egg changed the topic of #principia to: READ THE FAQ: http://goo.gl/gMZF9H; The current version is Cardano. We currently target 1.2.2. <scott_manley> anyone that doubts the wisdom of retrograde bop needs to get the hell out | https://xkcd.com/323/
<Virindi>
lamont: it also bugs me, I would like to be able to use mechjeb to execute burns, or at the VERY least know when I have expended the amount of deltav I have planned for the burn
<Virindi>
as it is now it is mandatory additional tedium
e_14159 has quit [Ping timeout: 180 seconds]
<Virindi>
(yes I know using mechjeb or whatever would not be "exactly" as predicted but I don't care, I would rather be able to do an approximate, automated burn then fix it later)
<lamont>
yeah i’m sorta going that direction wondering what APIs are available in Principia to write autopilot code
<Virindi>
plus flying around in lko is very close to patched conics anyway
e_14159 has joined #principia
<Virindi>
basically I love the idea of principia but I would not enjoy going back to the tedium of manual burns
<Virindi>
"lock the ship in the direction of the burn, start burning at half the total burn time before the node, and burn until the ship has expended an amount of deltav equal to the plan" is close enough for most maneuvers
<lamont>
yeah that’s the obvious solution to that problem
<lamont>
i’d also like to be able to hook into principia’s planner code and the orbit simulation code as well though ultimately
<Virindi>
yeah I would want to plot it in an IVA screen :)
MrSavage has joined #principia
<Virindi>
but that's more difficult since there isn't really an "orbital plane"
<lamont>
i’d just want to do trajectory optimization using principia itself to calculate the projected orbit after the burn
<Virindi>
any idea if the simulation is fast enough for that?
<lamont>
nope
<lamont>
one way to find out...
<Virindi>
well what do you mean by optimization, do you mean like some kind of search where you try different changes and try to move closer to some goal
<Virindi>
like "I want my next pe with body x to be y"
<lamont>
yeah just like interplanetary transfers
<lamont>
so doing porkchop plots which is a kind of search
<Virindi>
so like start with patched conics solution then mess with the parameters in some kind of hill climbing type fashion
<Virindi>
porkchop seems like it would be a lot of runs to do it that way
<lamont>
plus something like Levenberg-Marquardt to optimize the final trajectory
<lamont>
(which is what MJ does)
<Virindi>
hmm
<Virindi>
I just assumed it did an intrinsic calculation
<lamont>
and it should be okay to grind for awhile doing those calcs
<lamont>
not like days, or even minutes, but seconds at least
<Virindi>
you might need an external program :)
MrSavage has quit [Remote host closed the connection]
<lamont>
the alglib that MJ uses is ‘external’
<lamont>
its written in C# though so it links in
<lamont>
pincipia itself is external
<Virindi>
but you could precompute transfers out to a few hundred years or whatever for each solar system (stock, rss, etc)
<lamont>
i doubt you’d need to do that
<Virindi>
I just meant it might be better to run calculations out of game
<lamont>
yeah, but that’s where principia itself is external
<lamont>
so you could remove the KSP node object and its overhead and just hit the integrators that principia ships with
<Virindi>
any idea if principia calculates planets and moons based only on other planets and moons, or if it actually takes into account "game objects"...I once captured a really freaking huge spawned asteroid that would have been above floating point error for gravity effects
<lamont>
and i think principia simulates the whole solar system as a big n-body problem, and you don’t really need to recompute that every time, because the burn of your space ship isn’t going to noticably effect the orbit of earth or your destination (or whatever solar system + transfer)
<lamont>
lol yeah i guess there’s that if you’re actually moving asteroids around
<Virindi>
it was just above floating point error :)
<Virindi>
I was like "this thing is huge I wonder what the gravity would be..."
<lamont>
i think i would be happy to not solve the problem of setting up a grand tour of a class A asteroid…
<lamont>
*class E
<Virindi>
to capture that thing I had to burn half the mass of the asteroid with nuclear engines and it ended up in a terrible super far out orbit after a day of burning at 4x timewarp with like 50 lv-n
maccollo has quit [Quit: Web client closed]
<Virindi>
after awhile the problem solves itself as you mine the thing down to nothing :D
<Virindi>
anyway I wonder if asteroid flybys are enough to do anything to the solar system model
<Virindi>
put it slightly out so you can't calculate perfect transfers 500 years down the road
<Virindi>
probably not.
<Virindi>
but floating point can bite you :(
<UmbralRaptor>
!win32
<UmbralRaptor>
!win64
<UmbralRaptor>
huh
<UmbralRaptor>
!wa kittens
<Qboid>
UmbralRaptor: kittens (English word): 1 | noun | young domestic cat, 2 | verb | have kittens
rsparkyc has joined #principia
rsparkyc has quit [Ping timeout: 180 seconds]
rsparkyc has joined #principia
<rsparkyc>
I'm assuming that there are now contract packs that utilize principia
<rsparkyc>
i was about to extend ContractConfigurator to allow for putting sats in L1-L5, but didn't want to duplicate any work
regex has joined #principia
stratochief is now known as stratosleep
MrSavage has joined #principia
stratosleep has quit [Remote host closed the connection]
icefire has quit [Read error: Connection reset by peer]
rsparkyc has quit [Quit: Leaving.]
rsparkyc has joined #principia
rsparkyc has quit [Client Quit]
regex has quit [Remote host closed the connection]
MrSavage has quit [Quit: Leaving]
<lamont>
egg: i think i have a Cartan build but i’m getting NREs:
<lamont>
[EXC 22:57:31.168] NullReferenceException: Object reference not set to an instance of an object
egg|phone|egg has quit [Read error: Connection reset by peer]
<egg|zzz|egg>
lamont: strange
<egg|zzz|egg>
lamont: continuous nullref spam?
<egg|zzz|egg>
lamont: I don't really have a way to diagnose that, the C# gives shitty traces, so you'll have to add traces to find what breaks I'm afraid
<egg|zzz|egg>
rsparkyc: also FYI, you're using an old version :_p
<egg|zzz|egg>
s/_/-
<Qboid>
egg|zzz|egg meant to say: rsparkyc: also FYI, you're using an old version :-p
egg|zzz|egg changed the topic of #principia to: READ THE FAQ: http://goo.gl/gMZF9H; The current version is Cartan. We currently target 1.2.2. <scott_manley> anyone that doubts the wisdom of retrograde bop needs to get the hell out | https://xkcd.com/323/
<rsparkyc>
i would have a new version if i could build it myself :)
<rsparkyc>
but being on a mac...
<egg|zzz|egg>
aah mac right
<lamont>
yeah i’ve never figured out how to traces
<rsparkyc>
i need lamont for that
<lamont>
*add traces
<egg|zzz|egg>
our traces would be Log.Error(some string), or if you want them to show up in KSP's own logs, something something debug.log or something along those lines
egg|laptop|egg has joined #principia
<lamont>
ah just debug logging
egg|laptop|egg has quit [Client Quit]
egg|laptop|egg has joined #principia
<lamont>
i’d really like proper damn stack traces
egg|laptop|egg has quit [Client Quit]
<egg|zzz|egg>
me too
<egg|zzz|egg>
with bloody line numbers
<egg|zzz|egg>
like we have in the C++ when we CHECK fail/LOG(FATAL)
<lamont>
yeah pretty much anything else — ruby, java, etc
<lamont>
i think there’s a way to get it i just don’t know how to beat on mono enough to do it, and the info you can dig up on google is hard to apply to KSP
Iskierka has quit [Read error: Connection reset by peer]
<egg|zzz|egg>
lamont: I mean, I think C# does that too, it's just the garbled mono
<egg|zzz|egg>
unity has a shitty old mono
<egg|zzz|egg>
ksp has a shitty old unity
Iskierka has joined #principia
<rsparkyc>
lamont, have you been able to build Cartan on osx?
<GH>
[Principia] pleroy commented on issue #1347: retest this please https://git.io/v9OpT
<lamont>
yeah it builds
<lamont>
it doesn’t work though
<rsparkyc>
what's broken?
<lamont>
it might be in your backscroll ^
<lamont>
MrSavage left the chat room. (Quit: Leaving)
<lamont>
[11:02pm] lamont: egg: i think i have a Cartan build but i’m getting NREs:
<lamont>
[11:03pm] lamont: [EXC 22:57:31.168] NullReferenceException: Object reference not set to an instance of an object
<egg|zzz|egg>
lamont: ok, so those exceptions are a red herring then...
<egg|zzz|egg>
no idea what's going on though
<egg|zzz|egg>
lamont: try using the adapter dll from our build
<teabot>
Adaptear.
<egg|zzz|egg>
that should be cross-platform, and maybe mono builds it differently?
<lamont>
okay
<lamont>
good point
<lamont>
i’ll give it a shot after i get back, i need lunch + coffee
<egg|zzz|egg>
sure
<egg|zzz|egg>
I might need dinner too :-p
<lamont>
also thinking of uploading a build so rsparky can try it out, maybe its just me (video settings or something like that...)
<lamont>
are you europe?
<rsparkyc>
please do :)
<rsparkyc>
i'm is the US
<lamont>
east coast?
<egg|zzz|egg>
lamont: I am Europe personified, yes :-)
<lamont>
oh egg, yeah, europe
<egg|zzz|egg>
I'm in Zurich
<egg|zzz|egg>
so CET/CEST
<egg|zzz|egg>
currently UTC+2
<egg|zzz|egg>
nomztime
egg|zzz|egg is now known as egg|nomz|egg
<lamont>
okay makes sense — my boss is in the UK
Wetmelon has joined #principia
icefire has joined #principia
<Virindi>
normally msdotnet gives line numbers in stack traces when the pdb is present, mono uses "mdb" files instead so you may get line numbers if you generate an mdb...or unity might just not do it
UmbralRaptor is now known as GalacticRaptor
<Virindi>
with msdotnet it is magic, upon any stacktrace trying to look at the description it will try to load a pdb for the assembly and add line numbers to it
<Virindi>
you may just get line numbers if you generate a mdb and stick it alongside the assembly dll
<lamont>
yeah it was figuring out the magic of making the mdb and wiring it up that i couldn’t quite figure it out
<lamont>
so mono literally just looks for the same filename with s/.dll/.mdb/?
<lamont>
so there’s actually a ksp_plugin_adapter.dll.mdb
<lamont>
should that be ksp_plugin_adapter.mdb maybe?
<lamont>
i thought last time i looked at how to do this i hit a major stumbling block of how you were supposed to pass in cli flags or something to how you started the program up, and just adding them to the `open KSP.app` cli command on OSX didn’t work for me, i didn’t know what to hack up in the app to make it work or something like that...
Thomas is now known as Thomas|AWAY
<Virindi>
msdotnet looks for the same filename with .pdb
<Virindi>
google indicates mono can run in debug or not debug mode
<lamont>
yeah
<Virindi>
so unity may just have it turned off
<Virindi>
in release unity
<lamont>
it seems like adding the `—debug` flag would run the executable in debug mode which would then pick up the mdb files
<lamont>
the problem is getting through a couple of layers to add —debug to the mono command
<Virindi>
how about using the debug player
<lamont>
tried `open ~/ksp/ptest_1.2.2/KSP.app --args --debug` but while that opens up KSP it doesn’t give me stack traces from the dll that has the mdb
NolanSyKinsley has joined #principia
NolanSyKinsley has quit [Read error: Connection reset by peer]
<Virindi>
I mean there is a debug version of the unity player.
<lamont>
yeah, think i should get the build in shape and push one to rsparkyc first
regex has quit [Remote host closed the connection]
ferram4_ has joined #principia
<lamont>
oh egg: is it possible my old save from Cardano doesn’t work on Cartan?
<egg|nomz|egg>
!u —
<Qboid>
U+2014 EM DASH (—)
<egg|nomz|egg>
—debug might not work :-p
<egg|nomz|egg>
lamont: uh
<egg|nomz|egg>
lamont: I don't think it would do that?
<egg|nomz|egg>
though tbh I'm not sure we tested compatibility properly
<egg|nomz|egg>
but it should still work I think?
<lamont>
uh
<lamont>
well i just started a new save and i’ve got a principia window with text on it now
<egg|nomz|egg>
uh
<lamont>
yep
egg|nomz|egg is now known as egg|zzz|egg
<rsparkyc>
all this vector math is giving me a headache
<rsparkyc>
i'm trying to figure out if a vessel is in front of or behind a celestial body
<rsparkyc>
to differentiate between L4 and L5
<rsparkyc>
there are some many different vector objects on the vessle and the celestial body (and their orbits), i'm getting lost as to which one to use
regex has joined #principia
<rsparkyc>
mechjeb will give me phase angle, so i would think that's what i wanted, but trying to duplicate what they did does not seem to be working
<lamont>
brace yourself rsparkyc, cartan is coming…
<rsparkyc>
thx lamont :)
<regex>
egg|cell|egg, your dad cracks me up. Please pass on my regards.
<rsparkyc>
i'm heading home now, might mess with it either tonight or tomorrow