New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Child transforms move on load flight state with RSS #1421
Comments
I have no idea what is going on here; it might be related to #1404 and similar bugs, because unless we manage the vessel we shouldn't be doing things to its parts; it might be #1404, or it might be yet another condition under which the vessel should be unmanageable. I guess I might need to make an RSS install. |
Do you do any global changes, like changing coordinate systems, axes,
frames, etc? Changes that reposition everything?
…On Jun 1, 2017 12:37 AM, "Robin Leroy" ***@***.***> wrote:
I have no idea what is going on here; it might be related to #1404
<#1404> and similar
bugs, because unless we manage the vessel we shouldn't be doing things to
its parts; it might be #1404
<#1404>, or it might
be yet another condition under which the vessel should be unmanageable. I
guess I might need to make an RSS install.
Do any stock parts have child transforms? If that's not the case there
might also be something broken in the way we handle children (though there
still has to be a separate manageability issue, because we should really
not be touching the vessel until it leaves the @ferram4
<https://github.com/ferram4> sector).
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#1421 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AE5-Ig33WCfYcHxWT0A5oewSW7c-yUZeks5r_mougaJpZM4NseE_>
.
|
Well, we rotate the planets and tilt the universe... :-) |
Specifically, a vessel is mentioned when it is found unmanageable after having been managed, and the reason for the newfound unmanageability is given, so that log makes mismanagement of manageability quite obvious. |
Ok, gonna try making a weird-named vessel, launching, and posting log. |
Ok. I turned on journalling and increased stderr to max so INFO goes there too, then clicked VAB, built craft, launched, alt-F4'd. I don't see any mention of the craft in the log. |
Hmm; it's not a manageability issue then. Curiouser and curiouser! |
Interestingly, launching, crashing, then reverting to launch fixes the transforms (both on the satellite and on the SRB bell). |
Update: A lot of stuff happens between OnStart and OnStartFinished, so this might be a clue (perhaps we should wait for a bit when things start)?
|
More incomprehensible update: now tracking the FlightGlobals readiness, looking at both launch from VAB and revert to launch.
Revert:
I have no idea. |
More logging:
TODO(egg): what if we disable ourselves if mode is IDLE and last mode is TRACK_Phys? (and probably some of the other things so that we don't false positive) |
I do similar stuff in precalc for physics easing (rather, that's the signal to begin physics easing). |
I do nothing with easing, why? |
Well, try reproing with it disabled in the main menu settings page? That is
another thing that fires going from rails to TRACK_PHYS.
…On Sun, Jun 4, 2017 at 1:58 AM, Robin Leroy ***@***.***> wrote:
I do nothing with easing, why?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#1421 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AE5-IlAGC5BZ0deD5WiYizoMb7JbMDoYks5sAnG6gaJpZM4NseE_>
.
|
The bug is reproducible with physics easing disabled. |
Skipping Principia's AdvanceTime and SetBodyFrames when the active vessel is in PRELAUNCH with last mode TRACK_Phys and mode IDLE seems to do the job (we skip two frames).
|
Awesome work guys! <3 |
This (a) may be a duplicate of the Proc SRB Bell issue, if that's been reported, and (b) may be an issue, albeit smaller, with stock size systems (why would I touch stock, so I don't know).
On load to flight state (say, spawning a rocket, either from the VAB or via KCT), small child transforms of parts are offset. The test case here, as repeatedly demonstrated in the stream I just completed, is the ModuleJettison'd decoupler structure on the SXT Stayputnik core (in RP-0, the 20in X-Ray satellite). That transform is offset about 0.4m to the side (X axis) of where it should be. It is reproducible every launch.
The text was updated successfully, but these errors were encountered: