What happened to the reworked core systems?

[JoeSchmuckatelli]

You need the ever cheerful can-do guy on the team.  Otherwise nothing gets done. 

I've done my share of being critical of the Comm Strategy and state of Release - but honestly, the last thing we need is a glum and apologetic Nate Simpson. 

He needs to be 'Yeah, okay, release wasn't great - we get that (and I'm not going into why it was) - but we have a plan it's gonna be great, hang in there!' 

[GGG-GoodGuyGreg]

25 minutes ago, RocketRockington said:

Example here.  You can read more of what he posted on reddit under that account.  Unfortunately I don't know if the hypey stuff posted to the Kickstarter still exists anywhere.  This is him hyping the PA team after he barely joined Uber Entertainment and had almost no basis for this.

And this outlines my main gripe with the way the communications are being made to us around the KSP2 Launch.

 

Here's an exercise where I replaced PA with KSP2, and please let me know if it still holds up. I think it does look like it came straight from the communication we are all familiar with in the context of KSP2, and my point is it lack a honest and personal touch, and a good communication is one that can't fit anywhere else if you just replace the name of the product being discussed. 

Quote

I don't know that I'm equipped to answer you in detail, but I can speak to the general vibe around Intercept Games. I can tell you that these guys eat, sleep, and breathe KSP2, and the only thing they care about is making the best game possible.

Having worked in games for a couple of decades, I do know that there is often an IMMENSE gap between a game as pitched and a game as delivered. This is actually often a good thing, because as the developers experiment with the raw materials of the game, they often discover new kinds of fun that they hadn't anticipated. They also find out that some things they thought would be awesome are either boring or game-breaking.

I can say that by broad industry standards, KSP2 is way closer to the promised product than almost any other game I can name. But of course, when you launch something on Steam EA, you're going to be held to a higher standard of consistency.

The guys who make KSP2 are very, very committed to making the best game possible. I think they may still surprise you.

 

PLEASE NOTE: I don't mean this as a personal attack to Nate or the KSP2, I just think it proves my point that the communication style is a bit impersonal even after all this time when others have pointed out it's not the communication style we expect.

Edited May 16, 2023 by GGG-GoodGuyGreg

[Ferio]

This was 3 years ago.

 

 

[RocketRockington]

4 minutes ago, GGG-GoodGuyGreg said:

PLEASE NOTE: I don't mean this as a personal attack to Nate or the KSP2, I just think it proves my point that the communication style is a bit impersonal even after all this time when others have pointed out it's not the communication style we expect.

It's not the communication style that bothers me.  It's because the communication itself is completely bogus.  PA was a flop.  Nate was talking up a team that dropped the ball, really badly.  Every time people put their faith in Nate's words that KSP2 was going to be this grand thing - eg that the core systems would be a strong foundation for KSP2 (back on topic, sorta),  all those years of saying, essentially, that it was going to be better than KSP1 in every way, has utterly no basis in reality. 

From the contents of his other posts, it seems he would say the same thing about literally any game/company he was working with.  And he made himself the face of KSP2. If you total up the face time of everyone else from Intercept, I doubt you'd have half as much comms as from Nate.  Heck, Squad never featured some guy over and over - acting like he was the second coming of Will Wright.  They just made the game and let it speak for itself.

[GGG-GoodGuyGreg]

18 minutes ago, RocketRockington said:

[...] Every time people put their faith in Nate's words that KSP2 was going to be this grand thing - eg that the core systems would be a strong foundation for KSP2 (back on topic, sorta),  all those years of saying, essentially, that it was going to be better than KSP1 in every way, has utterly no basis in reality. 

[...]

Also to be back on topic, I'm not an expert and can't say I know if the core systems were recycled or replaced by newer, better ones.

But what anyone can tell you is once you have a foundation of such core systems in place and decide to launch it like that, then it's those that you will use until you start over and build a new game.

You can't just plug in or replace core systems, if those were a poor choice to begin with, they will decide the future of the rest of the game.

Edited May 16, 2023 by GGG-GoodGuyGreg

[Alexoff]

3 hours ago, Ferio said:

This was 3 years ago.

 

 

Looking at the title of the video, for some reason I remembered an old joke. One chef comes up to another in a restaurant, shows him the visitors and says - "Look, they are eating THIS!" 

High fps for medium computers for huge crafts, first impressions for beginners, docking, working multiplayer, and the last question about 2022 and success. This is both funny and sad...

 Interestingly, it was for this reason that T2 made him the face of an advertising campaign?

Edited May 16, 2023 by Alexoff

[RocketRockington]

1 hour ago, Ferio said:

This was 3 years ago.

 

 

Oh man I forgot about that video.  Prime example of the sort of untruths Mr Simpson found so easy to share.  Goes beyond just the standard 'talking things up for PR'  Just utterly making things up, spinning it like it was working internally.  That video would be a huge sack of lies if it was posted today, much less whatever state KSP2 was in 3 years ago.

Edited May 16, 2023 by RocketRockington

[Spicat]

12 hours ago, Scarecrow71 said:

I have to disagree here.  While I do think that things look great from orbit or during fly-by, I have seen nothing to this point that made me think that the graphics or terrain were awesome.  Most disappointing were the rings on Dres; I specifically went to Dres to see the rings, and all I got was a few flat, grey lines spinning around the planet.  I know the argument here would be something about performance, but what I saw was a real letdown. 

I was specifically talking about the terrain, not ground texture, rings, or something else..

And what they did is not the "same thing" as ksp 1, that's just false. They did in fact improve the terrain as I said (but still need to be completed with a better ground texture).

For more info on what they did, here a talk at GDC about the terrain. One piece of information that can answer a part of the initial topic, I found it pretty interesting:

 

Edited May 16, 2023 by Spicat

[Alexoff]

7 minutes ago, Spicat said:

And what they did is not the "same thing" as ksp 1, that's just false. They did in fact improve the terrain as I said (but still need to be completed with a better ground texture).

He seemed to be talking about other planets, we still have a space simulator. And there are a lot of boring places on Kerbin. But the surface is certainly better than in KSP1. And parts have more polygons.

[VlonaldKerman]

41 minutes ago, Spicat said:

And what they did is not the "same thing" as ksp 1, that's just false. They did in fact improve the terrain as I said (but still need to be completed with a better ground texture).

 

Maybe I should edit my post and clarify: I am NOT referring to the new terrain maps/textures themselves, especially on Kerbin. I'm referring to the PQS/PQS+ system that the terrain uses to exist, which, in my EXTREMELY low-level understanding, is limited in some fundamental way. They repainted KSP 1, but they didn't change the underlying system which, if they did change it, could yield better, more performant results. And, they talked up their new terrain system as offering unprecedented close-up detail.

Additionally I would certainly say that up-close planetary textures themselves are underwhelming as currently implemented- for instance, Duna's "sand". For now, I'll stick with KSP 1 + Parallax.

Edited May 16, 2023 by VlonaldKerman

[Scarecrow71]

1 hour ago, Spicat said:

I was specifically talking about the terrain, not ground texture, rings, or something else..

And in my experience with KSP2, I have seen nothing anywhere that indicates the terrain has been updated or is better.  Every place I've been, whether it's on Kerbin, Dres, Mun, or any other planet, just looks flat and dull.  Sure, things look good from orbit.  But once on the ground?  Nope.

[Razor235]

3 hours ago, Scarecrow71 said:

And in my experience with KSP2, I have seen nothing anywhere that indicates the terrain has been updated or is better.  Every place I've been, whether it's on Kerbin, Dres, Mun, or any other planet, just looks flat and dull.  Sure, things look good from orbit.  But once on the ground?  Nope.

This is my big thing with the new terrain.  I first noticed it with Minmus.  The textures admittedly do look good, especially from space, but Minmus without cliffs is just a tiny blue Mun with an ice skating rink.  I miss seeing the huge plateaus on the horizon and the little slot canyons on the edge of the flats.  I miss the huge mountains on Kerbin that go into the stratosphere, realism be damned.

Edited May 16, 2023 by Razor235

[PDCWolf]

On 5/15/2023 at 8:19 PM, Spicat said:

That's a huge understatement of what they did, it's not because it's the same algorithm that we have the same thing as ksp 1. (The new terrain algorithm will probably just be for optimisation)

Ksp 2 terrain is a way better than ksp 1 terrain, those realistic mountains are my favorite to capture:

[media snip]

That was just to nuance the argument a bit, I mainly agree with you with the rest, a lot of work (or implementation) still needs to be made to explain KSP2 existence.

8 hours ago, Spicat said:

I was specifically talking about the terrain, not ground texture, rings, or something else..

And what they did is not the "same thing" as ksp 1, that's just false. They did in fact improve the terrain as I said (but still need to be completed with a better ground texture).

[media snip]

For more info on what they did, here a talk at GDC about the terrain. One piece of information that can answer a part of the initial topic, I found it pretty interesting:

I have to remind you that they themselves confirmed that what they're using is PQS+, which funnily enough is where the biggest performance bottleneck was found (terrain drawcalls), and that CBT is yet unimplemented. There's a good chance CBT was, in a very early form, present in the media they used on GDC (yet the tools chapter has them first mentioning that they were still on PQS+).

What you're seeing is what can be done by extending KSP1s terrain system and I use "extending" as the most positive word possible.

Edit to add: As someone who doesn't watch much of GDC, I find it funny that they talk of PQS as theirs, when they took it from some indie a brazilian guy and his coworkers made and made it work hundreds of times worse.

Edited May 16, 2023 by PDCWolf

[Dakitess]

Seeing the images posted by Spicat, I'm really not enjoying the visual. It looks to me like an old naked satellite view, with false colors, artefacts, over sharpened yet lacking details, no proper lightning, low poly or faked high poly. A huge texture being applied on top of an HeightMap, maybe this is how Engines are exactly working for game like this, but it's really give this feeling of something being on top of a poor topology like a satellite view. And other games are dealing with this way (way (way)) better so I do not see that as an excuse.

Details are very unequal, not homogeneous, you immediately see where the texture feels stretched or not completely loaded, or something else. And it's like the mountains are not real shapes.

Anyway, I don't like it, it's not up to 2020+ standard by any mean, even from orbit. Let alone when landed where everything fall off except on very specific location where it's crisp and nice, even though lacking micro - medium topology.

Haaa... Graphics and terrain were what I was waiting the most  I really don't see incoming new dev / tech to really improve this, apart for performance that might improve by 2-3 times if we are lucky, which won't be enough at all.

Edited May 17, 2023 by Dakitess

[PDCWolf]

7 hours ago, Dakitess said:

Seeing the images posted by Spicat, I'm really not enjoying the visual. It looks to me like an old naked satellite view, with false colors, artefacts, over sharpened yet lacking details, no proper lightning, low poly or faked high poly. A huge texture being applied on top of an HeightMap, maybe this is how Engines are exactly working for game like this, but it's really give this feeling of something being on top of a poor topology like a satellite view. And other games are dealing with this way (way (way)) better so I do not see that as an excuse.

Details are very unequal, not homogeneous, you immediately see where the texture feels stretched or not completely loaded, or something else. And it's like the mountains are not real shapes.

Anyway, I don't like it, it's not up to 2020+ standard by any mean, even from orbit. Let alone when landed where everything fall off except on very specific location where it's crisp and nice, even though lacking micro - medium topology.

Haaa... Graphics and terrain were what I was waiting the most  I really don't see incoming new dev / tech to really improve this, apart for performance that might improve by 2-3 times if we are lucky, which won't be enough at all.

Meanwhile any random planet in Space Engine, being completely procedural and made for most of its life by a single dude.

 

[Scarecrow71]

I'm going to start this post by saying that I have not yet played Juno.  All of the opinions I am about to express are based solely on the videos I have seen, which includes a few tutorials.  With that said...

 

What I am really confused on here - or, rather, a few things I am confused on, really - is how Juno is capable of doing everything that the KSP2 team says KSP2 should be doing but isn't.  A few examples:

1. Procedural...well, everything.  Multiple tutorial videos show tanks, engines, wings, and other parts as being completely procedural.  Yet, the general consensus here (from what I gather, anyhow) is that procedural things beyond wings would be too complicated or cause too much stress on the game.  How is it possible that one game has been able to accomplish this while at the same time the largest game in the space (no pun intended) can't?  Or, perhaps, won't?  The biggest example I could possibly show people is the tutorial on how to get to orbit (which I won't link to in order to keep the forum as clean of non-KSP videos as possible).  Google it and watch it; the person running the tutorial uses both procedural engines and tanks on the same craft.  And we can't get that here?
2. Custom scripting language for automation.  Apparently, you can use Vizzy to automate just about anything on a ship, which should be the same as the Action Groups here.  And the Action Groups here in KSP2 just don't seem to work very well (if at all).  I'm not saying one is better than the other, especially without having used Vizzy.  But it seems that Vizzy is far more intuitive and scalable than the Action Groups.
3. Custom planet creation.  Why on Kerbin do we need mods for this in KSP2 but this is stock functionality for Juno?  Anyone can, apparently, create a planet in Juno and then share said world with the entire community.  I'm not saying that this is a live-or-die thing here...but really?  We are limited to the Kerbolar system unless we get mods or interstellar (which, apparently, Juno already has)?

And that doesn't even take into account the things that are currently broken in KSP2, like Maneuver Nodes and Trajectories.  And by broken I really do mean that they exist but are either far too fiddly and/or don't work properly.  And no, this is not me shilling for Juno as I have not yet played it.  I'm simply making the comparison between what we were supposed to get from KSP2 and what (apparently) Juno/SR2 already has.  And Juno dropped IN JANUARY, a full month before KSP2.

I guess I'm just confused as to how one game can pull this stuff off and the other can't.  I know there are different code-bases, but really?  Using the largest game in the space (again, no pun intended), and its code, and you can't get this stuff?  Am I missing something?

[GGG-GoodGuyGreg]

1 hour ago, Scarecrow71 said:

I'm going to start this post by saying that I have not yet played Juno.  All of the opinions I am about to express are based solely on the videos I have seen, which includes a few tutorials.  With that said...

 

What I am really confused on here - or, rather, a few things I am confused on, really - is how Juno is capable of doing everything that the KSP2 team says KSP2 should be doing but isn't.  A few examples:

1. Procedural...well, everything.  Multiple tutorial videos show tanks, engines, wings, and other parts as being completely procedural.  Yet, the general consensus here (from what I gather, anyhow) is that procedural things beyond wings would be too complicated or cause too much stress on the game.  How is it possible that one game has been able to accomplish this while at the same time the largest game in the space (no pun intended) can't?  Or, perhaps, won't?  The biggest example I could possibly show people is the tutorial on how to get to orbit (which I won't link to in order to keep the forum as clean of non-KSP videos as possible).  Google it and watch it; the person running the tutorial uses both procedural engines and tanks on the same craft.  And we can't get that here?
2. Custom scripting language for automation.  Apparently, you can use Vizzy to automate just about anything on a ship, which should be the same as the Action Groups here.  And the Action Groups here in KSP2 just don't seem to work very well (if at all).  I'm not saying one is better than the other, especially without having used Vizzy.  But it seems that Vizzy is far more intuitive and scalable than the Action Groups.
3. Custom planet creation.  Why on Kerbin do we need mods for this in KSP2 but this is stock functionality for Juno?  Anyone can, apparently, create a planet in Juno and then share said world with the entire community.  I'm not saying that this is a live-or-die thing here...but really?  We are limited to the Kerbolar system unless we get mods or interstellar (which, apparently, Juno already has)?

And that doesn't even take into account the things that are currently broken in KSP2, like Maneuver Nodes and Trajectories.  And by broken I really do mean that they exist but are either far too fiddly and/or don't work properly.  And no, this is not me shilling for Juno as I have not yet played it.  I'm simply making the comparison between what we were supposed to get from KSP2 and what (apparently) Juno/SR2 already has.  And Juno dropped IN JANUARY, a full month before KSP2.

I guess I'm just confused as to how one game can pull this stuff off and the other can't.  I know there are different code-bases, but really?  Using the largest game in the space (again, no pun intended), and its code, and you can't get this stuff?  Am I missing something?

This is because Juno is built from the ground up with new core systems, to stay on topic.

Whereas KSP2 at this point feels jerry rigged, and there are way too many bugs that were the same as bugs found in early KSP1 versions, which I hope it doesn't mean what I suspect it means.

 

If Juno adds what most people expect from KSP2 science update, then it will give KSP2 a run for its money, and bear in mind Juno is much more polished and stable game compared to how early KSP2 currently is.

Edited May 17, 2023 by GGG-GoodGuyGreg

[Davidian1024]

@Scarecrow71@GGG-GoodGuyGreg

Your posts just convinced me to look into Juno a bit.  After reading the critical Steam reviews from KSP vets, I get the impression that Juno is a dumbed down KSP1 with alot of it's own bugs.  Sounds like they rushed it out in order to beat the KSP2 EA release.

But then I haven't played it either...

[Scarecrow71]

49 minutes ago, Davidian1024 said:

@Scarecrow71@GGG-GoodGuyGreg

Your posts just convinced me to look into Juno a bit.  After reading the critical Steam reviews from KSP vets, I get the impression that Juno is a dumbed down KSP1 with alot of it's own bugs.  Sounds like they rushed it out in order to beat the KSP2 EA release.

But then I haven't played it either...

Again, I haven't played it.  And the amount of information available in the wiki is a bit thin; I'm going off of what I see in trailers and press releases.  It could be a dumbed-down version of KSP1...or it could be that KSP vets are just lashing out at it over their anger with KSP2.  Until/unless I play it, I can only question why it has things that were talked about for KSP2 but aren't there yet.

[Alexoff]

Juno may not be without flaws, but in steam this game has a very good rating. Perhaps this is due to the fact that the developers did not postpone the game for three years and did not promise something that did not appear in the game.

Edited May 18, 2023 by Alexoff

[Davidian1024]

13 minutes ago, Alexoff said:

Juno may not be without flaws, but in ыеуфь this game has a very good rating. Perhaps this is due to the fact that the developers did not postpone the game for three years and did not promise something that did not appear in the game.

I gather it was in early access for 6 years.  And KSP2 hasn't been released.  It's in early access.  Therefore, no promises have been broken... yet.

[Superfluous J]

I do own Juno (I bought it when it was called Simple Rockets 2).

I should install it and see how much better it's gotten in the past ... 5 years or whatever since I played it last.

I'll say 5 years ago the name "simple rockets" was very apt.

[Alexoff]

1 hour ago, Davidian1024 said:

I gather it was in early access for 6 years.  And KSP2 hasn't been released.  It's in early access.  Therefore, no promises have been broken... yet.

Juno has been early access since the beginning. Only a few people were involved in the development. KSP2 was supposed to come out first in 2020, then in 2021, then in 2022, then the full game turned into early access without new features. A days before the release, the developers of KSP2 posted a list of what will not be in early access - heating, aerodynamic effects, atmo scattering and other things. So I don't agree that promises are not broken. Here begins a strange story of half-truths and omissions.

[RocketRockington]

5 hours ago, Scarecrow71 said:

What I am really confused on here - or, rather, a few things I am confused on, really - is how Juno is capable of doing everything that the KSP2 team says KSP2 should be doing but isn't.  A few examples:

For point #1 and #2 - KSP1 is very much capable of doing those things.

  Proc part mods exist and work very well, for being mods.  No doubt if they were part of the game, they would work even better.  It was simply a design choice not to do them for KSP2.  I haven't even seen anywhere from the KSP2 team that the issue is technical complexity, anything you might have read about that being the case is likely uninformed fan speculation.

Custom scripting language for automation - it's like KAL was a start at this that didn't get far enough because that DLC was churned out fast.  The breaking ground mission creator is also a custom visual scripting language, albeit for a different reason.  In this case I think it's a feature KSP2 doesn't want because they're focusing on streamlining/dumbing down the gameplay from KSP1 for new user accessibility, vs adding more complex stuff for power users.

Point #3 - that one is baked into the KSP engine, afaik.  You can't dynamically load a new planet (or any new asset).  And they use prebaked assets for planet textures, height maps, etc.  Clearly it's not an impossible thing either, several games do planet gen now, but KSP2 is not doing it (opinions follow) because of several factors.

1. It's not a core feature they decided on early, because they thought things like multiplayer would sell better.

2. They decided to have a story mode, likely about finding alien artifacts in pre-planned locations and following them to different star systems and then the alien homeworld.  They probably thought proc planets don't work/aren't needed for this.

3. Proc planets are for people who are going to play the game a lot.  KSP2 is more focused on getting your $50 with early glitzy features, and maybe eventually modders will provide extended challenges for the game.

[Skorj]

This is the best KSP2 discussion thread I've seen in a long time.  Glad to see a reasonable discussion.  I think too much backlash has been directed at Nate though.  Not because I want to defend him, but because "there's a lot of red paint in this story" as the saying goes.  I was shocked after an hour's play just how low the standards of the dev team were.  Missing features aside, the kind of bugs we saw were such a bad sign.  As a rather experienced non-game dev, it just made me wonder how anyone can live with that sort of quality in the daily build everyone on the team is using.  I can't blame the junior devs, because you can't know better until you learn better, but where was the engineering leadership, where were the senior engineers? 

It really makes me wonder whether the senior dev guys left during that studio takeover (and it usually is the senior guys who go when the situation gets unpleasant, because they're the ones who can easily find jobs elsewhere).  That would also explain the lack of focus in KSP2.  I've heard that fragments of all of the feature sets we didn't get have been found in the code, the beginnings of career mode and colonies and multi-player and so on.  Working on everything at once and not finishing anything is such a basic leadership failure, I'm really wondering how that happened.