OHM LolaMFD for Orbiter 2010 (new corrections)

[JonnyBGoode]

Is it only the Europa base that this happens to?

 

[Tycho6]

NaNspace? Hah, I hadn't heard that one before, though I have seen it before. :lol:

I tried landing at Brighton Beach from LLO, and LolaMFD pulled it off perfectly. As I did that from a lower orbit (44 km v. 80 km) I tried again at Europa from an even lower altitude on the off chance there is some altitude sensitivity causing it. Alas, from 24 km above Europa I got exactly the same visit to NaNspace. If I leave the ship there for a minute or so, I also get a CTD. Whee!

Lola also successfully landed the DG at Olympus Base on Mars. I'm not sure what it has against Europa...hmm, I notice UCGO Base Europa has no pads, could that be a problem?
In addition, I've had several CTDs when selecting LolaMFD, as soon as I click on it, but before it comes up on the MFD.

 

[JonnyBGoode]

I also got the NaNspace bug... I was trying to land the Vulture add-on at Tranquility base. I think I might have had it on 10x speed when it happened, I'm not sure. But suddenly both the ship and the Moon disappeared and everything looked like Tycho6's screenshot.

Confirmed, I reran the scenario and got the same bug at the same place, in normal speed. About 3 seconds past changing from Braking phase to Approach phase, it does it. And Orbiter locks up such that I have to task switch to kill it.

So yes, it happens over the Moon too, not just over Europa.

 

[Rawash]

LolaMFD 2010 - New fixes

I was able to reproduce the bug (i think this is this one...).

Sometimes in landing AP, the convergence algorithm don't choose the nearest solution (LSQ approximation).
Divergence appears which causes state change (breaking phase to Approach phase) and finally NaNspace.
So forcing the closest solution seems to eliminate the problem.
But it doesn't appear optimize.



Thanks to all.

R.

 

[Wishbone]

The code is quite convoluted and it's been a long time since I read it. If you're after maximizing or minimizing a function, please take a look at the Precession MFD code, specifically the Brent's method. If you're doing minimization of a multi-arg function, www.alglib.net has Levenberg-Marquardt, BFGS etc. algorithms already coded. And it may be a wise idea to impose bounds on the maximands...

 

[Tycho6]

I just tested the latest LolaMFD, and it has fixed the NaNSpace problem, at least as far as my landings on Europa go. Nicely done! :tiphat:

If I may be permitted one minor gripe, is there any way to remove the text which now shows up on the lower left of the screen when LolaMFD is enabled? It's somewhat annoying in screenshots, though not the end of the world.

 

[Wishbone]

It helps in debugging, doesn't it?

 

[Spacethingy]

I don't do much of that myself... :lol:
I've got CTD a couple of times now when I try to load the MFD in the DGIV. Is this a bug? hsnap:
This MFD makes a much less nail-biting landing than BaseLand MFD - but the takeoffs are terrifying!

:edit: This is weird. I get the CTD whenever I load up LOLA in a new vessel. When I try again, it's all fine.

Edit 2: Ouch, just hit a NaNSpace teleport in the Arrow on final doing a balistic trajectory to Brighton Beach. Nasty.

 

[jgrillo2002]

Apparently this MFD doesnt work too well with some vessels. notably the XR2

 

[Carmen A]

This is an autopilot and remote piloting MFD? I have to check it out :3

 

[Rtyh-12]

I got to NaNSpace while trying a ballistic trajectory from UMmu Demo base on the Moon to Brighton Beach. At first everything ran fine. Then, while braking (I was quite close to landing) I decided to switch to a ground camera view. I paused the sim, placed the camera and then un-paused. I could see the Glider getting out of control with Lola trying to correct this, then everything went nuts and I got into NaNSpace.

The Sun had something like 1#e influence on me...

 

[Carmen A]

I think you were abducted by aliens

 

[Rtyh-12]

I think you know something that I don't

 

[NukeET]

I got to NaNSpace while trying a ballistic trajectory from UMmu Demo base on the Moon to Brighton Beach. At first everything ran fine. Then, while braking (I was quite close to landing) I decided to switch to a ground camera view. I paused the sim, placed the camera and then un-paused. I could see the Glider getting out of control with Lola trying to correct this, then everything went nuts and I got into NaNSpace.

NaNSpace?

 

[joertexas]

I found out that LOLA doesn't like tail sitters, I modified my custom lander, and it was working fine.

Now, though, I find that the Delta Vee numbers aren't adding up. From my research, Apollo used something like 2.1 to 2.3km/sec to land. My lander is up to 2.6km/sec fuel capacity, and it's still running out of fuel. I'm using the same ini file, and I only tweaked the fuel and mass numbers to fine tune the mission. I'm figuring the accel and fuel correctly, because the attached orbiter (in a lander/orbiter stack) works fine, with fuel consumption right on the money.

Any ideas?

Update: Okay, now I feel *really* stupid... I checked the lander while the sim was running, and found that it was severely overweight, as if it hadn't dropped the orbiter. I put a line of code in the ini file,a dn now the weight is correct. We will see how the landing goes...


JR

 

[HarvesteR]

NaNSpace?

NaN means Not A Number... it's a state that can happen to numbers on a program if you carry out illegal operations with it... like dividing by zero (although most times that equals CTD or compiler error)... or trying to do an operation with a variable that hasn't been properly initialized.

NaNSpace, if I correctly glean it's meaning (just heard of it too :lol: ) is where your ship goes after it's been fed a NaN somewhere in it's simulation parameters...

It's the nexus of the universe. It's also where all ballpoint pen caps go when we lose them. (same for umbrellas and guitar picks)

Cheers