Proposal to market KSP2 as actual rocket trajectory simulator.

[Exoscientist]

 Too bad KSP2 didn’t take off. Couldn’t more exposure and revenue be generated by marketing it as an actual commercial sim for spaceflight to orbit and interplanetary trajectories?

There is nothing in the open market that does this now in an accurate way. I like using Silverbidrastronautics.com payload estimator but it’s at best only within 10% accurate and it also has wide potential error bars. It also has poor accuracy for large SRB’s like the space shuttle, SLS, Ariane 5/6.

Governmental software like POST and OASIS are ITAR restricted and are highly non-user friendly. Believe me, I tried. KSP being European would not be subjected to ITAR.

 Bob Clark

[cubinator]

Have you tried GMAT? I haven't seen what it can do between ground and orbit, but it's highly versatile and professional for interplanetary stuff.

[kerbiloid]

And don't forget to integrate ISRINEX to compute nukes for the Orion drive.

[Meecrob]

Didn't you hear the news? Some people on the main forum think Elon Musk is gonna buy KSP2. I put in a bid as well. We will have to wait for all the other publishers to turn KSP down, but there's no way the IP is worth that much with the malarkey that went on in development.

[Nuke]

i think orbiter has a better integrator.

[PakledHostage]

8 hours ago, Meecrob said:

Didn't you hear the news? Some people on the main forum think Elon Musk is gonna buy KSP2.

That sounds about as wishful as my hoping in 1986 that Santa Claus would bring me a G.I. Joe aircraft carrier for Christmas... didn't happen then and ain't gonna happen now.

[HebaruSan]

13 hours ago, Exoscientist said:

marketing it as an actual commercial sim for spaceflight to orbit and interplanetary trajectories?

Isn't that a much, much smaller market?

[kerbiloid]

23 minutes ago, PakledHostage said:

That sounds about as wishful as my hoping in 1986 that Santa Claus would bring me a G.I. Joe aircraft carrier for Christmas.

Was it a decade earlier, I would suggest that Santa granted your wish, and they recruited you to the aircraft carrier crew.

[Exoscientist]

1 hour ago, HebaruSan said:

Isn't that a much, much smaller market?

 I remember asking on this forum once why Kerbal used Kerbin and Mun, rather the real Earth and Moon and other real planets in the Solar System. I remember being not really being satisfied with the answer. I think with the interest in universities and schools worldwide in spaceflight, there would be a great deal of interest in having a program where they could design real rockets for making real missions anywhere in the Solar System. 
 

  Bob Clark

 

[Nuke]

i really think that's the best you can do with 64 bit float. as far as gpu code thats more of a workstation gpu feature  (gaming gpus have intentionally gimped fp64 implementations). if you are designing a product for real astrodynamics, you are going to probibly need a workstation gpu and some really exotic multi-core processor with some serious instruction set extensions. pretty much supercomputer hardware.

so since your number format lacks a few orders of magnitude that are nice to have, you either have to use a really convoluted workaround, or simply scale things down so fp64 is good enough. not that these problems are impossible to solve or anything, its just you need extra steps that factor into your dev costs. since dev is probably the most expensive part of your dev team (of course you can spend a lot of money on voice acting now, eg if you hire keanu reeves to voice some kerbals). there is also considerable precedent for newtonian games scaling things down. newtonian is also a problem, you are going to want lagrangian or hamiltonian mechanics in a pro science sim product.

[Meecrob]

2 hours ago, PakledHostage said:

That sounds about as wishful as my hoping in 1986 that Santa Claus would bring me a G.I. Joe aircraft carrier for Christmas... didn't happen then and ain't gonna happen now.

Oh I know, I find it funny what people come up with. Like the guy doesnt have enough on his plate.

41 minutes ago, Exoscientist said:

 I remember asking on this forum once why Kerbal used Kerbin and Mun, rather the real Earth and Moon and other real planets in the Solar System. I remember being not really being satisfied with the answer. I think with the interest in universities and schools worldwide in spaceflight, there would be a great deal of interest in having a program where they could design real rockets for making real missions anywhere in the Solar System. 
 

  Bob Clark

 

I totally agree. This is a kids game. We need an actual spaceflight simulator. With Reentry the game for the IVA's.

[AVaughan]

3 hours ago, Exoscientist said:

I think with the interest in universities and schools worldwide in spaceflight, there would be a great deal of interest in having a program where they could design real rockets for making real missions anywhere in the Solar System. 

KSP 1 + Realism Overhaul + Principia 

[darthgently]

21 minutes ago, AVaughan said:

KSP 1 + Realism Overhaul + Principia 

+ kOS (kerbal operating system)

[cubinator]

10 hours ago, HebaruSan said:

Isn't that a much, much smaller market?

Definitely smaller. We are comparing the groups of people "I think rockets are cool and I like to dream about them" with "I build real rockets for a living at the rocket company".

Real rockets are a lot more complicated than KSP rockets and you're really going to be using a multitude of different softwares to simulate and evaluate them for real work, there's no one program that'll do it all and I think KSP wouldn't be a very good base for such a thing.

[Meecrob]

1 hour ago, cubinator said:

I think KSP wouldn't be a very good base for such a thing.

Spot on! We need a re-write fron the ground up. No point doing it unless we do it properly!

[Gargamel]

17 hours ago, Exoscientist said:

 I remember asking on this forum once why Kerbal used Kerbin and Mun, rather the real Earth and Moon and other real planets in the Solar System. I remember being not really being satisfied with the answer. I think with the interest in universities and schools worldwide in spaceflight, there would be a great deal of interest in having a program where they could design real rockets for making real missions anywhere in the Solar System. 
 

  Bob Clark

 

Probably because they built it piecemeal from the ground up.    They started with a Lego rocket builder, with arbitrary masses and sizes, and had it work in a 2d world, using earths gravity.   Then they realized they needed a globe to fly off of, and scaled that to match the parts, instead of retooling all the work they previously did.    By the time anybody said “Hey, this and this are not right”, it was too late to go back and fix them without a complete rebuild.    It’s amazing how far they actually got without ever having to restart from scratch with all the knowledge they gained.    
 There’s posts around here somewhere detailing the old design story.    I think even @Moach had posted a pic of the literal back of a napkin sketch they laid out for the process.  

[kerbiloid]

7 hours ago, Gargamel said:

They started with a Lego rocket builder, with arbitrary masses and sizes, and had it work in a 2d world, using earths gravity.   Then they realized they needed a globe to fly off of, and scaled that to match the parts, instead of retooling all the work they previously did.

And they saw that it was good.

7 hours ago, Gargamel said:

By the time anybody said “Hey, this and this are not right”, it was too late to go back and fix them without a complete rebuild.

Exactly that story.

[Exoscientist]

21 hours ago, AVaughan said:

KSP 1 + Realism Overhaul + Principia 

Could that for example take the specs for the Delta IV and Atlas V rockets and accurately output the payloads to LEO, GTO, etc.? The Silverbirdastronautix.com payload estimator can do that within about 10% accuracy I’ve found. But it does have wide cited error bars, i.e.,  you can only be sure of the answers within that rather large error range.

 Another thing I want to estimate is fast flights to the Moon or Mars.  For instance Elon once said the expendable Starship might have a dry mass of only 40 tons. That combined with a 380s Isp would allow 1 month flights to Mars with just chemical propulsion:

 That is just an estimate though based on the idea the departure speed is so fast the trajectory would be almost straight-line. It would be nice to have an actual trajectory sim that would calculate this more definitively.

With departure speeds that high it should mean near straight-line trajectories from LEO to the Moon also, with  greatly reduced travel times. We might want this with commercial flights to the Moon by, for example, SpaceX  coming into play in  a near term time frame. With just a 3.1 km/s escape velocity, we can reach the Moon within 3 days. How fast from LEO to lunar orbit with, say, 12 km/s delta-v available, taking into account you also have to slow down? Could the travel time be reduced to hours?

  Bob Clark

[darthgently]

4 hours ago, kerbiloid said:

And they saw that it was good.

Exactly that story.

Juno (simple rockets 2) seems on a similar track.  Eerily similar.  The Minmus correlate is called T.T. with no obvious explanation of what that stands for.  Hmmm

[PakledHostage]

16 hours ago, Gargamel said:

 There’s posts around here somewhere detailing the old design story.    I think even @Moach had posted a pic of the literal back of a napkin sketch they laid out for the process.  

A long time ago (but not in a galaxy far away), there was a massive forum glitch that took down the whole KSP forum for about 3 weeks. In the process of restoring it, about 6 months of forum history (as well as everyone's rep count) was lost. But I remember seeing that image of the napkin sketch too... Maybe that post got lost in that great forum purge? (I recently went looking for the Chelyabinsk meteor impact thread, but couldn't find it; I think it too was a victim.)

[Gargamel]

18 minutes ago, PakledHostage said:

A long time ago (but not in a galaxy far away), there was a massive forum glitch that took down the whole KSP forum for about 3 weeks. In the process of restoring it, about 6 months of forum history (as well as everyone's rep count) was lost. But I remember seeing that image of the napkin sketch too... Maybe that post got lost in that great forum purge? (I recently went looking for the Chelyabinsk meteor impact thread, but couldn't find it; I think it too was a victim.)

No, my first account got lost in that crash of ‘13, created this clone shortly after.    I remember seeing that post about the drawing after I became a moderator.    

[AVaughan]

On 7/11/2024 at 4:52 PM, Exoscientist said:

Could that for example take the specs for the Delta IV and Atlas V rockets and accurately output the payloads to LEO, GTO, etc.? T

With the specs you can build a replica of the real rocket, with the correct engine config, including thrust, mass, isp and fuel mixture.  You should also be able to give each stage with the correct mass fraction and hence replica rockets should perform pretty close to the real rockets.  I don't personally build replicas, so I'm not sure how close their performance stats are to the originals, but my guess would be less than 10% out (assuming the you start from accurate stats).   But the best answer is to install RO, and build a replica and compare its performance to the real life rocket.    https://github.com/KSP-RO/RP-1/wiki/RO-&-RP-1-Express-Installation-for-1.12.3  RO/RP-1 changes a lot of KSP's mechanics, so expect a steep learning curve, even without Principa.  The best place to ask questions is the KSP-RO discord.  (There is a link from the wiki page above). 

Principa is installed separately, and only really relevant if you want n-body gravity and orbital perturbations, rather than orbits on rails with KSPs sphere of influence model.  https://github.com/mockingbirdnest/Principia/wiki

[kerbiloid]

Spoiler

 

Found by google "kerbal moach kerbo".

[DerekL1963]

On 7/9/2024 at 7:08 AM, Exoscientist said:

Couldn’t more exposure and revenue be generated by marketing it as an actual commercial sim for spaceflight to orbit and interplanetary trajectories?

If KSP, in any flavor, was a reasonably accurate simulation...  That would be a reasonable question.  KSP isn't a reasonably accurate simulator.  It isn't even close.  (And even if it was, the market for such things is pretty small.)  As far as game value goes, as fidelity (and difficulty) increases the potential market decreases almost exponentially.

Seriously, very few KSP players grasp just how simplified and paint-by-numbers KSP actually is.