Kerbal Space Program 2 to be released in 2022 [Discussion Thread]

[Vanamonde]

Some posts having nothing to do with KSP have been removed. Please stick to the topic at hand. 

[Zozaf Kerman]

I don’t know if this observation is correct, but the absence of a show/tell or dev diary today could mean that this is the “quiet before the storm.” This means that the next video released will feature a release date or a push to 2023/2024. Thoughts?

[saxappeal89129]

Prior to release of KSP 2... and to try to spin some positivity in this room.... wondering/hoping the Dev team can add some atmosphere to the game before release... and a good start would be this 'sonification' of a black hole in Perseus...

https://www.nasa.gov/mission_pages/chandra/news/new-nasa-black-hole-sonifications-with-a-remix.html

We live in amazing times, with Musk reaching for the stars, and Cap'n Kirk finally going into space... exciting for people of a scientific mind, especially when we're seeing youtube episodes of a supposed "warp bubble" being created.

Love space, but I'll never get to go, thus the reason I'm so excited for KSP2!

Rock out with your rooster out!

[Bej Kerman]

On 5/5/2022 at 8:41 AM, Vl3d said:

I think the anger is because it's been one month since the Interstellar video came out and it's been complete radio silence since.

You only need a bit of patience. Just don't let your life revolve around hype.

[saxappeal89129]

3 hours ago, shdwlrd said:

I'm guessing you're talking about Starfield? I get that. All I can figure out is it's a space-based RPG. Ok, but what's the story? They have you hunting a "mystery." Ok, but why? Nothing. At least we know the ideas behind KSP. The systems behind the game may be unknown at the moment, at least we know what it's about.

Yeah but "RPG" can mean a lot of things to many different people. Some "Adventure" games are considered RPGs, even though there are no stats related to character progression etc; this is actually the impression that I got was that this was more of an adventure/discovery game, rather than Fallout4/Skyrim (in space), or like Outer Worlds. I could be wrong, but again the impression that I got of Starfield was more of a No Man's Sky / Puzzle scenario (nothing negative about No Man's Sky, it's turned into a fine game, with fun content)

I think this is what makes KSP1/2 so incredibly unique, and sets itself apart even from Space Engineers and Empyerion. KSP is really in its own genre at this point in time (a couple devs have tried to emulate unsuccessfully). KSP is all about the struggle to get into space, and a relatively accurate depiction thereof (without needing a supercomputer for true physics calcs and orbital mechanics). Space Engineers and Emperion are the "fantastical" depictions of where we could be in a couple hundred years... Kerbal is a depiction of where we could be in our lifetime if resources are dedicated to it. This is one of the reasons I stopped responding to the people in this very thread that are negative about the game and trying to use their force of will to create a minority voice to change the base formula of the game. It's not about "seeing large starships undock in orbit while viewing from the surface of a planet"... that's just not the aim/goal of Kerbal, it's about "near future experiences" if we have the vision and drive to get there.

Edited May 6, 2022 by saxappeal89129

[whatsEJstandfor]

I don't think it's safe to assume anything like that quite yet. There are still 7 months left in 2022; plenty of time to stage a long campaign between announcement and release. Supposing they launch in November, I don't think they'd start feeling obligated to announce the release date until June or July

[Zozaf Kerman]

Based on the size of intercept, I don’t think they could afford to have a 3-7 month campaign before ksp2 releases. It would most likely be 1-2 month at most before release.

[whatsEJstandfor]

Wasn't your original post suggesting that they're about to announce the release date and, therefore, have up to a 7 month campaign?

[t_v]

If you want precedents, it is not uncommon for no news to happen. News releases every 4 to 6 weeks, we just got one increase in pace this year and otherwise, throughout the first three months there has been one or less piece of news per month. After the last feature video, there was an 8 week stop to news, so I’d say this is to be expected. Plus, there is some speculation that the twitter feed just posted a screenshot at max graphics settings, so maybe we actually did get news, and that screenshot is also a good way to reach more people and start generating hype, so there is a chance this is the start of the marketing campaign. 

[Vl3d]

17 minutes ago, t_v said:

that screenshot is also a good way to reach more people and start generating hype,

455 likes on Twitter, 1080 likes on Instagram, 135 likes on Facebook. So much hype!

[gussi111]

4 minutes ago, Vl3d said:

455 likes on Twitter, 1080 likes on Instagram, 135 likes on Facebook. So much hype!

It was posted 6 hours ago?

[ballisticfox0]

1 hour ago, Zozaf Kerman said:

I don’t know if this observation is correct, but the absence of a show/tell or dev diary today could mean that this is the “quiet before the storm.” This means that the next video released will feature a release date or a push to 2023/2024. Thoughts?

Just a friendly reminder that the KSP 2 team doesn't have to give us any news.

 

[LHACK4142]

Hmm? What screenshot?

[Pthigrivi]

My feeling is they've been in the buckle-down stage since last fall and they'll be in it until they've got things nice and ironed out and stable and feeling confident enough to stick to a release date. Ive said it before but all that matters is that the game is good. Im happy to see them focus on that first and foremost. 

[DeadJohn]

5 hours ago, Zozaf Kerman said:

I don’t know if this observation is correct, but the absence of a show/tell or dev diary today could mean that this is the “quiet before the storm.” This means that the next video released will feature a release date or a push to 2023/2024. Thoughts?

Baseless conjecture. Maybe it will be released to great fanfare soon, maybe there are delays, maybe there's a catastrophe. It's just a game, one I look forward to, but there's no need to obsess over what-ifs.

[stephensmat]

Also, I recall hearing Financial Year, 2022, which could take us halfway into next year. Even if it's this year, we've still got more than six months left.

[Guest]

10 hours ago, Zozaf Kerman said:

This means that the next video released will feature a release date or a push to 2023/2024. Thoughts?

A combination of tarots and crystal ball scrying tells me that the game will release in 2042.

My prediction is just as accurate as any made here in this thread and others.

[Guest]

12 hours ago, saxappeal89129 said:

Yeah but "RPG" can mean a lot of things to many different people. Some "Adventure" games are considered RPGs, even though there are no stats related to character progression etc; this is actually the impression that I got was that this was more of an adventure/discovery game, rather than Fallout4/Skyrim (in space), or like Outer Worlds. I could be wrong, but again the impression that I got of Starfield was more of a No Man's Sky / Puzzle scenario (nothing negative about No Man's Sky, it's turned into a fine game, with fun content)

I think this is what makes KSP1/2 so incredibly unique, and sets itself apart even from Space Engineers and Empyerion. KSP is really in its own genre at this point in time (a couple devs have tried to emulate unsuccessfully). KSP is all about the struggle to get into space, and a relatively accurate depiction thereof (without needing a supercomputer for true physics calcs and orbital mechanics).

There's a whole lot of confusion going on for both the "RPG" definition, and the idea of "Space games" being a genre instead of just a setting.

It's incredible how a game being Fantasy immediately makes it considered "RPG" while the same rules and gameplay style in a Cyberpunk or Urban setting brings up other comparisons like GTA.

The same goes for Space Games, KSP is its own genre, but even comparisons between games like NMS and Elite: Dangerous make little sense, NMS has more in common with Minecraft than it has with ED, and ED is more similar to a mix between Sea of Thieves and Eurotrack Simulator than it is to any other space game.

 

[Gargamel]

This thread has been merged with the master thread.  

[t_v]

12 hours ago, LHACK4142 said:

Hmm? What screenshot?

Here is the image as well as the post I was referencing. Keep in mind, this is all particularly wild speculation, and there is not any solid supporting evidence that this is a screenshot. 

 

[SciMan]

So, I see a bit of an elephant in the room that nobody's acknowledging.

Orbital mechanics. Or, "Why does this game use patched conics".

It's not because of CPU load.
To prove that last (seemingly controversial at least on this forum) statement, I present another space flight simulator, set in the real solar system, with real N-body orbital mechanics.

Behold, Orbiter Space Flight Simulator. (Hail Probe!)

A game that got a notable UPDATE in 2006, and was ORIGINALLY designed to use DirectX 7.0 as the rendering engine.

I was in high school back then, and one day I decided to download it onto a flash drive at home and bring it into school with me so I could do something while I was supposed to be being bored to death by the intricacies of the English language (aka working on an English essay). The library had about 20 laptops available for use. Granted these were School owned laptops, so they were very much "just a laptop" type Dell machines, not at all a "gaming focused" laptop of any kind. I think the maximum clock speed of the CPU was like 2.4 Ghz at most, and it only had a single processor core to work with, and maybe 2 GB of RAM.

Do you want to know how many FPS I was getting with several hundred objects simultaneously in orbit (since I brought my save games from home with me)?
(Rhetorical question, I'm going to tell you anyways.)

Six. Hundred. Yes, I said that correctly. And that's at max time warp, where a full Earth year passes within maybe 10 seconds (IIRC, it's been a while), with SEVERAL HUNDRED INDEPENDENTLY ORBITING ARTIFICIAL SATELLITES, with the physics settings selected to even have gravity gradients make a difference in the craft's attitude, and the various mass concentrations of the Earth and Moon accounted for.
600 frames per second.
And that's probably because of some hard-coded limit within DirectX, and not a limitation of the CPU (because back then 60 or 75 hertz was all you were going to get, and that's from a CRT).So, again, that's

SEVERAL HUNDRED INDEPENDENTLY ORBITING ARTIFICIAL SATELLITES, with ALL KINDS of so-called sophisticated physics being calculated, all at 600 frames per second.

Mind you, Orbiter considers every non-planetary thing to be a single physical "object" and does not simulate collision physics between artificial satellites (so things can go thru one another unless they interact via docking ports or hit the surface of a planetary body), but that shouldn't matter to the statement "N-Body physics is too hard to calculate quickly", which is the most commonly brought up argument in favor of not including it in KSP or a KSP sequel.

There are only two simplifying assumptions you need to make to enable this.
The first is that the orbits of the planets themselves remain fixed and unchanging aside from the planet's progression along them (No, you can't deorbit an asteroid or moon in Orbiter. But you can't deorbit celestial bodies in KSP either, so that assumption is already fixed in place in both games).
The 2nd assumption you need to make is that artificial satellites (such as player-created ships) do not create their own gravitational fields, merely follow the path set to them by their current velocity vector and the gravitational forces applied to it by all the celestial bodies in the solar system.

Orbiter does the math that this requires by using one or more Runge-Kutta methods to solve the orbital dynamics calculations.
That is indeed somewhat complex math, but like I said, if Orbiter (a game that existed almost 15 years ago) could solve that math, on a one-lung single core CPU, at SIX HUNDRED FPS, I really don't think the math is what's gonna hold up N-body not being in KSP 2.

Now why IS it not in KSP 2? I couldn't tell you. But I CAN (and hopefully just have) tell/told you that "math too hard for CPU" is NOT the reason.

[TheOrbitalMechanic]

10 minutes ago, SciMan said:

I really don't think the math is what's gonna hold up N-body not being in KSP 2.

Now why IS it not in KSP 2? I couldn't tell you. But I CAN (and hopefully just have) tell/told you that "math too hard for CPU" is NOT the reason.

They've already elaborated some about their thoughts on N-body here: 

The station-keeping concerns are my main issue. I don't care about realism enough to want to have to check in on every single spacecraft I own every other orbit to make sure they're still doing what they're supposed to. KSP 2 is supposed to be a fun game, not a hardcore simulator, and besides, everything is at 1/10th scale anyway, so why draw the line here?

(Not to mention, if they were to introduce N-body physics, they'd have to overhaul the Jool system as well.)

[blankets]

26 minutes ago, SciMan said:

Now why IS it not in KSP 2? I couldn't tell you. But I CAN (and hopefully just have) tell/told you that "math too hard for CPU" is NOT the reason.

I'm sure Principia will be carried over to KSP 2 shortly after it releases if just N-body physics is your concern. I believe the devs mentioned something about N-body around Rask and Rusk, so it will be present.

The reason for not replacing patched conics with N-body is probably down to most players not wanting it.  Its a bit of a chore keeping satellites in orbit, for example.

[Vanamonde]

If at some point KSP2 is offered through the Epic store, the pros and cons of that distributor can be discussed. Until then, though, the Epic store is irrelevant to this discussion and those comments have been removed. Please stay on-topic. 

[SciMan]

@TheOrbitalMechanic

I understand your concerns regarding not wanting to have to micro-manage your satellites to make sure they're doing stationkeeping as you want them to do.
I fully agree with this, the player should not be required to actively visit each of their (likely many) satellites in their orbits simply to make needed adjustments to their orbits.

However, KSP 2 is already including a feature that I think may solve the problem without requiring N-Body physics to be abandoned entirely (except for Rask and Rusk).
The "automated supply run" feature, if it works more like a "vessels can be smart enough to correct their trajectories to arrive at their destination even if the maneuver burns weren't perfect" type of thing and not a "non-physical ships being simulated as just a flow of resources" thing instead, well there's your solution to station-keeping. Simply set the "departure" and "arrival" orbits to be identical and the automated supply run feature would probably interpret that as a "keep this thing in this orbit" instruction, in other words it would allow you to automate station keeping.

Additionally, if you can set conditions for other actions to happen besides just keeping a course, you could set it up so that satellites that are almost out of propellant would execute a pre-planned de-orbit burn or send themselves to a "graveyard orbit" with new satellites being automatically launched to replace them, or alternatively they could signal to send for a vessel to refuel the existing satellite (assuming you put a compatible docking port on it when you built the thing).

This would replicate quite well the #1 reason old satellites are decomissioned IRL (lack of fuel) without bothering the player beyond setting it up (unless they want to get involved in it manually, of course).
The #1 reason IRL satellites are decomissioned or go adrift is that they simply run out of propellant to be able to keep their "orbital slot" in geostationary orbit (like most geostationary communications satellites), or lose the ability to maintain orbital altitude if in LEO (like Starlink satellites are designed to gracefully deorbit themselves if they malfunction, since that large solar panel creates quite a lot of atmospheric drag (which is a thing at those low orbital altitudes)).

So while your concerns are valid, I think the solution is already in existence (tho it might need some minor modifications).

EDIT: And yes, of course, they would have to increase the orbit radius of most/all of the moons of Jool (and any other gas giant with similarly close-orbiting high mass moons).
However as far as I'm concerned that doesn't constitute a "change" to the KSP 1 solar system so much as it constitutes a change to make the physics make sense (because for a game that has a lot of realism in its orbital mechanics, the orbital configuration of Jool and its moons is not long-term stable AT ALL).

EDIT 2:
This is probably the SINGLE concession to realism that I think should be made in KSP 2, because it is the one that is most impactful to the "fly" aspect of orbital mechanics. Not life support, not random (or not so random) part failures, not "removing metallic hydrogen from the game because that doesn't work like that", none of that matters as much as N-Body physics (and along with that having the solar system have a configuration that is compatible with N-Body physics).

But then again, Orbiter ALSO solved the orbital drift problem. You see, it had another option for the physics, "orbit stabilization", which does pretty much what my "automated station keeping" idea does, but it does it fully automatically. I'm not sure how it works, maybe it just nulls out the influences of the gravity of the other celestial bodies when in a low orbit around a planet or moon (that's my best guess), but it sure did work to keep the ISS where it belongs around Earth without me having to do anything to it (back when I played Orbiter that is).

So maybe the best approach to N-Body for KSP is for it to only have an effect when it really makes a difference, which for my purposes would be "twice the orbital radius of geostationary orbit on a non-tidally locked planet or moon".

Because being able to park a satellite or an orbital VAB at someplace nice and stable like L4/L5 (or in an orbit around another lagrangian point, such as L1/L2) is just such a nice thing. Yeah the lagrangian points tend to be quite distant from a given planet, but there's just so many interesting things you can do with them, like the mentioned satellite or station (or colony) parking spots, or using them to create a "grand tour" trajectory that use a tiny amount of propellant because the assists all basically pass thru lagrangian points of various planetary bodies and moons (there's plenty more you can do with them, I'm just scratching the surface).

I just don't want a player to find a neat trick that NASA used with regards to orbital mechanics, and be unable to replicate it (or disincentivized from replicating it) because of KSP 2's simplified orbital physics modelling.

EDIT 3!, Give it up for Edit 3!:
You know, additionally, without non-spherical gravity sources, you can't even get a proper Sun-Synchronous Orbit (which is the ideal orbit for practically any kind of mapping satellite). Instead, you're forced to wait for the planet to turn under you, which if it's tidally locked will be a very long time indeed.
Now that might only matter if we get ScanSat-like functionality in KSP 2, but I see no reason we shouldn't get something like that. Using a satellite to map a planet is a thing that takes some amount of time significantly longer than a player would be likely to just wait for, instead they'd be doing something else in the background, it really shouldn't work like the ISRU orbital resource scanner (the one you put in a polar orbit) works at all, that thing's far too quick.

Edited May 9, 2022 by SciMan