Hopes and Wishes for KSP 2

[garwel]

One relatively minor thing that can make life easier for modders and mod users as well as for realism junkies (like myself): use of standardized real-life units in the game. In particular, switch to liters (or cubic meters?) for measuring volume of liquids and, as far as possible, gases and solids, and to joules or kilowatt hours for Electric Charge. It would help balancing parts mods.

[Xd the great]

5 hours ago, garwel said:

In particular, switch to liters (or cubic meters?) for measuring volume of liquids and, as far as possible, gases and solids, and to joules or kilowatt hours for Electric Charge. It would help balancing parts mods.

 

Spoiler

Americans: WHY NOT MILES?

 

Edited November 20, 2019 by Xd the great

[garwel]

9 hours ago, Xd the great said:

 

  Hide contents

Americans: WHY NOT MILES?

 

Yes, and gallons.

[jimmymcgoochie]

Rest of the world: because miles and gallons have no readily divisible units, and because pretty much everywhere else in the world uses the metric system which uses multiples of 10 and makes sense instead of 14 ounces in a pond, 16 pounds in a stone, 12 inches in a foot, 1760 yards in a mile  Do you want your VAB floor space measured in acres or hectares?
 

The international standards for measurement are SI units, and pretty much all of them are metric with orders of magnitude being easy to do- kilonewton, micrometre, megaton, gigawatts etc. so trying to shoehorn gallons or miles into it will only cause confusion, conversion calculation headaches and doesn't add anything to the game for the majority of people who either use metric natively, or who don't use it but don't care that the speedo is in m/s instead of miles per hour and your orbit is at 90,000m instead of 295275 feet 7 inches...

Now on to the part that I actually looked at this thread for- what I want to see in KSP 2:

- Built-in docking alignment indicator plus docking ports that can lock to specific angles (multiples of 15 degrees would probably be most useful e.g. 90 degrees) to build a space station that all points in the same direction. That alignment indicator mod is incredibly useful and yet so incredibly simple, it amazes me that it isn't baked into the game in the first place.

- A deeper, more meaningful tech tree that requires more than just grabbing 'science' and buying new stuff with it in the R&D Department- put specific goals in place like flying a crewed mission to the Mun and back, launch a craft into interstellar space, create a space station around Kerbin etc. and create a more conventional path to space that puts uncrewed probes and satellites before crewed missions rather than flying crewed rockets literally from day 1. I'd also like to see more things to do with planes and more of an emphasis on spaceplanes as satellite launch vehicles, plus micro/cube sats: think of the little school-Kerbals' faces when they see the little cube sat they built in engineering class floating above Kerbin (and the boost to your reputation for giving it a piggy-back ride during your latest 100 ton orbital launch); either that, or scatter a thousand of them in the upper atmosphere and then sit back and watch the show!

- More commercial options to run your space project commercially- get contracts to launch pre-built payloads into specific options, develop commercial rockets and run a business to get more money and greater prestige which in turn could boost the exploration side of things (and maybe some tech tree nodes could require a specific number of successful contracts/launches to complete or something like launch/return/relaunch the same ship a la Space-X). Do deals with the various suppliers to get their parts at a discount by limiting your use of a rival's products- this would require some partially redundant parts so there was enough overlap to be viable.

- Other space agencies operating at the same time. Kerbin needs a space race, and there's nothing like a bit of competition to drive innovation and push the boundaries! Other agencies could compete with you for commercial contracts, race you to achieve specific milestones first or even team up to launch joint or complementary missions like the ISS or US-European missions. Baikerbanur, Woomerang and Dessert Launch Site could become the HQ for rival space programmes, new sites could be created in places where there isn't much happening on Kerbin and you could have the embarrassing situation of landing your cutting-edge reusable booster on a rival's base instead of yours and they'd steal the designs! This would also work as an introduction to multiplayer games, could easily be disabled if you don't want them and combined with the more commercial side of things above would give you something to fill the time as your rover flies for 300 days to Duna or you wait for Jool to be in the right place to do a transfer.

- More science! I don't just mean more things to staple on the side of a ship- more deployed science modules (breaking ground has a total of 4 different experiments, and you can only use 3 at a time because one needs an atmosphere and another needs no atmosphere), long-term activities for crewed missions, anomalous sensor readings to investigate and so on plus some real life experiments such as the Hayabusa-2 comet sampler and hypothetical future missions such as a submarine for exo-oceans, whether on the surface or buried beneath it.

- Backstory and plot! There's no information on how KSP came into being and virtually nothing about Kerbal society in general except the little snippets you find in missions and the agency summaries within them. There's also plenty of stuff to talk about- where did those monoliths/that Arctic UFO/the pyramids and stone face on Duna come from? What made that huge crater on Kerbin? Why does Eve have oceans when it's over 400 Kelvin on its surface, and so many craters when it has such a thick atmosphere? Having some kind of story running through the game would also produce more events that could be required for research nodes on the tech tree, a better grounding for rival space agencies and suppliers plus a greater role for the characters in the game, whether that's the astronauts themselves or the rest of the team e.g. Gene or Wehrner.

- An expanded set of stock parts. Shiny new super-tech aside, there are still plenty of gaps in the parts catalogue of KSP that various mods have tried to fill out, and which KSP 2 would benefit from filling as standard: the 0.625m parts category is almost empty with just one stackable engine and one fuel tank (and a handful of little plane parts but I'm mostly talking rockets here), even with the addition of the new Mite and Shrimp boosters, and very few other parts that actually fit into that footprint (I think the QBE is the only probe core that a 0.625m heat shield actually covers completely); there are a lot of missing parts in the 1.875m size bracket too including a probe core(!), battery, reaction wheel, parachute and a short rocket engine like the Terrier that can be used for landers. There's one nuclear engine, one ion thruster and only one set of RCS thrusters which are too big for small satellites and too small to control big spacecraft/stations without using lots of them and bumping the part count really high. There's a sizeable opening for a set of space station parts- habitations, experiment modules/labs, life support modules and size adapters and/or a larger hub module. I'd also like to see Mk2 plane holds actually able to carry a 1.25m payload with some parts stuck on the outside without those parts sticking through the aircraft's skin and getting stuck when the doors open, and likewise a Mk3 cargo hold that you can a) load stuff in the ramp more easily and b) actually launch 2.5m parts without the same problems the Mk2 stuff has, plus a Mk1 cargo hold for small payloads.

- Performance and visuals. After putting together a large mission containing five probes bristling with science parts all attached to a single big craft that was purely for propulsion (admittedly with a few parts from mods) I started what was supposed to be a 1 hour burn, had breakfast, did some household chores, went shopping, had lunch and came back nearly 5 hours later and it still hadn't finished, but was now noticeably off course and would miss the target planet completely. First of all, the game was running super-slowly with 1 second of game time lasting 4-5 seconds of real time even though my PC can cope with pretty much every game I've thrown at it on maximum graphics settings. I've seen the same thing on space stations and large spaceships so either some mods have really poor optimisation, or else the game itself isn't using the available hardware efficiently. The second problem was that the result of the maneuver didn't match what it said it would do, possibly due to using an efficient but low thrust engine. Since KSP 2 promises interstellar travel, this issue could become exponentially worse if using more conventional parts (like the Dawn ion thruster) to go interstellar. Making the maneuver editor take your craft's size and thrust into consideration in more ways than just the time taken to do a specific maneuver would hopefully improve accuracy for long-distance travel, and adding the ability to do time warps while maneuvering, or switch vessels and continue the burn, would be a huge time saver.

I've also tried a variety of visual enhancement mods as the stock graphics look underwhelming especially in orbit/map view, will noticeably sharpen up as I fly over or switch back from map view and in some cases will create strange squares where shadows/sea/surface of Jool won't render correctly/at all, and the new graphics in v1.8 made the game crash every other time I switched to a vessel on the Mun or Minmus but looked really nice when it didn't. 
Without the need to maintain any backwards compatibility to existing players with older or less capable devices and with the option of pushing the required specs up to take advantage of modern hardware, KSP 2 could look so much better straight out of the box with clouds, weather, high-definition terrain and features and much more. If the trailer and demo videos were anything to go by, things look promising in this area!
A related point is the appearance of parts in the game, specifically the variants of those parts- or lack of them! I want to see a fully custom paint system for most parts or at the very least an expanded set of stock paints based on real rockets that aren't from NASA in the 1950s/60s- there are plenty to choose from! The stock parts often don't match each other properly (1.875m parts are really bad for this, they look completely different despite using the same hex codes for their colours) and many parts (3.75m tanks, oscar-b and the little orange fuel tanks etc) are lacking in variation. A fully custom paint scheme would make your spacecraft distinctive, would tie in to the commercial stuff above as a kind of branding (which would also make ship-sharing on the forums/steam better to look at) and to me would be a valuable addition to the game. Some of the gameplay demo videos and the images on the KSP2 steam page have brightly coloured fuel tanks, but I'm not sure if that was just to show what type of fuel was in them e.g. some kind of nuclear fuel, maybe?

And finally- it's better to release later with all the features fully working and tested as much as they feasibly can before releasing additional features at a later date than to release it early or rush everything into version 1.0 and end up with a really buggy and glitchy game that lets everyone down, followed by a series of hasty patches or removing certain features that didn't work. I'll wait a few more months for a more polished product (but if you're looking for a beta tester, I have some experience doing that and I actually do software testing as a job )

[Motokid600]

A proper day/night skybox transition is high on my list. Right now when the sun sets or before it rises the sky is pitch black. The stars will suddenly appear. Its absolutely awful to look at and I really, really hope KSP2 can fix it.

[Xd the great]

2 hours ago, Motokid600 said:

A proper day/night skybox transition is high on my list. Right now when the sun sets or before it rises the sky is pitch black. The stars will suddenly appear. Its absolutely awful to look at and I really, really hope KSP2 can fix it.

I am looking foward to launching my kerbals to doom in a beautiful, romantic sunset.

[Xd the great]

New idea: Give kerbal skydiving ability.

Basically, being able to toggle between a head-first dive/tummy-first (starship-style) dive for kerbals, and add maneuverability for kerbals while midair. I want to try nailing the flag pole from 30000m up

And add the ability for them to hold hands. Both in the air and on the ground. For skydiving, as well as cute photos of Jeb and Val holding hands.

Spoiler

Holding hands for the last time before a lander crushes them.

 

Edited November 24, 2019 by Xd the great

[B15hop]

Couple more things I thought of.. 

Skip time in addition to time warp. Something that allows us to jump to apoint rather than sit and wait extremely long time to get where we want in large orbits.

Also, and this one is a big deal I think. The ability to make bases on other planets and store resources to build parts. 

For example, one I beam could cost X amount of ore, etc..

That way we could farm materials and build what we can from what we work for.

[linuxgurugamer]

1 hour ago, B15hop said:

, and this one is a big deal I think. The ability to make bases on other planets and store resources to build parts. 

For example, one I beam could cost X amount of ore, etc..

That way we could farm materials and build what we can from what we work

You Do know that they already said there would be base building?

Edited November 27, 2019 by linuxgurugamer

[mcwaffles2003]

9 hours ago, B15hop said:

Couple more things I thought of.. 

Skip time in addition to time warp. Something that allows us to jump to apoint rather than sit and wait extremely long time to get where we want in large orbits.

Also, and this one is a big deal I think. The ability to make bases on other planets and store resources to build parts. 

For example, one I beam could cost X amount of ore, etc..

That way we could farm materials and build what we can from what we work for.

I'm hoping there's a bit more complexity to resources this time beyond base "ore" 

This way we would be incentivized to scan a planet from orbit, then maybe take samples in person (or maybe remotely, but I'd like kerbals to be more useful) to find premium locations for building bases or at least excavating.

[qwe9000]

I really hope KSP 2 will support DX12. From what I've seen, the current graphics are next-gen compared to the original KSP, but DX12 introduces a lot of new graphics options like ray tracing.
P.S Can anyone tell me if KSP uses any version of DX/D3D?

[Polnoch]

Could you please add a native support for VR? It will be amazing, in one hand, create a vehicles using VR, in the real 3d, by hands controllers, and in another way, to pilot vehicles in first-person view in the VR helmet. Also to visit Duna, eve, Minmus in VR, etc.

[Guest]

• Decent aerodynamics
• Decent aerodynamics
• Decent aerodynamics
• Decent aerodynamics
• Symmetrical and cambered airfoil wings
• Decent aerodynamics

[SciMan]

On 11/28/2019 at 7:09 AM, qwe9000 said:

I really hope KSP 2 will support DX12. From what I've seen, the current graphics are next-gen compared to the original KSP, but DX12 introduces a lot of new graphics options like ray tracing.
P.S Can anyone tell me if KSP uses any version of DX/D3D?

KSP already uses DirectX 9 by default, and you can force it to operate in DirectX 11 to get some performance benefits (but you need patches to fix some texture glitches that causes, some parts VAB icons get a strong blue tint to them without those patches).

To me, ray tracing isn't magic. The improvement that it gives is IMO only really noticeable in mostly still scenes with specific lighting situations. The current tech works well enough for "suspension of disbelief" purposes. I plan on getting an RTX 2070 within the next few months, so I'll have a card that supports it, however if KSP 2 has ray tracing but turning it on kills the frame rate even on a card that supports it, I'll turn it off.

On the other hand, there are many other performance benefits to be had that don't even need a switch to DX12. Two that come to mind are "don't hold the graphics textures in system RAM after you sent them to the graphics card", and "Only send textures to the graphics card when they are needed". Both of those are pretty much standard practice in video game design, and for good reason. The first one is just an oversight in the coding of the original KSP, and fixing it will massively reduce RAM usage. The second one is called "load on demand" and it can introduce some small amount of stutter, but I doubt it would be much more than we currently get when a craft loads in to physics range (unless the craft has modded parts that use giant textures for some reason. That would cause significant stutter, but that's a problem with the parts, not the game).

[Polnoch]

9 hours ago, SciMan said:

To me, ray tracing isn't magic. The improvement that it gives is IMO only really noticeable in mostly still scenes with specific lighting situations.

I agree. VR is more impressive technology. Especially for space sims.

 

Quote

DX12

Better to support vulcan. Similar benefits, including performance and multi-threading code, but supports other platforms, not only Win10: Linux, win7, OS X

Edited November 30, 2019 by Polnoch

[Skorj]

I'd really like to see the stars in the skybox fixed to be 1-pixel stars.  The big, blobby stars we have now, while pretty, are my personal "cavemen riding dinosaurs" realism pet peeve. 

On 11/28/2019 at 2:23 AM, mcwaffles2003 said:

I'm hoping there's a bit more complexity to resources this time beyond base "ore" 

This way we would be incentivized to scan a planet from orbit, then maybe take samples in person (or maybe remotely, but I'd like kerbals to be more useful) to find premium locations for building bases or at least excavating.

I was thinking just the same, but I've been struggling to think of what they might actually be.  All I can think of is metallic ores, organics (for fuel and life support if that's in game), and maybe He3?

But instead of  more realistic resources, we could also just have game-y ones, where you just need ore from specific places in order to progress, because of Kerbal physics.  I'd love that.

[mcwaffles2003]

24 minutes ago, Skorj said:

I'd really like to see the stars in the skybox fixed to be 1-pixel stars.  The big, blobby stars we have now, while pretty, are my personal "cavemen riding dinosaurs" realism pet peeve. 

Stars arent actually going to appear as single pixels to anything without an insanely big aperture. Eyes, camera, or even telescopes as images taken through apertures all are limited by the rayleigh criterion

http://astronomy.swin.edu.au/cosmos/R/Rayleigh+Criterion

It makes it so stars appear fuzzy to some scale no matter what and is what limits the resolution of any image captured. That said, I would also like the stars tightened up a bit too, but 1 pixel is much too far. 

please look over https://www.reddit.com/r/astrophotography/

24 minutes ago, Skorj said:

I was thinking just the same, but I've been struggling to think of what they might actually be.  All I can think of is metallic ores, organics (for fuel and life support if that's in game), and maybe He3?

But instead of  more realistic resources, we could also just have game-y ones, where you just need ore from specific places in order to progress, because of Kerbal physics.  I'd love that.

Even if its as base of systems as metal, air, chem fuel, and nuclear fuel I would be happy. There could be definite distinctions between fuel types as chem fuel (H2, methane, kerosene) and nuclear fuel (metallic H, He-3, and U-235). Mods could expand it out from there in an angel/bobs run of factorio type of way.

 

I see definite mod potential along this route as processing buildings could be made and a refinement process could take hold at colonies

Edited November 30, 2019 by mcwaffles2003

[Skorj]

32 minutes ago, mcwaffles2003 said:

Stars arent actually going to appear as single pixels to anything without an insanely big aperture. Eyes, camera, or even telescopes as images taken through apertures all are limited by the rayleigh criterion

In my limited understanding that's something different - it's saying that multiple light sources are all one "pixel" if they're under that resolution in separation.  Stars all appear to my human eyes as smaller than a pixel on my monitor, as they're so very much smaller than the resolution the eye is capable of. 

In any case, anything that shows a disk should be a planet, that's my particular pet peeve.

[mcwaffles2003]

52 minutes ago, Skorj said:

In my limited understanding that's something different - it's saying that multiple light sources are all one "pixel" if they're under that resolution in separation.  Stars all appear to my human eyes as smaller than a pixel on my monitor, as they're so very much smaller than the resolution the eye is capable of. 

In any case, anything that shows a disk should be a planet, that's my particular pet peeve.

This is the equation for how big a point source of light will appear projected through any aperture (human eye in our case)

Any difference below the limit theta for your light sources color/width of your pupil is indistinguishable. and most of our eyes cant see perfectly so the limit is higher than the absolute physical limits.

Single monitor pixels are normally smaller than any of us can see now individually so literal single pixel stars would be nearly invisible.

Though, if you're used to barely seeing stars from a city/suburb seeing stars as small as single pixels I could see your case. But as an amateur astrophotographer getting stars down to as small as your speaking even with decent equipment is difficult enough, our eyes are worse than the equipment taking these and the stars are much larger than single pixels

Edited November 30, 2019 by mcwaffles2003

[qwe9000]

9 hours ago, Polnoch said:

I agree. VR is more impressive technology. Especially for space sims.

 

Better to support vulcan. Similar benefits, including performance and multi-threading code, but supports other platforms, not only Win10: Linux, win7, OS X

Unfortunately, AMD Vulkan GPUs are less mainstream than Nvidia RTX GPUs. Most people who use ray tracing related things are probably on Win10.

Though I expect the devs to use DX11 or OpenGL because of KSP's simple graphics control interface.

[Skorj]

16 hours ago, mcwaffles2003 said:

Single monitor pixels are normally smaller than any of us can see now individually so literal single pixel stars would be nearly invisible.

I suspect we're talking past each other.  I'll try to clarify my terms.

There's no absolute minimum to the angular size we can see, it's a matter of brightness.   I can certainly see a single white pixel against a black background on my monitor, just as I can see some stars.  But perhaps you're talking about when we can no longer see the difference between 1 and 2 pixels.

For comparison, a "pixel" on my monitor (27" 1440p at ~36") is about 47 arc seconds (so the individual OLEDs are probably ~20 arc seconds), while Pollux (to pick a bright star at random) is about 0.008 arc seconds.  Now, even with my pupils fully dilated on a very dark night, Pollux (or any star I can see at night) is going to be smeared out across ~16 arc seconds by my "mark 1 eyeball", but that's basically the size of a pixel, no? 

I mean, sure, on a modest-sized 4k monitor a 2x2 pixel area looks about the same, as a 1x1 pixel area, but that's not at all what I'm talking about: there are stars in the KSP skybox that are 3mm across on my monitor.  Stars big enough to look like a disk instead of a pinpoint are everywhere.

[paul23]

On 11/14/2019 at 4:02 PM, DStaal said:

#1 I don't think is particularly important - by the time you're ready for a space elevator, launching to orbit is routine busywork.  Especially if it's a choice to get it built, giving players a way to avoid tedium as the game progresses is probably a good idea.  (And there would be new challenges with having it in orbit - particularly since in KSP equatorial orbits are easy and the default orbits you launch into.  If you've got a couple of space stations and decide to build a space elevator, well, first you'll want to make sure those stations don't hit the elevator...)

#2 is probably fatal to it however.

You do realize that space is not "high", it's closer than most cities are. - It's "fast".  xkcd shows.

 

An elevator won't really help you there: unless you wish to go to geostationary orbit. But that is just sheer impossible to create, we're going out of the science fiction and into the magical constructs if you go there, there is no molecule able to withstand the forces of itself.

 

And even if you do that, getting the stuff "up there" requires quite a bit of energy: incidentally, exactly the same as an ideal rocket would take. - Only no gravity loss and drag is probably also minimized.

[DStaal]

9 minutes ago, paul23 said:

You do realize that space is not "high", it's closer than most cities are. - It's "fast".  xkcd shows.

An elevator won't really help you there: unless you wish to go to geostationary orbit. But that is just sheer impossible to create, we're going out of the science fiction and into the magical constructs if you go there, there is no molecule able to withstand the forces of itself.

And even if you do that, getting the stuff "up there" requires quite a bit of energy: incidentally, exactly the same as an ideal rocket would take. - Only no gravity loss and drag is probably also minimized.

Sure - I mean, I play KSP.  I've put rockets into space without putting them into orbit.     However, an elevator helps with the 'fast' as well - as it drops you in geostationary, at speed.  (And if you want to transfer to someplace else, you can now use a high-ISP vacuum engine...)  And while it does take the same amount of energy as an ideal lifter, it would allow you to choose from a much wider range of possible engines, and you wouldn't have to lift any fuel as well.  (It could be lifted otherwise, or you could send power through the elevator.)

It's also right on the boundary of science fiction - we think we can make materials strong enough, but not in the lengths required.  For that matter, an elevator on the Moon or Mars would be much easier (within current materials), and have many of the same benefits.

All of which is really irrelevant to this discussion - Unless there's a more-major-than-we-expect overhaul to the KSP engine, KSP2 isn't likely to be able to support a space elevator any more than KSP1 can.

[strider3]

The ability to build a space station without the idiotic "death wobble". I have tried various construction techniques/auto-strutting/EAS struts until I'm red (not blue ) in the face. Nothing really seems to work except a massive amount of ugly EAS struts...if I can get them attached before the thing "wobbles out". That's a very big if, BTW.

[MR L A]

On 11/21/2019 at 8:47 AM, garwel said:

Yes, and gallons.

Because NASA uses metric  and tbf... gallons and miles are literally defined using a metric value