A Taste of Science

[PDCWolf]

50 minutes ago, cocoscacao said:

That was sarcastic response to those who claim it was made by 1 person, and KSP 2 is done by the whole team. I'm glad that part is finally cleared up now.

Not good enough attempts, apparently. "Kerbal" has nothing to do with it. It can only attract more people initially, but if the game is good "Naughty Spacenouts" can even surpass it.

Seems to me you're clearly confusing your willingness to try alternatives with general market trends. Alternatives have been made, they were good (specially when compared against EA ksp1), they failed.

[RocketRockington]

This is a pretty classic case of  First Mover Advantage

 

 

Edited May 15, 2023 by RocketRockington

[cocoscacao]

20 minutes ago, PDCWolf said:

Seems to me you're clearly confusing your willingness to try alternatives with general market trends. Alternatives have been made, they were good (specially when compared against EA ksp1), they failed.

Yes, but not good enough... Microsoft flight simulator -> X plane.

14 minutes ago, RocketRockington said:

This is a pretty classic case of  First Mover Advantage

It will attract more people initially, but if the game is good (people like it), they will play it.

[Alexoff]

36 minutes ago, VlonaldKerman said:

Is this really true? I don't see why they couldn't just implement the Scatterer mod-level stuff, as a starting point.

Of course, I can't know what's going on behind the doors of the IG office, but they showed atmospheric scattering back in the pre-alpha video. Since then, no scattering has been shown or told to us anywhere. As far as I understand, at that time there was no single build of the game, each group of developers had their own small build for testing, no one checked how these builds would work with each other. They began to bring them into one only this winter. Most likely it turned out that the scatterer they developed drops FPS to very low values and it would be necessary to specify not 3080, but 4090 in the recommended system requirements.

43 minutes ago, VlonaldKerman said:

Scatterer proves that it can be done to a reasonable level of effectiveness and performance

To do this, they need to have the appropriate competencies. Engine exhausts and almost all parts were made by a modder from our forum. It's scary to think what would have happened without him. The creator of the scatterer could not or did not want to be lured into the number of developers, so there is no scattering in the game. Helplessness and hopelessness ¯\_(ツ)_/¯

[The Aziz]

"missed the only shot"

It's funny really.

Like some people forgot how games tend to gather more audience as features get added, even before 1.0. Sometimes after.

[VlonaldKerman]

1 minute ago, Alexoff said:

Most likely it turned out that the scatterer they developed drops FPS to very low values and it would be necessary to specify not 3080, but 4090 in the recommended system requirements.

Interesting. I just made a post about this kind of issue- I feel like we've seen/heard about a lot of systems that aren't yet well integrated with each other, which is concerning considering that this under the hood stuff is the most important part of the game. Also directly contradicts some Nate Simpson pre-launch statements that I think were in the vein of, "The core game is mostly finished, so we're releasing into EA to refine game features and improve polish" or something to that effect.

 

[Alexoff]

3 minutes ago, VlonaldKerman said:

I feel like we've seen/heard about a lot of systems that aren't yet well integrated with each other, which is concerning considering that this under the hood stuff is the most important part of the game.

This is the taste of KSP2...

[PDCWolf]

6 hours ago, The Aziz said:

"missed the only shot"

It's funny really.

Like some people forgot how games tend to gather more audience as features get added, even before 1.0. Sometimes after.

A lot of the "sense of finality" comes from KSP2 still being stuck in Unity, and taking a lot of KSP1 with it, the most egregious example being PQS+ bursting at the seams and taking the performance with it. For a lot of people, me included, the entire point of KSP2 was a much better codebase and foundation to actually build stuff upon. That ship sailed now the game is out, and it has the wrong stuff onboard.

Edited May 15, 2023 by PDCWolf

[cocoscacao]

1 hour ago, PDCWolf said:

A lot of the "sense of finality" comes from KSP2 still being stuck in Unity

Someone wiser than me will have to answer how much would such a move affect the modding scene.

[PDCWolf]

5 minutes ago, cocoscacao said:

Someone wiser than me will have to answer how much would such a move affect the modding scene.

For starters, the asset streaming required for mods won't be much different from KSP1, it just can't be much different. You can look at all other moddable Unity games (Cities: Skylines comes to mind) and whilst the community has achieved amazing things, the problems are still the same: loading everything on ram at once. KSP2 might provide a better, faster, or more open API (or all 3 at the same time) but that's still not magic.

 

 

[cocoscacao]

@PDCWolf I was thinking more in terms how many people would be modding. Working with C#/Unity is easier than C++/Custom engine

[Strawberry]

7 hours ago, Alexoff said:

As I understand it, the developers could not integrate atmospheric scattering into the game and instead added distant fog. Which I saw in games 20 years ago.

Do you remember that?

We do have scatter though? It's super subtle on kerbin but its pretty obvious on Eve. I think part of this confusion is that most of that video seems to be on gurdamma and from what we know gurdamma likely has a very thick atmosphere. 

[Alexoff]

32 minutes ago, Strawberry said:

We do have scatter though? It's super subtle on kerbin but its pretty obvious on Eve.

It's a distant fog, scattering is a little harder to do

[PDCWolf]

4 hours ago, cocoscacao said:

@PDCWolf I was thinking more in terms how many people would be modding. Working with C#/Unity is easier than C++/Custom engine

That's not language related, that's API related. A game can be a headache to mod in any language if the devs don't provide a good API. Minecraft is Java and for the longest part (not sure, haven't modded in a long while) making native mods meant decompiling then wading through completely obfuscated code, yet Java is one of the most popular languages on Earth.

[Strawberry]

44 minutes ago, Alexoff said:

It's a distant fog, scattering is a little harder to do

What's the difference? 

[Alexoff]

3 minutes ago, Strawberry said:

What's the difference? 

Here is scattering

Here is distant fog

Fog is part of the scattering, but there are other parts as well.

[Socraticat]

1 hour ago, Strawberry said:

 

1 hour ago, Alexoff said:

It's a distant fog, scattering is a little harder to do

In this photo I see a white flag that is lit with purple light. Am I seeing this wrong? Am I misunderstanding what scatter is? Does scatter not "scatter" light particles? Like green Joolshine and Purple Evosphere effects?

[Strawberry]

18 minutes ago, Alexoff said:

Here is scattering

Here is distant fog

Fog is part of the scattering, but there are other parts as well.

Oh wow that is a very foggy first picture, that aside, I mean you can argue that the scatter isnt as good as it should be, but like, the game does do things like scatter the light from sunsets and stuff like that. 

[EvelynThe Dragon]

On 5/12/2023 at 1:22 PM, Scarecrow71 said:

I do like that it is all-in-one

I don't think that we know that, and I actually hope it isn't. 

[Alexoff]

42 minutes ago, Socraticat said:

In this photo I see a white flag that is lit with purple light. Am I seeing this wrong? Am I misunderstanding what scatter is? Does scatter not "scatter" light particles? Like green Joolshine and Purple Evosphere effects?

In KSP1, the illumination on the planets was set by simply adjusting several colors for the filters. You could do pink shadows and an orange sky. Checking scatter is pretty easy - will the flag turn a different color at sunset? If not, then in KSP2, several light filters have been added to the planets for different situations.

33 minutes ago, Strawberry said:

I mean you can argue that the scatter isnt as good as it should be, but like, the game does do things like scatter the light from sunsets and stuff like that.

Many players do not see retracing, so developers who keep up with the times can also not add it to games, limiting themselves to the usual reflections of the 2008 level. But in this case, there is no need to talk about RTX

[RocketRockington]

8 hours ago, cocoscacao said:

Yes, but not good enough... Microsoft flight simulator -> X plane.

It will attract more people initially, but if the game is good (people like it), they will play it.

That's not true.  The games that tried to follow in KSPs footsteps reached quality levels that were significantly better than early KSP without attracting anything like the same amount of interest from players or media sources that would help advertise them for free.

What you're basically suggesting is that if those games were as good as KSP, right now, they could attract an audience - but KSP got a lot of funding + attention + mod support + player feedback along the way to get where they got, and other games aren't getting that.  And no external party, like a big publisher, is stepping in to provide the funding.  

All we can hope for now  is for a Cities Skylines sort of thing to happen - after EA screwed up SimCity and stopped developing it, another developer saw value in building a new version.  But that cycle took a long time to happen.

[cocoscacao]

47 minutes ago, RocketRockington said:

What you're basically suggesting is that if those games were as good as KSP, right now, they could attract an audience - but KSP got a lot of funding + attention + mod support + player feedback along the way to get where they got, and other games aren't getting that.  And no external party, like a big publisher, is stepping in to provide the funding. 

It has that special something. Similar gameplay isn't enough. What makes people tick? If I knew that, I'd be rich.

[Xeiki]

13 hours ago, PDCWolf said:

A lot of the "sense of finality" comes from KSP2 still being stuck in Unity, and taking a lot of KSP1 with it, the most egregious example being PQS+ bursting at the seams and taking the performance with it. For a lot of people, me included, the entire point of KSP2 was a much better codebase and foundation to actually build stuff upon. That ship sailed now the game is out, and it has the wrong stuff onboard.

This right here is the main thing. Seeing bugs and issues KSP1 had in the early stages cropping up again is just a huge red flag. That plus the performance issues. Like, yes, performance optimization comes last in development, we all know that - but if basically nobody is able to run your game with more or less all features missing after such a long development time ... ooof, just oof. It all speaks to very poor planning. Like assets not being done (missing cockpits, etc.). And fine, you can say "They spent that time on more important things like coding", but then, there's almost no features in the game. Not even reentry heating. So clearly, that stuff's not done either. So then one could say "Ah, but they spent all that time building a better base to build upon", but, uh ... see point one ... And one does ask themselves what all this development time was actually spent on, and what it'll be spent on in the future, and if the fundamental issues can be overcome.

[RocketRockington]

1 minute ago, Xeiki said:

Like, yes, performance optimization comes last in development, we all know that

This actually isn't the truism it used to be.  It depends a lot on what kind of optimization you need.

For instance, if you're making a shooter that needs to run at 60fps on a console, you don't want till the last 3 months to try to make your game run at 60fps on a PS5 when it's running at 10 on a 4090.   You make it run at 60fps from the start and you keep it running at 60fps.  

What its true of is you don't do fine-grained, code-obfuscating optimizations until the end.  But if you need to cut your frame time by 50% to reach your target framerate - you better not be counting on fine grained optimizations to get you there, especially in a game where there's a wide variety of time sinks rather than just one or two trouble spots.

With a game like KSP - their ideal strategy would have been to focus their prototyping time on getting a better, cleaner version of KSP1's key systems working to build from.  I don't think any such approach was taken, based off of the screenshots/bullshots from the early days of the project, both during the Star Theory days & Intercepts time, it seems like they focused on building stuff they could demo/hype instead.

[Socraticat]

5 hours ago, cocoscacao said:

It has that special something. 

The post that inspired a short.

We get a little something here... a little something there... next thing you know, we'll be at 1.0! Kerbal Enlightenment!