[WIN/MAC/LINUX] KSP Trajectory Optimization Tool v1.6.9 [New MATLAB Version!]

[Arrowstar]

4 minutes ago, Drew Kerman said:

Finally ready to migrate to v1.2.2! So now using the pre-release more often and can finally report this issue I noticed a while ago but on the older releases so had to wait to see if it had been fixed in the recent one... I get this error after startup when I try to launch the application from the Start menu (have it pinned there):

But when I start by just dbl-clicking the EXE, it loads fine.

Cool.  So when you start it from the start menu, is it just a shortcut to the application?  How does it work?

[Drew Kerman]

10 minutes ago, Arrowstar said:

Cool.  So when you start it from the start menu, is it just a shortcut to the application?  How does it work?

Yup, it's an application shortcut. Did some more experimenting and added a shortcut I created myself to the Start menu, and it works fine. So... yea I dunno. Lowest priority on this one, or just disregard.

Here's a real one tho - Mission Architect still asks you to save a recently-saved file when you try to close it via the X button

And I know you're caught up in the n-body but wondering if you've thought more about being able to add/remove bodies from the MAT file without having to close MA and load a new bodies.ini file. Could be another way for users to optimize things, but for me its mainly so that when I plot the orbit of a body it doesn't keep crashing into itself

[Arrowstar]

10 minutes ago, Drew Kerman said:

Yup, it's an application shortcut. Did some more experimenting and added a shortcut I created myself to the Start menu, and it works fine. So... yea I dunno. Lowest priority on this one, or just disregard.

Here's a real one tho - Mission Architect still asks you to save a recently-saved file when you try to close it via the X button

And I know you're caught up in the n-body but wondering if you've thought more about being able to add/remove bodies from the MAT file without having to close MA and load a new bodies.ini file. Could be another way for users to optimize things, but for me its mainly so that when I plot the orbit of a body it doesn't keep crashing into itself

1) Okay, not sure what the issue is there.  I'll leave it be for now, but if you can recreate it, let me know how and I'll look into it.

2) Noted, I'll take a look hopefully this weekend.  I bet it's an easy fix.

3) Haha, yes.  Okay, I will look into it this weekend too.  Here's the catch: no body can be removed that is already referenced anywhere in the mission.  If you do, instant crash on the application side most likely.  So if you want to remove a body, you'll need to make sure you're never in it's SoI, or it's not a coast reference body, or anything like that.  Will that work?

[Drew Kerman]

Just now, Arrowstar said:

Will that work?

That would work, since the whole point IMO of removing the body is that you don't need it to be there in the first place. Again, this just makes it a bit cleaner in the form of being able to load a bodies file once and then just make any removals in Mission Architect itself. I would see this as a thing people do at the start of their mission planning, chucking out the stuff they don't need to work with, but having it available in case they change their minds. Appreciate the time to look into it!

[Drew Kerman]

Ok here's another issue, I found some weirdness in the in-flight import from an asteroid on a hyperbolic orbit through Kerbin's SOI. Here are the mission files. Also here is the orbit data from the SFS in case you want to copy it onto another body (I use Custom Asteroids so can't give you a stock SFS):

ORBIT
{
  SMA = -25825757.894448847
  ECC = 1.4745822912273383
  INC = 116.162228616633
  LPE = 259.63362163652266
  LAN = 236.55986034674197
  MNA = -2.2671605615830068
  EPH = 18797728.15589368
  REF = 1
}

"JJP Flt.mat" has data imported from the game, while "JJP SFS.mat" has data imported from the save file. The True Anomaly field is the only one I saw a major change in.

[Arrowstar]

6 hours ago, Drew Kerman said:

Ok here's another issue, I found some weirdness in the in-flight import from an asteroid on a hyperbolic orbit through Kerbin's SOI. Here are the mission files. Also here is the orbit data from the SFS in case you want to copy it onto another body (I use Custom Asteroids so can't give you a stock SFS):

ORBIT
{
  SMA = -25825757.894448847
  ECC = 1.4745822912273383
  INC = 116.162228616633
  LPE = 259.63362163652266
  LAN = 236.55986034674197
  MNA = -2.2671605615830068
  EPH = 18797728.15589368
  REF = 1
}

"JJP Flt.mat" has data imported from the game, while "JJP SFS.mat" has data imported from the save file. The True Anomaly field is the only one I saw a major change in.

Okay, so the SFS file segment you posted above doesn't match the SFS .mat file you posted.  Is there any way you can give me a simple stock SFS file I can play with that demonstrates the issue?

[Drew Kerman]

12 hours ago, Arrowstar said:

Okay, so the SFS file segment you posted above doesn't match the SFS .mat file you posted. 

hrm.... I must have... I dunno. Will def be looking at more hyperbolic asteroids in the near future & will keep an eye out for a recurrence of this issue & try to do a better job of nailing down a reproducible example.

Did you ever find anything wrong with that Mun encounter issue where it wasn't showing up in Mission Architect?

[Arrowstar]

7 hours ago, Drew Kerman said:

hrm.... I must have... I dunno. Will def be looking at more hyperbolic asteroids in the near future & will keep an eye out for a recurrence of this issue & try to do a better job of nailing down a reproducible example.

Did you ever find anything wrong with that Mun encounter issue where it wasn't showing up in Mission Architect?

So when I said it doesn't match, that's actually okay.  It just means there might be a problem reading the SFS file, which is what I'm trying to investigate.  A stock SFS file and .Mat file that demonstrates the issue will be super helpful to tracking any bugs down!

And no, I didn't find anything wrong with the Mun encounter scenario you sent previously.  That .mat file never was going to have an encounter because the trajectory never got close enough to the Mun for an SoI transition.  You can run Graphical Analysis (distance to reference celestial body) to see what I mean.

[Drew Kerman]

sorry to go slightly off-topic here but @Arrowstar was wondering if you saw me tag you in this post regarding using GeoGebra for creating my own dynamic graphical body views for my Flight Tracker since matlab doesn't seem to have any web plugin (which I still can't believe, BTW :P). I figure anything similar to what you're doing is in the .fig files I can't access, but I think with a trial version of matlab I should have enough time to dig into the equations and make my own version. What do you think?

[Arrowstar]

39 minutes ago, Drew Kerman said:

sorry to go slightly off-topic here but @Arrowstar was wondering if you saw me tag you in this post regarding using GeoGebra for creating my own dynamic graphical body views for my Flight Tracker since matlab doesn't seem to have any web plugin (which I still can't believe, BTW :P). I figure anything similar to what you're doing is in the .fig files I can't access, but I think with a trial version of matlab I should have enough time to dig into the equations and make my own version. What do you think?

So what is it you're trying to do?  Just use my MATLAB plotting code in your own application, or at least as inspiration for your own application?  Give me a bit more detail and I'll see how I can help.

[Drew Kerman]

29 minutes ago, Arrowstar said:

Just use my MATLAB plotting code in your own application, or at least as inspiration for your own application?

Yea, pretty much the same as how I managed to re-purpose your matlab code for my JavaScript version of figuring out position, speed, etc on orbit, I'm thinking about looking into your .fig files to see if I can create a similar output in GeoGebra. Just wanted to make sure that's where I need to be looking

[something]

Has this tool ever been tested with GNU Octave? https://www.gnu.org/software/octave/

I am just asking since octave is Open Source and mostly compatible with matlab code - and mostly it is the differences that do mess up the code...

Also, octave is available for Linux, Mac and Windows which would make it universally useable.

[Arrowstar]

19 hours ago, something said:

Has this tool ever been tested with GNU Octave? https://www.gnu.org/software/octave/

I am just asking since octave is Open Source and mostly compatible with matlab code - and mostly it is the differences that do mess up the code...

Also, octave is available for Linux, Mac and Windows which would make it universally useable.

Sadly it is not. I used a precompiled library and some matlab functions that don't exist in octave. The UI is also not compatible. 

[storm6436]

*blinks* Arrowstar, you are my freaking hero. I've been grumpy about TWP presuming a perfect equatorial, circular orbit forever and had been grousing about building my own orbital planner once I made it through Classical/Modern Physics and my Astrophysics class** I'm supposed to be taking here soon-ish... and then I found this.  Seriously, I owe you I don't know how many beverages of your choice.  THANKS!

 **= I'm currently doing the "Go back to college in your mid-30s route" and I'm a astrophysics/applied math dual major with 2 years left on the physics major and 4 classes on the math major side.  As a result, I'm rather poor... otherwise I'd happily send you some money as a consolation prize for the number of SAN points everyone involved had to sacrifice to produce this tool.

[Arrowstar]

2 hours ago, storm6436 said:

*blinks* Arrowstar, you are my freaking hero. I've been grumpy about TWP presuming a perfect equatorial, circular orbit forever and had been grousing about building my own orbital planner once I made it through Classical/Modern Physics and my Astrophysics class** I'm supposed to be taking here soon-ish... and then I found this.  Seriously, I owe you I don't know how many beverages of your choice.  THANKS!

 **= I'm currently doing the "Go back to college in your mid-30s route" and I'm a astrophysics/applied math dual major with 2 years left on the physics major and 4 classes on the math major side.  As a result, I'm rather poor... otherwise I'd happily send you some money as a consolation prize for the number of SAN points everyone involved had to sacrifice to produce this tool.

You're welcome!  Let me know if you have any questions. :-) 

Edit:  What's a SAN point? :-) 

Edited April 11, 2017 by Arrowstar

[storm6436]

42 minutes ago, Arrowstar said:

You're welcome!  Let me know if you have any questions. :-) 

Edit:  What's a SAN point? :-) 

 From an old RPG game system called Call of Cthulu -- Not sure if you're familiar with HP Lovecraft's Cthulu mythos, but the short/sweet on it, is that as you discover more and more information on "Things Man is not Meant to Know" you steadily lose sanity.  The system is kinda insidious... the less you know, the less you're able to defend against eldrich horrors... but if you want to not be some extra-dimensional tentacle-beast's snack cake, you need to learn what's going on.  So the big question in the game is:  How much do you value your sanity compared to your life? 

It's a fun/funny game of "Do I read this book of arcane secrets?  I mean, I might pick up a psychosis or schizophrenia ...  but on the other hand I might also learn how to actually defeat these guys."

Also explains why people who work on terrifically complex math/physics problems don't seem quite "normal" to the normal people.  Personally, "normal" is overrated in most use-cases.

[Drew Kerman]

On 3/25/2017 at 2:42 PM, Arrowstar said:

A stock SFS file and .Mat file that demonstrates the issue will be super helpful to tracking any bugs down!

Okay I had another occurrence of this issue, here is the stock save file.

Load that, switch to the probe and when you query the orbital data direct from the game to the initial state and then add a Coast Event to PE no trajectory should show up. however if you load from the SFS it will work. Seems the TA parameter is what's being affected

[Arrowstar]

9 hours ago, storm6436 said:

 From an old RPG game system called Call of Cthulu -- Not sure if you're familiar with HP Lovecraft's Cthulu mythos, but the short/sweet on it, is that as you discover more and more information on "Things Man is not Meant to Know" you steadily lose sanity.  The system is kinda insidious... the less you know, the less you're able to defend against eldrich horrors... but if you want to not be some extra-dimensional tentacle-beast's snack cake, you need to learn what's going on.  So the big question in the game is:  How much do you value your sanity compared to your life? 

It's a fun/funny game of "Do I read this book of arcane secrets?  I mean, I might pick up a psychosis or schizophrenia ...  but on the other hand I might also learn how to actually defeat these guys."

Also explains why people who work on terrifically complex math/physics problems don't seem quite "normal" to the normal people.  Personally, "normal" is overrated in most use-cases.

Got it. Well, I work on this in my spare time when I'm in the mood, so very little sanity lost! :-) 

8 hours ago, Drew Kerman said:

Okay I had another occurrence of this issue, here is the stock save file.

Load that, switch to the probe and when you query the orbital data direct from the game to the initial state and then add a Coast Event to PE no trajectory should show up. however if you load from the SFS it will work. Seems the TA parameter is what's being affected

Got it, thanks!  I'll take a look. :-) 

[Arrowstar]

18 hours ago, Drew Kerman said:

Okay I had another occurrence of this issue, here is the stock save file.

Load that, switch to the probe and when you query the orbital data direct from the game to the initial state and then add a Coast Event to PE no trajectory should show up. however if you load from the SFS it will work. Seems the TA parameter is what's being affected

Alright, give this new pre-release a try and let me know if the issue resolves itself.

[Drew Kerman]

16 hours ago, Arrowstar said:

Alright, give this new pre-release a try and let me know if the issue resolves itself.

Yup, looking good over here!

[Arrowstar]

19 minutes ago, Drew Kerman said:

Yup, looking good over here!

Glad to hear it!

[Drew Kerman]

Also did you mean to have the "Use Stricter SOI Search" under Script->Execution Settings to now be unchecked by default?

[Arrowstar]

4 minutes ago, Drew Kerman said:

Also did you mean to have the "Use Stricter SOI Search" under Script->Execution Settings to now be unchecked by default?

Yes I did. I found that turning it off by default improved things a bit without too much of a hit to run time. :-) 

[Drew Kerman]

yea I did notice the run time hit, which made me investigate, but no complaints if it's better overall for things

[Arrowstar]

11 minutes ago, Drew Kerman said:

yea I did notice the run time hit, which made me investigate, but no complaints if it's better overall for things

You can certainly turn it back on, it still works well in most cases. :-)