KSP2 EA Grand Discussion Thread.

[ConsoleCoder]

I’m getting about 40-20 FPS on average. It can go down to like  5 FPS sometimes, but that’s relatively uncommon. I would say 2% of my gameplay is a bug or intense lag. The otter 98% is my game running at 30 FPS with no bugs. For some reference, I’m running KSP2 on a GeForce GTX 1660 ti. KSP2 is currently my favorite game, even more then KSP1. I don’t mind the lack of features, cause I’m not a professional. But I’m also not a noob. Been planing KSP1 for about 1-2 years, so I’m pretty decent at it. My highest delta-v was 27,000 if you don’t have a payload attached, but that’s with the advent of hydrogen engines. In KSP1, I would play sandbox mode anyways, and never really did any ore mining or anything, so the lack of these features in KSP2 is really no different to my normal KSP1 gameplay, except for the obvious stuff like lag, bugs, and hydrogen engines.

[didkodidko]

Any news on 0.1.5? Wen?

[KUAR]

10 minutes ago, didkodidko said:

Any news on 0.1.5? Wen?

The last stated plan was 6-week intervals between releases. I've no idea if that's still the target. If so, that would place 0.1.5 on or about Friday 13 October.

[stephensmat]

I'm waiting for someone to say something about CM Nerdy_Mike's comments on the Discord Dev Tracker. Am I looking in the wrong threads? 

[Superfluous J]

11 hours ago, KUAR said:

Friday 13 October

How appropriate

[Alexoff]

I think 0.1.5 will be at the very end of October or beginning of November, and then by Christmas the developers will try to present the science. If science is also very early access, then the option of stopping development is possible

[Astr0Guy5]

4 hours ago, Alexoff said:

If science is also very early access, then the option of stopping development is possible

Hey, Rome wasn't built in a day. I mean, KSP2 at its full potential will prove to be one of the greatest, most complex games ever developed. Patience we must have, hmm.

[Alexoff]

5 minutes ago, Astr0Guy5 said:

Hey, Rome wasn't built in a day.

It's a survivor's mistake.

6 minutes ago, Astr0Guy5 said:

KSP2 at its full potential will prove to be one of the greatest, most complex games ever developed

I prefer not to have my head in the clouds, but to look at what has been done since 2017, when development of the game began.

[uglyduckling81]

I'm blown away by how dead these forums are.

Ksp2 has really done a number on ksp franchise popularity.

[Royalswissarmyknife]

I agree

It just feels empty and depressing.

[Turbo Ben]

What is there to talk about that hasn't already been said?

There's nothing new in KSP2 and hardy anyone is playing it. There isn't even anything from the devs to discuss.

 

[Royalswissarmyknife]

15 minutes ago, Turbo Ben said:

What is there to talk about that hasn't already been said?

There's nothing new in KSP2 and hardy anyone is playing it. There isn't even anything from the devs to discuss.

 

The issue is that Ksp-1 also feels dead.

Edited September 23, 2023 by Royalswissarmyknife

[Vanamonde]

Overlapping threads have been merged. 

[Alpha_star]

On 9/14/2023 at 11:58 PM, ConsoleCoder said:

I’m getting about 40-20 FPS on average. It can go down to like  5 FPS sometimes, but that’s relatively uncommon. I would say 2% of my gameplay is a bug or intense lag. The otter 98% is my game running at 30 FPS with no bugs. For some reference, I’m running KSP2 on a GeForce GTX 1660 ti. KSP2 is currently my favorite game, even more then KSP1. I don’t mind the lack of features, cause I’m not a professional. But I’m also not a noob. Been planing KSP1 for about 1-2 years, so I’m pretty decent at it. My highest delta-v was 27,000 if you don’t have a payload attached, but that’s with the advent of hydrogen engines. In KSP1, I would play sandbox mode anyways, and never really did any ore mining or anything, so the lack of these features in KSP2 is really no different to my normal KSP1 gameplay, except for the obvious stuff like lag, bugs, and hydrogen engines.

Yes please. I had tons of fun playing with my 3070 E-GPU and i7-8700K. The frame rate never gets to anywhere below 20 in my starship moonship replica mission and is steadily above 25 for another fighter jet creation I have done. Still, it’s far away from the promised “playable without a dedicated GPU”, but I’d say the frame rate is 2 times higher than in the initial version.

[THE_KERBAL]

Most of KSP 2’S development cycle was working on everything but the core game itself.  Leaving no time for reentry or a proper rendering engine or code optimization. In the original game files for 0.1.0 there were multiple leftovers from colonies showing that they were mostly complete. There were also leftovers from interstellar travel and small bits of multiplayer code. The early access release had to be put together with an unfinished base game in less than a year most likely underfunded. This is Take Two we are talking about. They are a not good company. So be Patient and wait. They already have a lot of work done on colonies, interstellar and multiplayer. The hardest hurdle for the development team will be finishing the base early access up to colonies.

Edited September 24, 2023 by THE_KERBAL
Adding some stuff to my message

[MarcAbaddon]

14 hours ago, THE_KERBAL said:

Most of KSP 2’S development cycle was working on everything but the core game itself.  Leaving no time for reentry or a proper rendering engine or code optimization. In the original game files for 0.1.0 there were multiple leftovers from colonies showing that they were mostly complete. There were also leftovers from interstellar travel and small bits of multiplayer code. The early access release had to be put together with an unfinished base game in less than a year most likely underfunded. This is Take Two we are talking about. They are a not good company. So be Patient and wait. They already have a lot of work done on colonies, interstellar and multiplayer. The hardest hurdle for the development team will be finishing the base early access up to colonies.

I doubt that's true and I had a look at the code myself back then. There are some things about colonies (mainly delivery routes) in, but I wouldn't call it mostly complete. The multiplayer stuff was mostly standard lobby stuff, nothing super specific. From the latest sets of Nertea it seems there is quite a lot of work left to do on colonies.

The state of the non-implemented features is one of the largest unknowns at the moment for me, and it's one of the reasons why some people don't have a lot of trust left, since there were interviews where the devs stated the new features were done and now it was just about putting them all together... 

On a different note Take Two remains the company who has the final say. Even if you trust the original team, as long as you think Take Two is completely untrustworthy there's no reason to be very optimistic.

[GGG-GoodGuyGreg]

18 hours ago, THE_KERBAL said:

Most of KSP 2’S development cycle was working on everything but the core game itself.  Leaving no time for reentry or a proper rendering engine or code optimization. In the original game files for 0.1.0 there were multiple leftovers from colonies showing that they were mostly complete. There were also leftovers from interstellar travel and small bits of multiplayer code. The early access release had to be put together with an unfinished base game in less than a year most likely underfunded. This is Take Two we are talking about. They are a not good company. So be Patient and wait. They already have a lot of work done on colonies, interstellar and multiplayer. The hardest hurdle for the development team will be finishing the base early access up to colonies.

I kept seeing this narrative shortly after launch, and less and less since then, and unfortunately I’m not smart enough to take a look in the code and judge for myself if it’s like that, but I do find this scenario intriguing to this day.

I just see a small flaw in this scenario.

 

Do you mean to say that they could get colonies done but the re-entry heating is what’s been giving them issues for the past months?

 

Then us skeptics ( or realists as I like to call it ) are either going to be eating our words very soon or the people responsible for the done features have moved on to other projects. 

[The Aziz]

It's possible that they went into implementing as much as possible (colonies, progression, interstellar mechanics, planets, resource processing etc), but as a result they didn't have enough time left for basics. Once the workflow switched to EA roadmap, they had to cut most of the game out of the game, and realized they don't have much left because they weren't focusing on it at all. I see it very well in a game I do the QA for, but unlike KSP, they are aiming for full release and I can assure you, they're in panic mode now.

[GGG-GoodGuyGreg]

All I’m saying is if they had the winning card (core game that ties together all those supposedly complete features) they would have used it sooner than later, as their public image is going downhill since launch. 
 

And even if this is what they are doing now, they are stumbling, and if it’s going to be done anytime soon, it definitely won’t redeem them as much as if they would have done so earlier. 

Edited September 25, 2023 by GGG-GoodGuyGreg

[Royalswissarmyknife]

1 hour ago, The Aziz said:

I see it very well in a game I do the QA for, but unlike KSP, they are aiming for full release and I can assure you, they're in panic mode now.

 

Spoiler

Please dont be Cities: Skylines 2

 

[PDCWolf]

On 9/2/2023 at 7:36 PM, PDCWolf said:

So, they have this seemingly inefficient system that doesn't scale well with partcounts, they're really trying to stop us from lego-ing solutions to certain stuff like we would in KSP1, and they also have a clear aversion to tell us the partcounts they're aiming for...

Once again my nose picks up a sour smell.

Once again, my nose was right. I'm not gonna bother linking the corresponding posts, because this reddit post has them already. Now I invite you to ask yourselves genuinely, if their optimizations so far have come 90% of the time from culling, deleting, hiding and downgrading stuff on screen rather than proper progress on the performance front... What's the timeframe for the obvious "rework" this system is gonna need?

Every single word that comes out of them turns this into more and more of an amateur mess. And yes, I do finally feel using that word is appropriate.

Edited October 5, 2023 by PDCWolf

[cocoscacao]

On 9/25/2023 at 1:50 PM, The Aziz said:

It's possible that they went into implementing as much as possible (colonies, progression, interstellar mechanics, planets, resource processing etc), but as a result they didn't have enough time left for basics.

Highly unlikely... I believed that initially... in February... but even according to recent AMAs, that's not the case. Aside from multiplayer (which I still think is irrelevant), code for the core loop is the hardest part. The rest are just parts and (relatively) straightforward calculations.

[VlonaldKerman]

On 9/25/2023 at 7:50 AM, The Aziz said:

t's possible that they went into implementing as much as possible (colonies, progression, interstellar mechanics, planets, resource processing etc), but as a result they didn't have enough time left for basics

Preface: I’m not a game dev…

BUT I feel like things like the joint physics system are so fundamental that you would be worried about those first? Like, why would I try and figure out how the game stores information about active vessels and their orbits in order to have persistent resource consumption and physics applied if I don’t know how part joints work? Maybe that example is bot actually a problem but in general I’d have to imagine that an ideal dev arch follows the (fundamental systems) —> gameplay implementation/system integration —> cosmetics and optimization timeline, perhaps with cosmetic aspects like part designs and music being developed at the same time as the technical stuff.

This is what I’ve made several posts (and threads) about already: the problem isn’t that there is a bug that causes rocket wobble, the problem is that rocket wobble will probably be addressed by implementing exactly the kind of back-end fundamental system upgrade that the entire game is predicated on, and that should have been done AT THE BEGINNING! These backend systems are what will make or break the game- modders can add in all the fancy parts they want, but it will be as buggy as modded KSP 1 unless the backend of the game is totally fresh.

[Periple]

11 minutes ago, VlonaldKerman said:

Preface: I’m not a game dev…

BUT I feel like things like the joint physics system are so fundamental that you would be worried about those first? Like, why would I try and figure out how the game stores information about active vessels and their orbits in order to have persistent resource consumption and physics applied if I don’t know how part joints work? Maybe that example is bot actually a problem but in general I’d have to imagine that an ideal dev arch follows the (fundamental systems) —> gameplay implementation/system integration —> cosmetics and optimization timeline, perhaps with cosmetic aspects like part designs and music being developed at the same time as the technical stuff.

This is what I’ve made several posts (and threads) about already: the problem isn’t that there is a bug that causes rocket wobble, the problem is that rocket wobble will probably be addressed by implementing exactly the kind of back-end fundamental system upgrade that the entire game is predicated on, and that should have been done AT THE BEGINNING! These backend systems are what will make or break the game- modders can add in all the fancy parts they want, but it will be as buggy as modded KSP 1 unless the backend of the game is totally fresh.

Unity is packaged with relatively feature-complete game physics, including rigid-body physics, which are good enough for most purposes. Replacing that with something else, let alone your very own implementation, would be a very big and very risky undertaking that requires some highly specialized skills. So unless there's a REALLY good reason, you will want to work with the physics system that's already there, plugging in what works and finding workarounds for what doesn't. This is the case with joints. 

So their options were:

1. Use Unity physics and find workarounds where necessary
2. Use Havok physics, pay for the extra middleware cost, and find workarounds where necessary
3. Make your own rigid-body physics

They clearly decided, very early on, to go with option 1. From there on out, this becomes a matter of finding a good-enough compromise in terms of gameplay feel and computational cost. There are lots of ways they could approach this and without trying them it's impossible to say for certain which one works best. They could do some variant of hidden autostrut. They could add more joints. They could have some kind of auto-welding system, or they could just make some joints rigid (e.g. joints between parts of equal diameter) and see how that feels.

In any case, I'm sure the problem can be solved, and while the choice to go with option 1 was fundamental, addressing this particular problem within the constraints of that choice isn't: it's a tuning/balancing/workaround/compromise problem just like many other similar ones.

[MarcAbaddon]

I agree with this take, but I also feel that if you decide to go with 1 from the start you shouldn't make claims about how you are going to or at least trying to slay the kraken. Another problem I have here is that I feel that it's something that should been tackled earlier in the Dev process. It's one of those central things you have to plan around early when you make another KSP using Unity physics. It'd be one thing if they tried a solution early on and it turned out a dead end, but it seems they are talking conceptually. Which again, seems a bit late.