When I'm doing interplanetary burns I often have nodes that lie 20 days+ into the future, sometimes 200 days+

Sometimes after clicking "execute next node" in Maneuver Planner when a node lies very far into the future, the auto-warp seems to miss the node and 'brakes' too late. I've had overshoots of more than 2 hours worth.

It still aligns the craft and performs the burn, but the resulting orbit is totally out of whack due to the late burn.

setup details:

• I'm using 2.3.1.270 (Jul 25, 2014 7:45:30 PM) an have KSP 0.24.2 via Steam
• I have no other mods installed
• MJ2 seems to otherwise work correctly for me (besides some quirkiness with delta-V displays, but that could also be caused by my crazy staging schemes)

to reproduce:

• attempt node execution for a node that lies 100 days+ into the future (sometimes it doesn't occur so one might have to give it a few tries)
• in case it's hard to reproduce I could try to capture it on XSplit and post a link in the comments if that's helpful

This has to have either been fixed or been an underlying KSP issue that was fixed or people would be screaming about it.

Of course when trying to reproduce the SOB suddenly performs exemplary. Please indulge me with some more patience, I do intend to provide you with a true video example.

Could it be that the ship design you were seeing this with was very slow
to rotate ?

I have seen nodes overshot before but this is more the fault of
suboptimal ship design, rather than a MechJeb bug perhaps?

On 02/09/14 22:16, fdepee wrote:

Of course when trying to reproduce the SOB suddenly performs
exemplary. Please indulge me with some more patience, I do intend to
provide you with a true video example.

—
Reply to this email directly or view it on GitHub
#402 (comment).

@chriscamacho:

No I don't think this is the cause. My problem is that the auto-warp (fast forward of time) misses out. So we're talking about a state where rotation of the craft is not even activated. Simply fast forwarding to the burn time goes wrong, and it ends up several hours after the node T time has passed. Only then is it starting to rotate, but the problem has of course already occurred by then.

I uploaded a video of me playing KSP where the overshooting happens twice: http://youtu.be/szY86eCgkjk

Around 10:15 it happens in map mode so it's clear to see the craft is already past the node when warping ends.

I haven't tried your suggestion yet @sarbian but when I get around to it I'll post the result here

Just a hunch: could it be related to the Kerbal day length vs. Earth day length setting?
Is it possible the calculation is botched if you use Kerbal time instead of Earth time?

@fdepee have you tried changing the setting and see if it fixes the problem?

I doubt it, the setting is a pure display thing. Internal time are all in seconds anyway.
As before upgrade to newest dev release where is fixed a few thing that may be related.

Ok @sarbian I tried the latest dev build: Build 319 (Sep 10, 2014 9:40:05 AM) but the problem persists:

• the delta-v of my node jumped from ~250 m/s to ~1000 m/s as the craft achieved Eve orbital insertion.
• auto-warp was more than 1 hour late for the burn at T.

I made a video, please have a look it's only 3 minutes. Maybe you spot something that I'm doing wrong.
http://youtu.be/jLAYYYDZCJw

im sure this is caused by the compute not being able to play the game with enough FPS when warping. fps look ok tho in your vid odd.

oh, good idea there joebopie.

@joebopie Yes I have to say the problem does not occur when I play with very simple space craft, so that would seem to support that notion.

Would fiddling with the physics setting in the menu make a difference? Should I turn it up or down?

looking at the video he is approaching EVE witch has a lot of gravity and hes trying to circularise fairly close. i think as its warping it accelerating to a point that for a split second there is not enough FPS for MJ to realise it has to stop.
this used to happen to me all the time the more parts the worse it gets. i have yet to see it more that a couple of times after getting a much faster CPU.
i don't think changing the phsics delta time would do anything as its not calculation physics at the point anyway. best to not tick that auto warp box and warp yourself slower. MJ will still stop to burn when its time.

Ok, it's true that it only is a problem with huge ships (in terms of part count), so I'm ok with accepting this as a limitation of my pc. Should I close this issue or should it remain open for reference?

no, leave it open. we need a fix for big ship/slower PC.

bump

Does anyone recognize this problem? I'm still experiencing the overshoot with far planned interplanetary nodes. I can't be the only one with this problem, right? Is this perhaps related to any of the other issues? Can anyone confirm this is a problem, or am I doing something wrong?

Thanks for the reply. I'll try your suggestion. I'll also try to capture it on video as I was able to reproduce it yesterday in my current save game. In a couple of days I'll have more info posted here.

Sorry, I forgot about you :(

I could not duplicate it. I wonder if it's somehow related to framerate.

Can you try the latest dev version ( http://jenkins.mumech.com/job/MechJeb2/lastBuild/ ) ?
I made the wrap controler lower wrap rate a few frame earlier and I added 2 Info Item in the thrust category :

• Node Burn Countdown
• Node Burn Length

They show when MJ expect to burn the node and when it will start it. If the length is negative that may explain your problem too.
(you can add those in any MJ info windows by clicking the "E" in the window corner)