New Release Interplanetary Modular Spacecraft RC9

[jedidia]

So you made it work finally? good job!

 

[PeterRoss]

I really hope so. It was 4 a.m. when I finished it, I hope it wasn't a dream :lol: It needs testing anyway.

---------- Post added 29-03-13 at 01:42 ---------- Previous post was 28-03-13 at 20:23 ----------

UMMU-capable test lander:

Covers are used as belly landing plates - some use for covers at last!
EVA exit/return were performed as planned.

No bugs so far. Tried to build classic space-only vessels too just to check if everything works as it was. It seems it is.

What I did to achieve this was less messing with SetCenterOfGravity and more involvement in saving and loading functions. For example, there was a comment in your code saying that shifting CG in PostCreation results in CTD. But it works perfectly if performed BEFORE creation.

 

[jedidia]

But it works perfectly if performed BEFORE creation.

You've got to be kidding me :lol:

I guess you had to modify docking ports, attachment point and thruster initialisation, then?

EVA exit/return were performed as planned.

Well, next we'll have users killing UMMUs by inapropriate airlock placement...

 

[PeterRoss]

You've got to be kidding me :lol:

I guess you had to modify docking ports, attachment point and thruster initialisation, then?

I'm just calling initial ShiftCG not from first PreStep but from clbkLoadStateEx directly. It saves me from vessel being created at loaded coordinates and then being shifted to a newly calculated CG. I just save last CG shift into scenario (that's what I used MASSCENTER string for) and then when loading I shift my vessel to the last saved CG instead of recalculating it between the PostCreation and first frame. It saves me from all kinds of shifts and catapults at startup.
Of course there's a trick with saving correct CG - it's two different cases when you save it landed or in flight.

Well, next we'll have users killing UMMUs by inapropriate airlock placement...

:lol:

 

[BruceJohnJennerLawso]

You've got to be kidding me :lol:

I guess you had to modify docking ports, attachment point and thruster initialisation, then?

Oh no, has the master been :ninja:?

:lol:

Well, next we'll have users killing UMMUs by inapropriate airlock placement...

Well a UMMUs lifespan probably drops by decades the instant it steps out of any airlock.


I unfortunately have a bit of a problem that I should report, one of the most bizarre ones Ive ever seen. It seems that if I run Orbiter at high timewarp (10000x I think), certain IMS vessels will crash the simulation, and they always crash it at the same time. They usually are in systems trouble with cooling overloads, multiple system failures, and a handful of dead crew, but for some reason they crash the sim at the same time every time I run it. Do you want me to set up an example to demonstrate what I mean?

I also created a sort of station module of my own, but I wasnt thrilled with the results, so things will take a bit longer as well. Im stumped on the logo right now, so Im just working over the first round of edits to a paper copy of Spacecraft Design for Dummies. Its a truly incredible piece of writing, but I feel that I may need to work it over a bit more in certain places, so long as you dont mind that Jedidia. It really is excellent in thermodynamics and other systems related topics, but it needs some more meat in explaining logistics & low thrust navigation to be useful.

On an related tangent, have either of you ever seen a movie called The Blues Brothers?

 

[jedidia]

Do you want me to set up an example to demonstrate what I mean?

I'm not really participating currently, but that is always helpful. You can of course also run it in the debugger and see if the crash happens within IMS code. If not, IMS might still be at fault, but in a more indirect way than I'm hoping...

but it needs some more meat in explaining logistics & low thrust navigation to be useful.

Que "Lateralus" from Tool: SPIRAL OUT!!

sure, go on, work it over.

On an related tangent, have either of you ever seen a movie called The Blues Brothers?

Yes I have, multiple times. And no, I have no Idea where that tangent is coming from nor where it's going... :shifty:

 

[BruceJohnJennerLawso]

I'm not really participating currently, but that is always helpful. You can of course also run it in the debugger and see if the crash happens within IMS code. If not, IMS might still be at fault, but in a more indirect way than I'm hoping...

I can say with 99.9 % certainty that it is IMS, I just cant figure out why it would happen at a particular time every time. I should set up my source copy & weed out the bugs you originally laid out for me.

Que "Lateralus" from Tool: SPIRAL OUT!!

sure, go on, work it over.

Not sure I understood that reference :lol:

Its really a terrific guide, I just think it needs some work in order to be even better. For an average user who is used to DGIVs, XRs, & the like, I think it would really help to work them through the headaches of flying low thrust missions, and what sort of things can be done to avoid them. The WWDGN analogy seemed a bit inadequate to deal with the problem, especially since one of our goals is to have users able to do their rough planning without ever firing up Orbiter.

Im also thinking of including a good delta-vee map in that section would be a good idea as well.

Yes I have, multiple times. And no, I have no Idea where that tangent is coming from nor where it's going... :shifty:

:lol: Perfect, you're gonna love it.

 

[Scav]

:lol: Perfect, you're gonna love it.

Somehow I doubt an image of a 1974 Dodge sedan with Illinois plates floating through space is going to wash very well . . .

 

[PeterRoss]

Oh no, has the master been :ninja:?

I wouldn't say so. Jedidia and I were solving different problems hence different results. If he were about to make touchdown points support he would have come to the same solution. Or maybe a better one.

I unfortunately have a bit of a problem that I should report, one of the most bizarre ones Ive ever seen. It seems that if I run Orbiter at high timewarp (10000x I think), certain IMS vessels will crash the simulation, and they always crash it at the same time. They usually are in systems trouble with cooling overloads, multiple system failures, and a handful of dead crew, but for some reason they crash the sim at the same time every time I run it. Do you want me to set up an example to demonstrate what I mean?

Sure, post a scenario here and describe conditions leading to CTD. Are you talking about RC2 version or is it about the one modified by me?

I should set up my source copy & weed out the bugs you originally laid out for me.

Hey, we have another working version of IMS out, remember?:hello: This means that RC2 sources are obsolete, and fixing bugs in these is unreasonable.
I'm going to try to fix these bugs in my sources. And we gonna find a way to share latest sources. Do you know any?

 

[jedidia]

Not sure I understood that reference

the song:

No, it doesn't have anything to do with low-thrust navigation per se, but it's a kick-ass song (and one of the hardest I ever covered with my Band), and there's this nice bridge where the text goes "we'll ride the spiral to the end, and might just go where noone's been". And then maynard repeatedly yells "spiral out! keep going!" And low-thrust navigation is all about spirals, after all. I hum that bridge practically every time I'm departing earth with a VASIMR under time compression.

Oh no, has the master been ?

My mastery is greatly overrated. Just because I'm the one person in the world that currently knows the source code best doesn't mean I'm a better coder...

And we gonna find a way to share latest sources. Do you know any?

As I said, if more than one is working on the same code, you need a version control system.

Face is kind of our local authority figure on the topic and he's pretty fond of mercurial.
A bit more commonly known because older is tortoiseSVN

There's others too, of course. Basically what you do is create a repository on the net in which every version is stored seperately. People submit changes and these get stored as a new version, and you'll have an easier time merging the changed versions. That's about as far as I understand it. I never worked with VCS, but if there's more than one having his hands in the code it's pretty much indespensible.

 

[PeterRoss]

No, it doesn't have anything to do with low-thrust navigation per se, but it's a kick-ass song (and one of the hardest I ever covered with my Band), and there's this nice bridge where the text goes "we'll ride the spiral to the end, and might just go where noone's been". And then maynard repeatedly yells "spiral out! keep going!" And low-thrust navigation is all about spirals, after all. I hum that bridge practically every time I'm departing earth with a VASIMR under time compression.

When I launch another big rocket there's that Lucyfire song comes to my mind:

Thousand million dollars in the fire
Thousand million dollars take me higher

:RnR1:

 

[Scav]

...which in turn spawns memory of a Коrsика song:

(A) thousand miles through (a) flaming sunset / each rider has crossed
Wild wind, and old friends behind / it's a code of the road . . .

 

[PeterRoss]

...which in turn spawns memory of a Коrsика song:

(A) thousand miles through (a) flaming sunset / each rider has crossed
Wild wind, and old friends behind / it's a code of the road . . .

I see you keep exploring Russian bands :lol:


Back on an on-topic note:

Hey, we can have modular planetary bases now!:blink: Like this one. The only condition is not to use docking ports when landed, but as far as I know APs are still there even after finalisation. I think I'll give it a try.:hmm:

 

[jedidia]

Ah well... that'll be a whole can of worms. Of course you can just make modules that look like base modules and put them together, but you better forget half the thermodynamics code while doing it, as well as some other stuff. Really, making this work in a realistic way would take quite a bit of effort.

 

[PeterRoss]

Ah well... that'll be a whole can of worms. Of course you can just make modules that look like base modules and put them together, but you better forget half the thermodynamics code while doing it, as well as some other stuff. Really, making this work in a realistic way would take quite a bit of effort.

I figure. It doens't take into account the planetary shadow when calculating sun heating, isn't it?
Well, I can use GetShadow or GetSunEnergy which you're using for solars only. There's clbkGetRadiationForce too which returns force due to radiation pressure while taking planetary shadows into account.

EDIT: On another hand, I believe it takes planetary shadows into account. At least radiators aren't heated by the sun when in shadow, I think. And... I had a glance at thermodynamics code... it's way too much for me, I don't understand a thing there :lol: At least for now.

 

[jedidia]

I figure. It doens't take into account the planetary shadow when calculating sun heating, isn't it?

It does, that's not the problem. It doesn't take into account ground heating, heat disposal through the ground, and last but not least atmosphere temperature and conduction where applicable. You'd have to develop a model to calculate those things first to do it accurately. The current system would halfway work on atmosphereless bodies (you'd totally need to have rotatable radiators, though), but would utterly fail you anywhere else.

I had a glance at thermodynamics code... it's way too much for me, I don't understand a thing there

I think that's mostly because of poor documentation. There's a whole lot of undocummented equations in there.

 

[BruceJohnJennerLawso]

I wouldn't say so. Jedidia and I were solving different problems hence different results. If he were about to make touchdown points support he would have come to the same solution. Or maybe a better one.

I really didnt mean anything by it, but it seems funny that the solution was so unexpected. Obviously we owe Vchamp a big thanks for the original code, but sometimes working from non-original source can be a headache.

Sure, post a scenario here and describe conditions leading to CTD. Are you talking about RC2 version or is it about the one modified by me?



Hey, we have another working version of IMS out, remember?:hello: This means that RC2 sources are obsolete, and fixing bugs in these is unreasonable.
I'm going to try to fix these bugs in my sources. And we gonna find a way to share latest sources. Do you know any?

Running RC2 original (we can call your new version RC2.1 I guess). Ill post the scenario when I get a chance.

Yep, I just wasnt sure whether setting up a repository would really be worth the hassle, but once we get going, code changes will be happening left and right. Maybe we can start by posting your RC2.1 source for me to download?

Somehow I doubt an image of a 1974 Dodge sedan with Illinois plates floating through space is going to wash very well . . .

Ah :lol:., its not that actually, but that would be funny in a certain way. I would really love to see a UCGO car of "that dodge"!

Ah well... that'll be a whole can of worms. Of course you can just make modules that look like base modules and put them together, but you better forget half the thermodynamics code while doing it, as well as some other stuff. Really, making this work in a realistic way would take quite a bit of effort.

Not sure I follow. Are you talking about the add-on he posted, or about using IMS like that? In case you were wondering, I checked the package for IMS module potential, and there isn't much. Im converting some of the realistic modules that were on Orbiter Francophone (Destiny, Cupolav2, Polaris, etc), but we're really starting to scrape the bottom of the barrel in terms of finding good ones. Hopefully Hlynacq can change that for us.

It does, that's not the problem. It doesn't take into account ground heating, heat disposal through the ground, and last but not least atmosphere temperature and conduction where applicable. You'd have to develop a model to calculate those things first to do it accurately. The current system would halfway work on atmosphereless bodies (you'd totally need to have rotatable radiators, though), but would utterly fail you anywhere else.

So IMS wouldnt really be accurate for ground bases without some major code additions? I think that might be doable eventually though, assuming that we can learn the necessary thermodynamics to do it.

 

[jedidia]

I think that might be doable eventually though, assuming that we can learn the necessary thermodynamics to do it.

Oh, it's doable alright, don't think I never played with the thought. In fact I considered it a logical extension of the application, but I would make it a separate add-on. And preferably after a core re-write, things are confusing enough as they are...

 

[PeterRoss]

Speaking of, as much as I love 'IMS' term, it will be quite obsolete at the time of core rewrite, because it'll be much more than just Interplanetary Spacecraft. Something like Universal Modular Platform maybe.

 

[BruceJohnJennerLawso]

Oh, it's doable alright, don't think I never played with the thought. In fact I considered it a logical extension of the application, but I would make it a separate add-on. And preferably after a core re-write, things are confusing enough as they are...

Speaking of, as much as I love 'IMS' term, it will be quite obsolete at the time of core rewrite, because it'll be much more than just Interplanetary Spacecraft. Something like Universal Modular Platform maybe.

Well essentially the core IMS framework can be expanded well beyond its original scope to do a lot of things. I think that it would make the most sense to fork each project as an independent add-on, specifically optimized for each application. The big three that I see would be IMS (the current project), a similar application designed for simulating unmanned space probes that a user builds part by part (think building a probe in KSP)




Sorry about that :lol:

The last one would be MBS, Modular Base Simulator, basically just IMS tweaked specifically for creating surface bases, and probably allowing users to sim things like digging underground, setting up solar farms, & finding resource deposits.