[Poll] So what are we thinking about 1000 part ships?

[jastrone]

the main problem with many parts in a ship is the parts being "floppy". there is no reason for surface atached parts like ladders need as much ḿovement simulation as the fuel tanks. and the rockets could even be completely stiff and the game would be a lot less laggy.  I honestly think the devs made a mistake making the parts not rigidly attached. i know its realistic that rockets are somewhat moving but not to the point in ksp or ksp 2. it is probably not even noticable so adding it to the game doesnt add anything. and floppy tockets is the reason behind kraken attacks. and in iterstellar travel floppyness would add nothing exept  the giant swinging  making sure you cant do your manoever node

[Lyneira]

Some way of simplifying complex builds would definitely be needed if it is desired to get large and complex 1000 part builds running smoothly. Some possibilities:

• Some sort of dynamic part welding that aggregates parts during times of low stress - during collisions, RUDs and other extreme events the frame rate would still get really bad, and the aggregation may be another source of bugs.
• Reduce the need to make high part count vessels by combining common functions  into a single part - but this leads to an explosion of multipurpose parts and will be more difficult to balance. And it takes away from the challenge of building something functional out of individual parts with a specific function. I'd prefer if most parts of a rocket do just one thing and do it well.
• Allow players to do some form of part welding in the VAB for collections of parts that make sense - there should be some kind of tradeoff for this so it doesn't simply become a no-brainer to weld everything to make your ship unbreakable. The possibility of your vessel breaking apart into a whole bunch of pieces is core to KSP.
• Keeping vessels in a docking bay on a mothership welded until they are undocked.
• There could be a role for part welding in making multi-engine assemblies with the engine plates. Treating all of the engines on the plate as one big engine for the purposes of physics and fuel calculation should certainly help.

Edited April 4, 2023 by Lyneira

[mcwaffles2003]

On 3/24/2023 at 10:54 AM, regex said:

Why would you need 1000 parts on a single vessel?

Cause I wanna do more than I have before

Spoiler

 

Edited April 5, 2023 by mcwaffles2003

[Whackjob]

I'm late, but I'm here.  Picked it up yesterday.  Been crazy busy with college, will be getting worse as year goes on, but will try to build during the occasional break.

Give me a few days to see what I can do with things.  To explain, I'm taking 5 classes this semester, on top of a job, got four this summer, and potentially six this fall.  On top of work, the house, and my family. 

I am not a great multitasker, or even a good one.  But, I'll either get better at it, or go mad.

Edited April 7, 2023 by Whackjob

[B15hop]

If they simplified the physics a bit and allowed for "welding" parts together essentially instead of having the game consider them joints, it would cause much less strain on systems.

This could be an option for larger structures such as colonies perhaps? Maybe something you could do after something gets put into a stationary and locked in position. Could even be an option. "Would you like to make this station permanent" (YES/NO) then bam, it executes some kind of weld command. However, that would obviously make the station have many fewer joints and would be much less fun to watch if you "accidentally" ran something into it, as it would only break at a few major locations or even be considered just one object.

There was a mod for that in KSP 1 called something "Weld" you could use to make a massive ship, execute the weld command and it made the FPS absolutely beautiful for it. But like I stated, it then considered the ship one large item.

[Little 908]

well so far my rocket is 700 parts and runs at... 4 fps 

Edited April 8, 2023 by Little 908

[mcwaffles2003]

8 hours ago, B15hop said:

If they simplified the physics a bit and allowed for "welding" parts together essentially instead of having the game consider them joints, it would cause much less strain on systems.

This could be an option for larger structures such as colonies perhaps? Maybe something you could do after something gets put into a stationary and locked in position. Could even be an option. "Would you like to make this station permanent" (YES/NO) then bam, it executes some kind of weld command. However, that would obviously make the station have many fewer joints and would be much less fun to watch if you "accidentally" ran something into it, as it would only break at a few major locations or even be considered just one object.

There was a mod for that in KSP 1 called something "Weld" you could use to make a massive ship, execute the weld command and it made the FPS absolutely beautiful for it. But like I stated, it then considered the ship one large item.

I wonder if it could be possible to do something like dynamically weld and unweld joints based on stresses or static/dynamic loads. Maybe the mechanism telling joints to weld/unweld would take up more processing... but maybe not? It works on terrain and terrain at the end of the day is a map of vertices like and RBD is a tree structure of vertices...

[Whackjob]

I remember partwelder.  Never used it.  It was all about the engineering challenge, for me.  It has to fly, it has to land, and it's gotta have the right end pointed the right way bare minimal fifty-one percent of the time.

Engineering challenges such as this:

Now, I did some stress tests with the lander leg (stump?) aspect.  And it held up.  But that was with the weight of one fuel can and six of them motors.  When I put this on the pad to try to give it its maiden flight (even though I've barely started the building UP portion) the weight of it IMMEDIATELY makes it liquefy straight into the pad.  That's engineering challenge number one.

It could take me months.

 

And if a dev is eggin' it through here, can I ask that we get an option to bring up the flight report on a crash WHEN WE WANT TO?  It's an absolute irritation for me to try to launch this ship, over and over again, and I can't see what's initially breaking because somebody's cheerfully shoving a clipboard in my face to try to get me to feel better about everyone dying.  That's what we're here for!  I bought the ticket, let me sit down and enjoy my ride.

Edited April 8, 2023 by Whackjob

[J.Random]

On 3/24/2023 at 6:03 PM, The Aziz said:

Note that these were shown as development footage, it wasn't just cinematic trailer. And the framerate wasn't slideshowy

It's called a bullshot. I think I recall somebody mentioning that the scenes were made in the editor and prerendered for the trailer, so no fizzicks, no real-time lighting, just a bunch of parts set in a way to suggest that the game could handle it. In other words, a lie and a fake. You can actually notice that the lighting is quite a bit off in the second shot (look at the shadow "terminators" on the round tanks, leftmost tanks are about "half-moon", while the rightmost ones are so obviously in a "waning moon" state): it seems that whoever made it used point/omnidirectional light and put it way too close to the contraption.

[Guest]

On 3/25/2023 at 3:10 AM, RocketRockington said:

Intercept credit for basically overhyping their game for since 2019 - 3.5 years

This is fun because I distinctly remember years of people wining because of no communication at all coming from the studio.

 

BTW 3 pages and nobody managed to bring up a single example of a dev actually promising 1000+ parts ships.

I'm not saying it doesn't exist, it probably does, I don't know, but sure enough looks like nobody is even doing the most basic research before claiming that the devs promised x or y.

The reply to Shadowzone about part counts is a masterpiece in diplomatic responses, not only he never specifies an actual part count, but  it doesn't even come close to confirming 1000 parts ships or say that they will be possible.

 

The first step to a 1 km long ship isn't a system capable of managing 1000 1m parts, but huge parts that allow you to build such a ship in 1-200 parts at most, probably the ability to "containerize" landers into kits/hangars/containers and orbital colonies making you skip the part in which you need 10 times more parts to actually lift that ship to orbit from the KSC.

[Observe]

I see higher parts counts being mostly relevant to space stations and surface colonies. Let's say I have a 500 part space station and I want to dock a 200 part spaceship. That's already 700 parts and could easily reach 1000 if I want to dock multiple ships. So I don't think 1000 parts is at all unreasonable. This is one of the weaknesses in KSP1 and hopefully KSP2 will improve on. However, I don't see it happening, unless the game adopts a different approach to physics calculations and their accompanying negative impact on performance.

[Alexoff]

1 hour ago, Master39 said:

BTW 3 pages and nobody managed to bring up a single example of a dev actually promising 1000+ parts ships

In fact, it is difficult to find something specific that would be promised to us. Will there be many planets in new systems? We were called quite a few of them, you can count on the fingers. If the new star has only 2-3 planets, then it will be rather boring, and we will have nothing to argue with. We weren't promised anything! We were not promised high performance, we were not promised huge complex ground bases, we were not promised any interactive objects on the planets. We know almost nothing specific about science and multiplayer. And we are not given any dates. But this does not mean that there are no reasonable limits from below.

[RocketRockington]

4 hours ago, Master39 said:

BTW 3 pages and nobody managed to bring up a single example of a dev actually promising 1000+ parts ships.

Did you miss this post?  Or all the videos where they show off extremely high part count ships?  Or is this going to be a pointless semantics argument where you try to argue those weren't 'promises'.  Which I'll acknowledge - they never specifically promise anything, they just hyped things without biding themselves to anything legally.

[Guest]

6 hours ago, RocketRockington said:

Did you miss this post?  Or all the videos where they show off extremely high part count ships?  Or is this going to be a pointless semantics argument where you try to argue those weren't 'promises'.  Which I'll acknowledge - they never specifically promise anything, they just hyped things without biding themselves to anything legally.

 

Please read my whole message. You'll discover I've already addressed both your points.

 

9 hours ago, Alexoff said:

10 hours ago, Master39 said:

 

In fact, it is difficult to find something specific that would be promised to us. Will there be many planets in new systems? We were called quite a few of them, you can count on the fingers. If the new star has only 2-3 planets, then it will be rather boring, and we will have nothing to argue with. We weren't promised anything! We were not promised high performance, we were not promised huge complex ground bases, we were not promised any interactive objects on the planets. We know almost nothing specific about science and multiplayer. And we are not given any dates. But this does not mean that there are no reasonable limits from below.

Point is, if you listen to the reply Nate gave to Shadowzone he treads around the thing without committing to anything specific. He outright states that you'll be able to make the game slow down with high part counts, only saying that the number of parts needed for that is going to be higher.

 

I see this thrown around a lot, but actually look at that ship, a crew cabin, something behind it, 4 rings, 2 big habitats, maybe something else in the middle, let's say 10 parts and we've already finished the crew section.
Let's say the long structure is not modular and made up of 5 segments, let's make it 10, we're at 20 parts and 2/3 down the ship, 6 containers in a 6x symmetry (I don't see the ones below, but, let's pretend), another 36 parts, we're at 56, then we have 3 tanks in another 6x simmetry, 74 parts so far, 82 with the bigger tanks, 83 with the single-part huge engine.

Now, on the station, I see 160 parts in panels alone, and with the truss structures, if they're as small as they appear, we can get up to 250/270 parts easily. To that we add 24 huge spherical tanks, 24 containers, 8 big habitats, 9 rings, 2 smaller inflatable modules, 1 orbital VAB (the whole assembly the big ship is attached to) and, why not, an additional bonus 100 parts for whatever that thingie at the end of the rings is plus antennas, RCS thrusters and all the small bits not present in this image, another 168 parts to the count.

We have 524 parts.

And we can cut half of those out if we have procedural trusses, solar panels, or if we scrap the ISS look and do the right thing by bringing a nuclear reactor to our fusion fueled colonization ship shipyard which is sitting in jool orbit.

 

It may not be a huge difference, but we're sitting at half of what people is saying was promised with that image, and half of that are aesthetics (again, solar panels on a fusion-power capable station in Jool orbit)

 

The most important thing I take away from counting the parts is that barely a third of the parts are actually functional, the rest is all piling on of smaller parts to get to the desired size and a ridiculous amount of solar panels in a 8x symmetry (seriously, it looks like trying to use the OX-STAT to make the ISS arrays). The ability of welding truss segments together, or them being procedural is going to dramatically reduce the part count of everything, the same goes for solar arrays. 

 

Not saying that the game shouldn't be able to deal with 1000 parts ships, only that optimization goes both ways, a lot of effort goes into reducing the amount of parts needed. The orbital VAB alone is going to do a lot for that, given that most user examples of 1000+ ships posted here are on the launchpad and including the rocket needed to take those contraptions to space (BTW I get the challenge of launching an already assembled ISS from the ground, but docking ports are a thing, no need to launch a single 1000 part rocket when you could launch 3 400 parts ones).

[Alexoff]

3 hours ago, Master39 said:

We have 524 parts.

This frame was just to demonstrate the capabilities of the game. It's kind of like the gameplay was in 2019. There is no functionality in this, the orbital stations both in the game and in real life are generally meaningless (for some reason, nothing about the ISS is ever reported in astronomical areas, but maybe I'm just inattentive). Simple - look KSP1 fans, in KSP2 you can build monstrous space stations with good performance!

3 hours ago, Master39 said:

Not saying that the game shouldn't be able to deal with 1000 parts ships, only that optimization goes both ways, a lot of effort goes into reducing the amount of parts needed.

Well, of course it is, Stratzenblitz will surely be able to make craft from 10 parts to fly to all the planets of new star systems. But maybe someone wants to make a space battleship Yamato, and for this he will have to return to KSP1? Many forum users in the voting wrote that there should be the possibility of building megaships. Nevertheless, this is a continuation of KSP1 and we would like to see more features, and not stories that we don’t need it, you can fly around everything on a simple ship, put a flag on each planet and delete the game.

[Guest]

2 hours ago, Alexoff said:

This frame was just to demonstrate the capabilities of the game. It's kind of like the gameplay was in 2019. There is no functionality in this, the orbital stations both in the game and in real life are generally meaningless (for some reason, nothing about the ISS is ever reported in astronomical areas, but maybe I'm just inattentive). Simple - look KSP1 fans, in KSP2 you can build monstrous space stations with good performance!

I don't get your point.

The huge station in the image doesn't even come close to the 1000 parts target.

It's a quarter of that if they weld the truss segments together and provide big procedural solar panels (or if the player building a interstellar shipyard doesn't choose to power it with solar while in Jool's orbit).

 

2 hours ago, Alexoff said:

Well, of course it is, Stratzenblitz will surely be able to make craft from 10 parts to fly to all the planets of new star systems. But maybe someone wants to make a space battleship Yamato

You seem to think I'm claiming that players should limit their part counts.

I'm not.

I'm saying that a big portion of optimization goes through providing parts appropriately sized for the ships you're intended to build.

There's no art or creativity lost if they give us a 50m truss segment instead of forcing us to line up 10 5m ones, or a 500m one for what matters.

The game supporting bigger ships and station doesn't mean that it needs to be able to withstand 10000 parts.

[regex]

6 hours ago, Master39 said:

Now, on the station, I see 160 parts in panels alone, and with the truss structures, if they're as small as they appear, we can get up to 250/270 parts easily. To that we add 24 huge spherical tanks, 24 containers, 8 big habitats, 9 rings, 2 smaller inflatable modules, 1 orbital VAB (the whole assembly the big ship is attached to) and, why not, an additional bonus 100 parts for whatever that thingie at the end of the rings is plus antennas, RCS thrusters and all the small bits not present in this image, another 168 parts to the count.

We have 524 parts.

And we can cut half of those out if we have procedural trusses, solar panels, or if we scrap the ISS look and do the right thing by bringing a nuclear reactor to our fusion fueled colonization ship shipyard which is sitting in jool orbit.

We can cut it all out if we treat it like a single colony structure, which appears to be what they'll be doing according to Nate. That orbital station is now one big part (essentially). So in-frame we've got 100, maybe 200 parts at most, in orbit around Jool, with only Jool visible. Of course it's going to run really smooth.

[Guest]

1 hour ago, regex said:

We can cut it all out if we treat it like a single colony structure, which appears to be what they'll be doing according to Nate. That orbital station is now one big part (essentially). So in-frame we've got 100, maybe 200 parts at most, in orbit around Jool, with only Jool visible. Of course it's going to run really smooth.

That may hamper the ability to use colony parts in normal crafts, but yes, I imagine having a difference of some sort between crafts and buildings would simplify things a lot.

And there's also the matter of what Nate said recently about wobbly rockets, ships and stations that size aren't supposed to wobble at all, which is a great source of optimization if you just weld them together.

[regex]

3 minutes ago, Master39 said:

That may hamper the ability to use colony parts in normal crafts

I don't see an issue there if that's ever allowed. Colony parts are for colonies and craft parts are for crafts. If you're going to reuse an asset for either you might as well define it specifically for the purpose its used for (in this case once for each) instead of trying to make it one thing with dual uasge. Much simpler implementation.

[darthgently]

20 minutes ago, regex said:

I don't see an issue there if that's ever allowed. Colony parts are for colonies and craft parts are for crafts. If you're going to reuse an asset for either you might as well define it specifically for the purpose its used for (in this case once for each) instead of trying to make it one thing with dual uasge. Much simpler implementation.

Are orbital colonies going to be a thing?  O'Neill cylinders?  Hope so

[regex]

31 minutes ago, darthgently said:

Are orbital colonies going to be a thing?  O'Neill cylinders?  Hope so

No idea about O'Neill cylinders but orbital colonies are most definitely going to be A Thing because you will need to do orbital assembly for the bigger ships. That implies storage of resources, a workforce, and a VAB with "launchpad".

[Alexoff]

3 hours ago, Master39 said:

The huge station in the image doesn't even come close to the 1000 parts target.

Yes, I got it. But should consumers be counting the details in this station and the docked interstellar spacecraft fueled with solid hydrogen? For example, no one told us that this size of the station would be the limit for the game, or rather, such a number of parts. To tell after the release of the game that it was the fans themselves who fantasized about high performance is a dubious thought.

4 hours ago, Master39 said:

There's no art or creativity lost if they give us a 50m truss segment instead of forcing us to line up 10 5m ones, or a 500m one for what matters.

Nate didn't promise us treatment tanks. But unstable connections, he even promised. I wonder if a kraken will eat such a station in three seconds in KSP2?

4 hours ago, Master39 said:

You seem to think I'm claiming that players should limit their part counts.

It's more like users should lower their expectations to the level of the developers' ability, and not to the level that fans imagined from 2019 trailers. Which apparently the enemies of T2 uploaded to YouTube.

[RocketRockington]

[snip]

Anyway, the solution of just making all the parts bigger isn't a good one.  I don't buy a Legos to I have 3 pre built pieces to stick together in a pre-defined fashion.  More parts offer the possibility of more creativity, less parts means less creativity, its as simple as that. 

And I didn't want to have to read between the lines of every dev answer that pointed toward better physics with bigger craft as if I was listening to a politician speak and give "diplomatic" answers.  Regardless, 80% of the people answering the poll expected higher part counts with reasonable performance - whether it was because of developer-created hype or personal expectations.

Edited April 10, 2023 by Snark
Redacted by moderator

[Strawberry]

8 minutes ago, Alexoff said:

Nate didn't promise us treatment tanks. But unstable connections, he even promised. I wonder if a kraken will eat such a station in three seconds in KSP2?

While wobbliness from acceleration is still very much a thing and can easily destroy parts, kraken spontaneously crumpling up non accelerating large spacecrafts seems to be gone in my experience (though the wobbliness from acceleration can cause parts to crash into each other). In terms of physics, struts feel weaker but besides that it feels more stable overall. I cant speak much to the performance of large parts because GPU is by far the main bottleneck for me and thing killing my frame rates.

[Guest]

15 hours ago, Alexoff said:

Yes, I got it. But should consumers be counting the details in this station and the docked interstellar spacecraft fueled with solid hydrogen? For example, no one told us that this size of the station would be the limit for the game, or rather, such a number of parts. To tell after the release of the game that it was the fans themselves who fantasized about high performance is a dubious thought.

I don't think if the game will ever be able to sustain a 1000 parts launch from Kerbin, still, due to a number of choices of game design, I'm more than convinced that such stations and ships as the one shown in that footage are more than possible.

Again, the station is half of the 1000 parts target, the Devs clearly claimed that bigger rockets should not be wobbly, meaning less things to simulate at that size, bigger parts will reduce the need for spam of smaller ones too, further reducing part counts, and, finally, orbital VABs and delivery routes means you can build stuff directly in orbit, removing hundreds of parts of launch veichles from the equation. 

All of that points towards a game in which ships and colonies as big as the one shown will be possible well within KSP-1 parts count limits, even disregarding potential future optimization of KSP-2 code.

[snip]

 

15 hours ago, RocketRockington said:

Anyway, the solution of just making all the parts bigger isn't a good one.  I don't buy a Legos to I have 3 pre built pieces to stick together in a pre-defined fashion.  More parts offer the possibility of more creativity, less parts means less creativity, its as simple as that. 

That only works in your over-generalizarion.

As if most 1000+ ships weren't actually almost entirely made out of 3 or 4 different MK3 fuselages (admittedly we evolved a bit part that, now we can include fairings and flags too, so what's it, 6 parts?), have you even seen the Shadowzone bit you pointed me to in your last reply?

Wasn't everyone happy about procedural wings because they drastically reduce the parts count of planes while also increasing the creativity?

And it's not a new thing I'm making up on the spot just to adere to some narrative, I've been talking about "mosaics of MK3 parts" at least since 2020, and repeatedly after that.

 

15 hours ago, RocketRockington said:

And I didn't want to have to read between the lines of every dev answer that pointed toward better physics with bigger craft as if I was listening to a politician speak and give "diplomatic" answers.  Regardless, 80% of the people answering the poll expected higher part counts with reasonable performance - whether it was because of developer-created hype or personal expectations.

That's quite far away from they've promised us 1000+ parts ships" and, again, I'm not even claiming they've not said it

[snip]

Edited April 10, 2023 by Snark
Redacted by moderator