ferram4 changed the topic of #RO to: Welcome to the discussion channel for the Realism Overhaul (meta)mod for KSP! Realism Overhaul Main Thread https://goo.gl/wH7Dzb ! RO Spreadsheet http://goo.gl/Oem3g0 ! Code of Conduct http://goo.gl/wOSv2M ! | Maximal & soundnfury's RP-1 Race Into Space Signup: http://bit.ly/2DEVm2i [15:01] <soundnfury> Straight Eight Stronk (and) RP-0/1 is basically "Space Agency Spreadsheet Simulator"
<taniwha>
egg|laptop|egg_: I dunno about the part ids, but I do suspect that EL creating and destroying a vessel in the same frame might be giving trouble
<taniwha>
I know that used to give FAR trouble
<taniwha>
(until I convinced ferram that he does need to check that vessel is still valid)
<egg|laptop|egg_>
that might be it, there also are issues with vessels in a strange state where they are unpacked but their parts are not actually there, we do not like that
<taniwha>
The only thing I can think of for that is timing
<egg|laptop|egg_>
(as in their parts don't have a rigidbody)
<taniwha>
EL either uses ShipConstruction.AssembleForLaunch (when building), or root.AddComponent<Vessel>() then vessel.Initialize (after setting the name) when getting the build cost
<egg|laptop|egg_>
yeah this sounds very very similar to the KIS thing
<egg|laptop|egg_>
at least at five in the morning it does, I should sleep eventually
<taniwha>
yeah, the one time I don't let KSP do all the spawning is when getting the build cost, and that vessel gets destroyed within the same frame
<taniwha>
and I don't spawn individual parts
<taniwha>
(unless it's a one-part craft file)
<taniwha>
egg|laptop|egg_: btw, while I'm not very good at it yet, I have been learning tensor math
Blu3wolf has joined #RO
<egg|laptop|egg_>
nice
<egg|laptop|egg_>
Principia has made me learn about elliptic functions & elliptic integrals recently
<taniwha>
I bet
<egg|laptop|egg_>
because we added persistent rotation, and it turns out doing that correctly requires elliptic integrals
<egg|laptop|egg_>
(note: PhysX does *not* do rigid body rotation correctly)
<egg|laptop|egg_>
(we wasted a couple of months trying to understand a bug because we thought we were doing minor corrections atop physics, but in fact we were doing its job for it)
<taniwha>
PhysX expects to operate in very flat space
<egg|laptop|egg_>
that isn’t really the issue here though
<egg|laptop|egg_>
well
<egg|laptop|egg_>
I suppose if things hit the ground before they can spin it matters less
<egg|laptop|egg_>
but basically it makes no attempt at solving Euler’s equation
<egg|laptop|egg_>
things spin with constant angular velocity rather than constant angular momentum
<egg|laptop|egg_>
so no tennis racket theorem if your tennis racket is a single rigid body
<egg|laptop|egg_>
mind you, doing it right (solving Euler’s equation formally with elliptic integrals) took us a year, sending finding some sign errors in the only usable papers we could find, and sending bug report to an expert on elliptic integrals, so it is not for the faint of heart
<egg|laptop|egg_>
a physics engine could reasonably symplectically integrate Euler’s equation though, that should yield plausible results
<taniwha>
heh, debugging papers
<egg|laptop|egg_>
and Toshio Fukushima’s FORTRAN as well :-p
<taniwha>
debugging code is to be expected, especially FORTRAN :P
<taniwha>
egg|laptop|egg_: oh, one thing that might help: EL sets the name of the temporary vessel quite distinctly:
<taniwha>
craftVessel.vesselName = "EL craftVessel - " + craft.GetValue ("ship");
<taniwha>
so if the problem vessel's name looks like that, you know it's the temp vessel, otherwise it's the actual build
<egg|laptop|egg_>
yeah the thing with spawning vessels is that letting KSP do the spawning is not necessarily enough to ensure that properties that are always true in stock remain so (viz. the KIS issue)
<egg|laptop|egg_>
mostly because stock doesn’t spawn vessels all that much
<egg|laptop|egg_>
(at least not unpacked ones)
<taniwha>
my point there was to help identify where the problem is
<taniwha>
then can go from there
<egg|laptop|egg_>
somewhat unclear from glancing at the logs
<egg|laptop|egg_>
we don’t appear to log the vessel name
<egg|laptop|egg_>
we have part names, but I don’t know if that helps you
<egg|laptop|egg_>
ah nevermind we have vessel names as well
<egg|laptop|egg_>
no EL craftVessel
<egg|laptop|egg_>
the only vessel I see is Launcher
<egg|laptop|egg_>
> Skipping AdvanceTime and SetBodyFrames while waiting for the vessel to be fully ready (see #1421). I
<egg|laptop|egg_>
hmm
<egg|laptop|egg_>
doing that on a manageable vessel might be dodgy
<egg|laptop|egg_>
I think in stock it only happens when the active vessel is unmanageable anyway
<taniwha>
part names don't
<taniwha>
ok, so it seems to be the built vessel causing trouble
<egg|laptop|egg_>
so we have a vessel that we have had and known about for more than 10 s, we have this skipping of advancetime, and then we destroy the pile up in the next frame or that one
<egg|laptop|egg_>
I wonder if the ancient 1421 workaround might be messing with our lifecycle
<egg|laptop|egg_>
wait how is it 6 in the morning already
<egg|laptop|egg_>
I should sleep
Mike` has quit [Ping timeout: 190 seconds]
<taniwha>
yes, sleep
<taniwha>
I had that responding to the EL issue :P
Mike` has joined #RO
egg|laptop|egg_ has quit [Remote host closed the connection]
egg|laptop|egg has joined #RO
egg|laptop|egg_ has joined #RO
egg|laptop|egg has quit [Ping timeout: 189 seconds]
AmbulatoryCortex has joined #RO
egg|laptop|egg_ has quit [Remote host closed the connection]
TM1978m has joined #RO
TM1978m has quit [Client Quit]
AmbulatoryCortex has quit [Quit: Leaving]
egg|laptop|egg has joined #RO
egg|laptop|egg has quit [Ping timeout: 204 seconds]
Blu3wolf has quit [Ping timeout: 190 seconds]
Blu3wolf has joined #RO
Blu3wolf has quit [Read error: Connection reset by peer]
Hypergolic_Skunk_ has joined #RO
TM1978m has joined #RO
Mike` has quit [Ping timeout: 204 seconds]
Mike` has joined #RO
egg|laptop|egg has joined #RO
Hypergolic_Skunk_ has quit [Quit: Connection closed for inactivity]