiddlier but lag-compatible
<discord-_>
Zeusbeer. — I do this by reducing flight length and setting history+predicted path close to 0
<discord-_>
Zeusbeer. — But reducing the planners length is done with sliders
<discord-_>
Zeusbeer. — Quite infuriating on a voyager like plan
<discord-_>
Damien. — I'd also like a lamborghini, a supermodel girlfriend and a pony
<discord-_>
egg. — Zeusbeer, the question is not how *you* can make the game less laggy. The question, if we are discussing changing things in Principia, is how *Principia* can be less laggy. We are not interested in hacky workarounds, and it should be clear by now that we have never been and never will be.
<discord-_>
Zeusbeer. — fair point
<discord-_>
Zeusbeer. — But my guess was that no matter how much you optimize principia, it will always be laggy to render 10 years of noodles in n-body space
<discord-_>
Al2Me6. — I personally like having both the slider and increment/decrement buttons, because it's easier for me to make fine adjustments with a button instead of dragging a slider small amounts.
<discord-_>
egg. — @Damien I do think there is room for improvement: right now the flight plan will block until it recomputes whatever you have asked for; perhaps some of it could be done asynchronously, or, in order to avoid overshoot if that is an issue, we could perhaps do an inaccurate synchronous computation, and then refine asynchronously
<discord-_>
Zeusbeer. — I find the slider very pleasant for the exponential nature of increasing an orbit on a burn
<discord-_>
Zeusbeer. — I find the slider very pleasant for the exponential nature of increasing an orbit with a burn (edited)
<discord-_>
egg. — I recommend against guessing; while there is computational cost that is hard to reduce, the issue is not that directly, rather it is waiting for it
<discord-_>
egg. — long predictions used to be unusable, now they can be quite long before they are costly (mostly because of the rendering code, which is blocking)
<discord-_>
Zeusbeer. — Would it be possible to use the gpu to render the predictions?
<discord-_>
egg. — (trajectories in general, not just predictions). Possibly, but we have tried to avoid that particular rabbit hole. Maybe we could make more of the rendering asynchronous (transforming to the correct reference frame doesn’t need to happen synchronously, maybe we could even precompute the interpolations), perhaps some other steps could benefit from laziness.
<discord-_>
egg. — (on the other hand for KC remote sensing I would expect dark GPU magic to improve things quite a lot; perhaps I can trick @Butcher into conjuring some magic there someday)
<discord-_>
Zeusbeer. — GPUs really are quite magical at doing certain tasks
<discord-_>
Zeusbeer. — yeah and it would be practically unusable for systems with integrated graphics / old video cards
<discord-_>
Zeusbeer. — and they are just weird to program, at least from my experience with Python
<discord-_>
Festive CringeName. — I mean it's not really slowing down it's just not getting cheaper no?
<discord-_>
Zeusbeer. — yes but demand is so high right now that it feels slow to actually benefit from the changes
<discord-_>
Zeusbeer. — but improvement goes hand in hand with lower prices for same performance, but this is offtopic with this channel
egg|cell|egg has joined #principia
egg|cell|egg has quit [Ping timeout: 198 seconds]
egg|cell|egg has joined #principia
egg|cell|egg has quit [Ping timeout: 189 seconds]
<discord-_>
Zeusbeer. — how much delta V does it cost to do a direct transfer + capture with Mercury?
<discord-_>
badgermasher. — It looks like I used 15.9 km/s without any attempt at optimization. I had the margin so I used it. Not really sure that's a principia question though.
<discord-_>
Butcher. — I don't find integrated gfx to be that bad to program, though if they're pre-DX11 then all bets are off (but that goes for dedicated cards too).
<discord-_>
Butcher. — They're just *slow*.
<discord-_>
Butcher. — @egg KC remote sensing? This your new stuff?
<discord-_>
Butcher. — Also transforming predictions to a different reference frame sounds like the sort of thing GPUs are very good at. 🤔
mic_e has quit [Ping timeout: 189 seconds]
mic_e has joined #principia
<discord-_>
egg. — yeah, see the maps I posted earlier
<discord-_>
egg. — (a lot of that sounds very much like the sort of things GPUs are good at, since it is mostly answering the question « can the satellite see this, at what resolution, and, is it sunlit » very frequently and for a lot of points on the surface)
mic_e has quit [Ping timeout: 189 seconds]
<discord-_>
Butcher. — Yes, it does rather.
<discord-_>
egg. — Yes; but that kind of dependency sounds like more rabbit holes at the bottom of the Principia rabbit holes (building Principia can already be a challenge; just keeping the third-party dependencies up to date and upgrading the build system is daunting)
<discord-_>
Butcher. — I'm not sure how one injects additional rendering into ksp for these sorts of things off hand.
<discord-_>
Zeusbeer. — is this only capture? or also the transfer burn to mercury from Earth
<discord-_>
Butcher. — It's typically about 6-7 km/s out and then maybe 8-9 to capture for a direct transfer.
<discord-_>
Butcher. — But it varies by your window.
<discord-_>
lpg. — my latest flyby probe needed 5.1km/s to leave, and would have needed ~9 to capture
<discord-_>
Butcher. — I'm pretty sure I had a ~6 capture, but it was a slightly more expensive TMI the usual.
<discord-_>
leudaimon. — yeah, I got a flyby for less than 5.3 as well... didn't even check how much it was for capture
<discord-_>
Zeusbeer. — yeah I feel like doing multiple flybys is dubious for the amount of effort it takes
<discord-_>
Zeusbeer. — also time
<discord-_>
Butcher. — It's fine if you have a few years to spare.
<discord-_>
Butcher. — But a direct transfer is around 100 days, so it feels long.
<discord-_>
Butcher. — Just go for a big transfer burn and aerobrake using the Sun. 🧠
<UmbralRaptop>
🌞⛵❔
egg|anbo|egg___ has quit [Read error: Connection reset by peer]
egg|cell|egg has joined #principia
<_whitenotifier-edb>
[Principia] rnlahaye commented on pull request #2981: Optimizations for DiscreteTrajectory - https://git.io/JsfgR
mic_e has joined #principia
<_whitenotifier-edb>
[Principia] pleroy commented on pull request #2981: Optimizations for DiscreteTrajectory - https://git.io/JsfKH
<discord-_>
Kobymaru. — Yes, that makes sense to me
<discord-_>
Zeusbeer. — Are the inclinations the same?
<discord-_>
Kobymaru. — Pretty much, 1° off
<discord-_>
Zeusbeer. — Are you doing rendezvous with the lander or the orbiter?
<discord-_>
Kobymaru. — The main Issue I have right now is that in the target local reference frame, the maneuver planner doesnt work
<discord-_>
Kobymaru. — In "target local", he dotted line stays the same
<discord-_>
Kobymaru. — In "target local", he dotted line stays the same if I change flight plan duration or maneuver. And it's over the current trajectory, which doesn't make sense (edited)
<discord-_>
Kobymaru. — Lander is in low moon orbit, Orbiter is in elliptical orbit and does the Rendez-Vous
<discord-_>
Zeusbeer. — so anything you do with the planner has no impact to what you see on the spiral?
<discord-_>
Kobymaru. — correct
<discord-_>
Kobymaru. — but only in "target local" reference frame, it works fine in "moon centered inertial"
<discord-_>
Zeusbeer. — wow you must be playing on 4k, the manouver node is really tiny
<discord-_>
Zeusbeer. — I'm a bit confused how it doesn't do anything, if it shows the prediction lines
<discord-_>
Kobymaru. — 2k. I can zoom in and make better screenshots if needed
<discord-_>
Zeusbeer. — no it's fine
<discord-_>
Kobymaru. — Yeah, me too... That reference frame worked like a charm for me in LEO, now I don't know what's up. Is it the moon that's confusing? Or the retrograde-ness of the orbit?
<discord-_>
Zeusbeer. — can you delete the flight plan, keep the reference frame, and just add another flight plan, so we can start from a blank slate?
<discord-_>
Zeusbeer. — I'll cheat some spacecraft in lunar orbit and try it myself
<discord-_>
Zeusbeer. — or ideally, you could livestream what you see in a discord voice channel
<discord-_>
(ᴇxᴘᴇʀɪᴍᴇɴᴛᴀʟshells). — time to update the channel description
<discord-_>
Paculino (ŝi/ri/she/they). — First, get close (the launch timing alone should get this close enough), then switch the reference frame to target to see the relative positions over time to gauge when to make the next adjustment. When you are ready to make the adjustment, switch back to the old reference frame, make the burn to get the orbits intersecting, and go to the target reference frame again to easily
* raptop
has bad flashbacks to E&M
<discord-_>
Zeusbeer. — 🌚
<_whitenotifier-edb>
[Principia] eggrobin opened issue #2982: The flight plan should not be silently truncated in target LVLH when the prediction of the target is too short - https://git.io/JsJUm
<discord-_>
Zeusbeer. — I think I'm beginning to understand it
* raptop
wants to say that Principia isn't really putting any effort to be in CKAN because there are lots of potential mod conflicts (outside of the namespace colliding RSS)
<_whitenotifier-edb>
[Principia] pleroy commented on issue #2922: [Info] FARc 0_16_0_1_Mader can lead to a Principia Fatal error in KSP 1.11.1 - https://git.io/JsJ3O
egg|anbo|egg has joined #principia
egg|anbo|egg has quit [Remote host closed the connection]