Sigma88 changed the topic of #Kopernicus to: #Kopernicus | Release 1.3.0-4 - https://git.io/v0Gg5 | Pictures: https://git.io/vHqLy | List of packs: http://git.io/vWjiY | GitHub: http://git.io/vWAFE | Examples: http://git.io/vWAF9 | Visual Editor: http://git.io/vWAbq | Funny Quotes: https://git.io/vHqt8 | <Kopernicus> Kopernicus/master 91e89a4 thomas-jenkins: Kopernicus 1.2.2-6 - "404 - Title not found"
ferram4_ has joined #Kopernicus
ferram4 has quit [Ping timeout: 383 seconds]
Mead_ has joined #Kopernicus
Mead has quit [Ping timeout: 183 seconds]
TonyC has joined #Kopernicus
egg|zzz|egg is now known as egg|afk|egg
<Kopernicus> [Kopernicus] Olympic1 commented on issue #209: I agree https://git.io/v549z
pkmniako has joined #Kopernicus
<pkmniako> o/
<Sigma88> 0/
<pkmniako> I got the surface to look OK yesterday
<pkmniako> Didn't save
<pkmniako> D:
<Sigma88> hahaha
<Sigma88> !c "Kerbin".GetHashCode()
<Qboid> 1529818993
<pkmniako> What does VertexHeightNoiseVertHeightCurve2 do?
<Sigma88> no idea
<Sigma88> noise I would guess
<Thomas> noise multiplied with curve
<Kopernicus> [Kopernicus] Sigma88 opened issue #212: flightglobalsindex vol.2 https://git.io/v5BU4
<pkmniako> Good enough for me
<Kopernicus> [Kopernicus] StollD commented on issue #212: The current system has an advantage over this one: You can change the name of a planet without killing the whole savegame. Additionally, this approach has two rather big flaws:... https://git.io/v5BTI
<Kopernicus> [Kopernicus] StollD commented on issue #212: The current system has an advantage over this one: You can change the name of a planet without killing the whole savegame. Additionally, this approach has two rather big flaws:... https://git.io/v5BTI
<pkmniako> So when I try to edit VertexHeightNoiseVertHeightCurve2 with Kittopia, the planet turns black :/
<Kopernicus> [Kopernicus] Sigma88 commented on issue #212: I agree that the switch to a new system would create a lot of issues for savegames... https://git.io/v5BIc
<pkmniako> What dds I have to use for normal maps?
<pkmniako> compression I mean
<Sigma88> the NM one
<Sigma88> DTX5_NM or something like that
<pkmniako> Thanks
<pkmniako> I wanted to be sure
<Sigma88> but make sure you export it properly
<pkmniako> Flipped?
<Sigma88> I don't think you can simply take the kittopia output and export it
<pkmniako> awww
<Sigma88> at least, in my photoshop I need to:
<Sigma88> 1- separate RGBA
<Sigma88> 2- copy A and paste it over R
<Sigma88> 3- clear B (all white)
<Sigma88> 4- remove A
<pkmniako> Yeah, I remember that
<Sigma88> 5- export as dtx_nm
<Sigma88> that should be enough
<pkmniako> How come KSP ended up with two different renders of planets, the close one and the SpaceSpace one?
<Thomas> Because PQS calculation is expensive
<Thomas> if the planet is just one pixel big, you dont need to generate the full PQS
<pkmniako> Isn't that why we have LOD?
<Sigma88> LOD is taking the concept of scaledversion to the extreme :D
<pkmniako> lo
<pkmniako> l
<Sigma88> basically, when the planet is so small that you don't even need the textures, they are taken off the ram (or something like that, I'm don't have a very good grasp on how that stuff works)
<pkmniako> Well, just taking the .png from exports and converting .dds "works", but neither the ocean nor the mesh gets into ScaledSpace
<Thomas> No, LOD improves RAM usage, Scaled Space improves CPU speed
<Thomas> Sigma88: When the planet isn't rendered anymore, the textures are unloaded, yes
<Sigma88> I was thinking on a broader level, both features allow the game to run better
<Sigma88> :D
<Thomas> Sigma88: For the PQS it is similar, if the map wasn't used for some time, it is unloaded
<Thomas> Yeah
<Thomas> Sadly, more and more people seem to be deactivating LOD
<Sigma88> I think it has issues with linux/mac
<Sigma88> not necessarily LOD
<Sigma88> but one of the requirements for LOD
<Sigma88> and with 64bit I guess the RAM issue is less common
<pkmniako> oh oh
ferram4_ has quit [Ping timeout: 383 seconds]
<Sigma88> thomas, I took a look at the KSP code yesterday for the sigmabinary issue
<pkmniako> Can't find the alpha channel
<Sigma88> I think squad changed something
<Sigma88> pkmniako: layers > new mask layer > from transparency
<Sigma88> try that
<pkmniako> true
<Sigma88> for the sigmabinary issue, it seems like SQUAD changed the code for calculating closest approach
<Sigma88> I will have to go through all the code >_>
<Sigma88> comparing it to the old code to see what changed
<Kopernicus> [Kopernicus] StollD pushed 2 new commits to master: https://git.io/v5BLK
<Kopernicus> Kopernicus/master c3fe713 Dorian Stoll: Clean up some things
<Kopernicus> Kopernicus/master b2ca01a Dorian Stoll: Merge remote-tracking branch 'origin/master'
<Qboid> Starting build #116 for job Kopernicus-Dev (previous build: SUCCESS)
<Qboid> [640b7] title: Add contractWeight option... by Dorian Stoll | Additions: 26 | Deletions: 1 | https://github.com/Kopernicus/Kopernicus/commit/640b77766b945d05f0d764d3e413bc287a8dfe1a
<Qboid> Project Kopernicus-Dev build #116: SUCCESS in 48 sec: https://jenkins.tmsp.io/job/Kopernicus-Dev/116/
<Qboid> Dorian Stoll: Clean up some things
<Sigma88> nope
<Sigma88> <3
<Sigma88> cool
<Sigma88> implemented!
<Sigma88> thanks for the tip :)
<Sigma88> now I just need to figure out the stupid closest approach bug
Rokker has quit [Quit: Connection closed for inactivity]
<Sigma88> thomas do you plan to release that for 1.3.0?
<Sigma88> or to wait 1.3.1
<Sigma88> ?
<Thomas> Depends when 1.3.1 releases
<Thomas> *on
<Thomas> If I can get enough features for a release until 1.3.1 I will make a 1.3.0-7
<Sigma88> I would like to get SB out fixed for 1.3.0 with this in addition
<Sigma88> (even if I can't manage to do it before 1.3.1)
<Sigma88> just for ppl that want to stick with 1.3.0
<Sigma88> I can use a plugin to do that if you don't plan to add it to kopernicus
<Thomas> Well, it is already in kopernicus, so I have no problem with releasing it for 1.3.0
<Thomas> The thing is: I would like to have some more things before I release it
<Sigma88> ok, you can wait for more features to get in
<Sigma88> I still have no clues on how to fix sb
<Sigma88> >_>
<Sigma88> I mean, I have clues, I don't know if they'll lead me to a solution :D
<Thomas> I still dont quite understand your bug :D
<Thomas> or rather the bug you are trying to fix
<Sigma88> squad has a piece of code to calculate closest approach of orbits
<Sigma88> that method freaks out when the soi of the current body is bigger than the soi of the parent
<Sigma88> which is what happens in SB
<Sigma88> this is the code I used in 1.2.2
<Sigma88> basically I reverted KSP to use the old method (which didn't have this issue)
<Sigma88> but now even the old method throws exceptions
<Thomas> Can I see the exception?
<Thomas> And the stacktrace?
<Sigma88> I don't have my laptop with me
<Sigma88> I have the bugreports from an user but I don't know how you can trust those
<Sigma88> https://forum.kerbalspaceprogram.com/index.php?/topic/116058-wip130-galactic-neighborhood/&do=findComment&comment=3161395
<Thomas> Sigma88: It seems that this is a different method
<Sigma88> no it's definitely that one since I just silenced it yesterday and the exceptions were gone
<Sigma88> but I took a look at the code
<Thomas> Then your user has a different bug
<Sigma88> that method also calls other methods
<Sigma88> so maybe it's one of those?
<Sigma88> if you look at Orbit.FindClosestPoints_old you will see that it calls other methods
<Thomas> No, because it doesnt show anywhere in the stacktrace
<Sigma88> weird
<Sigma88> when I commented out this line the issue was gone
<Sigma88> this one
<Thomas> Because if you dont find a closest point, you cannot run the code that uses theese points
<Thomas> and if that code has errors...
<Sigma88> which methods do you see in the stacktrace?
<Thomas> getObtAtUT infinite UT on elliptical orbit UT: Infinity, returning NaN
<Thomas> at System.Environment.get_StackTrace()
<Thomas> at Orbit.getObtAtUT(Double UT)
<Thomas> at Orbit.getPositionAtUT(Double UT)
<Thomas> at Orbit._SolveClosestApproach(.Orbit p, .Orbit s, Double ByRef UT, Double dT, Double threshold, Double MinUT, Double MaxUT, Double epsilon, Int32 maxIterations, Int32 ByRef iterationCount)
<Thomas> at PatchedConics._GetClosestApproach(.Orbit p, .Orbit s, Double startEpoch, Double dT, .SolverParameters pars)
<Thomas> at PatchedConics._CheckEncounter(.Orbit p, .Orbit nextPatch, Double startEpoch, .OrbitDriver sec, .CelestialBody targetBody, .SolverParameters pars)
<Thomas> at PatchedConics._CalculatePatch(.Orbit p, .Orbit nextPatch, Double startEpoch, .SolverParameters pars, .CelestialBody targetBody)
<Thomas> at PatchedConicSolver.Update()
<Sigma88> so getObtAtUT ?
<Thomas> No, solve closest approach seems to pass an invalid universal time to the other methods
<Sigma88> ok
<Sigma88> I'll take a look then
<Sigma88> how can you tell the origin is _SolveClosestApproach and not some of the others methods mentioned?
<Thomas> The origin could be everything
<Thomas> in fact I cant find any difference between the 1.2 and the 1.3 version of these methods
<Sigma88> there is a difference in how the method I use is applied
<Sigma88> now there is a FindClosestPointsDelegate
<Sigma88> I don't think it's anything major
<Sigma88> but that's the only difference I found
<Sigma88> wait
<Sigma88> I may know what's going on
<Sigma88> <<starting at the closest encounter UT>>
<Sigma88> if that closest encounter never occurs (because the ship will never enter that soi (like in my case)
<Sigma88> this could mean that the UT is calculated as "infinity" ?
<Sigma88> meaning, you will never encounter
<Sigma88> so it should be enough to make sure the UT can never be infinity
<pkmniako> How to make Update Mesh permanent?
<Sigma88> you can't
<Sigma88> if kopernicus finds a mesh file it will load that one
<Sigma88> instead of generating a new one
<Sigma88> just remember to delete the cache files everytime you load ksp
<pkmniako> Thanks
<pkmniako> There a lot of stuff that I've forgotten since 2015
<pkmniako> So expect more stupid quiestions
<Sigma88> there are no stupid questions
<Sigma88> there are only stupid people asking totally legitimate questions
<Sigma88> :D
<pkmniako> *raises hand*
<pkmniako> Maybe having 8K normal maps isn't the brightest idea
<Sigma88> 2k are perfectly fine
<Sigma88> I mean 2048x1024
<Sigma88> 4096x2048 are better of course
<Sigma88> but I've never even considered 8192x4096
<pkmniako> I made the textures in 8192x4096 just so I could scale them back
<Sigma88> wouldn't it be better to export directly the size you intend to use?
<pkmniako> Forgot
<Sigma88> :D
<pkmniako> ¯\_(ツ)_/¯
<Sigma88> anyways, scaling 50% should not ruin the texture too much
<pkmniako> You can see just from the number of commits in my Github that it's my first time in the page lol
<Sigma88> which page?
<pkmniako> 19 commits
<pkmniako> Just 5 file
<pkmniako> s
<Sigma88> want a suggestion?
<pkmniako> Yeah
<Sigma88> don't include dds files in the github repo
<Sigma88> it gets very huge very quickly
<pkmniako> But I want the download on github
<pkmniako> Is there an alternative for uploading the textures somewhere else but be able to go there from github?
<pkmniako> nvm i have an idea
GregroxMun has joined #Kopernicus
<GregroxMun> o/
<GregroxMun> Maybe I should make the moon mission happen in an alternate solar system in real scale so that I can fine tune the travel time to the moon and the surface characteristics of it
<Thomas> pkmniako: Best would be to finish the textures and upload them when you are less likely to change them
<GregroxMun> for instance, if I make it a smaller-than-moon object at 6 g/cm^3 it's still at possible density for a rocky/iron body but now orbital period around it if you're in a spacecraft is a lot shorter
<pkmniako> I'm going to upload them onto Google Drive
Rokker has joined #Kopernicus
<pkmniako> How do one adds a imgur album on the forum?
<GregroxMun> probaly cant anymo
pkmniako has quit [Ping timeout: 180 seconds]
<Sigma88> 0/
<Sigma88> I back
<Sigma88> I'm*
<GregroxMun> o/
<GregroxMun> i made a thing with python
<GregroxMun> I'm in a computer programming class this semester that uses python
<GregroxMun> https://pastebin.com/tWgEQV4L I made this planetmaker.
egg|afk|egg is now known as egg
<Thomas> python \o/
<UmbralRaptor> ???
<GregroxMun> a project to send a small biography of the person that a given lunar crater was named after to their crater on the moon
<GregroxMun> for instance, Tsiolkovsky's biography will be flown to the center of tsiolkovsky crater
<GregroxMun> (which I hope I'm remembering is a real crater)
<UmbralRaptor> It is, somewhere on the far side.
<GregroxMun> that's what i thought
<GregroxMun> lots of russians on the side seen by, iirc, Luna 3
<Sigma88> Definitely not luna 3
<Sigma88> If you look at wiki there are a gazilion luna that just blew up
<Sigma88> A ton never reached orbit
<GregroxMun> but they werent named/numbered until they reached orbit
<GregroxMun> but some failed before the flyby or something like that
<Sigma88> Huh
<Sigma88> I definitely remember reading about that
<Sigma88> I thought they were named
<Sigma88> I just went back and you are right
<GregroxMun> https://en.wikipedia.org/wiki/Luna_3 wikipedia calls the flyby "3"
<Sigma88> Yes I noticed
<Sigma88> Idk what i remember
<GregroxMun> no you're sort of right too
<GregroxMun> a lot of Luna spacecraft blew up during the launch
<GregroxMun> they just weren't officially considered "Luna" until after
<GregroxMun> after the launch, not after the mission
<Sigma88> I'm back for realz now
<Sigma88> thomas, what does it mean when there's a "ref" in a method?
<Sigma88> mymethod(ref int x)
ferram4_ has joined #Kopernicus
ferram4_ is now known as ferram4
<Sigma88> fudge
<Sigma88> the bug was always there
<Sigma88> I guess nobody ever noticed
<Sigma88> >_>
<Sigma88> fffff
<GregroxMun> s/fudge/but i didn't say fudge
<Qboid> 4d 0h 0m 0s left to event #8: Intelsat 37e & BSAT 4a/Ariane 5 ECA [at 2017-09-05 21:51:00]. Say '!kountdown 8' for details
<Sigma88> GregroxMun: how do I enable the framerate counter on ksp?
<GregroxMun> i dont kno
<Sigma88> I figured it out :D
<Sigma88> thank god
<Sigma88> thomas, I have solved the SB issue so I am ready to release
<Sigma88> I will add the contract thing in SB itself
<Sigma88> so you don't need to release kopernicus
egg is now known as egg|zzz|egg