Science in KSP2

[Pthigrivi]

On 11/15/2021 at 11:56 PM, SOXBLOX said:

The nodes should be unlocked by throwing time and funds (or the equivalent thereof) at them, through a constrained development pipeline. Experiments shouldn't directly unlock new tech; they should pave the way for more research.

This is a common ask, and its not crazy in principle, if you can solve the time-warp problem. I think you have to solve that problem anyway through because of colony scale ISRU. After all, why build up mining capacity or make things more efficient if you can just time warp to infinite resources? You need some manageable disincentive to endless time-warp that doesn’t pinch players later in the game when they’re on long transfers. Life support works well for this despite some of the controversy, but there are other solutions as well. 
 

To me it seems like there should be a built-in but manipulable rhythm to time warp that scales to progression as players move deeper into space to match longer and longer transfers. For instance right from the get go your first few missions getting to orbit are only going to take a couple of hours in game to complete. Then half a dozen round trips to the Mun and Minmus will only take you 60 or 70 days. But your first Duna window doesn’t happen until day 236, and then its a 300 day transit. And maybe you might want to do something in the meantime like get your station going but why would you when you could just fast forward to the probe landing to get a whole batch of better parts? You end up with these herky-jerky development periods where a lot happens and then you time warp for months on end. It be nice if this was smoothed out in a more manageable way that really made you think carefully about time as a resource. 
 

The truth is real R+D is obviously too complicated and unpredictable to be emulated exactly and be fun. In any case we’re talking about a simplified abstraction, and all things being equal the simpler the better. The way Ive been thinking about it designing an electric wheel is trivial. Whats not trivial is knowing enough about the surface of the moon to make an electric wheel that performs well there. You need to understand its consistency and composition, temperature range, radiation exposure for sensitive electronics. This to me is what the KSP science system is emulating, a recursive process of exploration and tighter and tighter understanding of design constraints. You gather data about environments and then your kerbals build up a knowledge base and experience level that allows them to produce new parts. And to me thats where the time should come in—after data has come in and until it’s processed into the abstract quality of “science”. You fly your first high altitude flight and gather data, and then you can either fly again right away with the same parts or wait a few days for the data to be processed and get new parts to work with. Over time you can upgrade your R+D facility or establish science labs offworld to process data faster and speed up your development cycle to handle more and more complex technologies. Its not an exact emulation, nor could it be. But it smooths out your program’s development in in-game time and doesn’t hang players up in with time-warps at inopportune times like between a settled design in the VAB and going to the launchpad. 

Edited November 18, 2021 by Pthigrivi

[Staticalliam7]

On 11/14/2021 at 4:53 PM, Pthigrivi said:

So I took a stab at a better tech-tree system. This one would use Boom events to unlock each tech-tier, after which you could spend Science to unlock the nodes within that tier. I started with 5 main tech lanes: rockets, exploration, structure, electronics, + aerodynamics, which would open up and expand as more exotic engines and technologies became available. I've also just given the experiment parts as Boom rewards. This would take you through your first interplanetary missions and unlock most of the parts in KSP1. It's too hard to guess what would come after. 

Start - (Command Pod, small chute, basic fin, Flea, small hardpoint, Swivel, small 1.25m tanks, thermometer, barometer, surface sample)
Initially available tech nodes > 

Basic Rocketry - (Hammer, Terrier, Large 1.25m tanks)
Survivability - (1.25m heat shield + service bay, radial chutes)
Stability - (Basic decouplers, Strut connector, launch clamp)
Electronics - (Stayputnik, small PV + batteries, Communotrons)
Aerodynamics - (Winglets, nosecones, 1.25m fairing)

First orbit unlocks > Terrain + Biome Scanner, Dosimeter 

General rocketry - (Reliant, Thumper, Ext. Fuel Duct.)
Space Exploration - (Mk1 Lander can, ladders, lights, LT-1 struts, command seat, foldable wheel)
General Construction - (Adapters, couplers, girders)
Probe Tech - (Basic probecores, Rovemate, medium PV + batteries, S2 Wheel, HG-5)
Basic Aviation - (First tier plane parts)

Landing on either Mun or Minmus > Infrared + Resource Scanner, Core Samples

Heavy Rocketry - (Skipper, Poodle, Kickback, 2.5m tanks)
Propulsion Systems - (Thud, Twitch, Spark, External tanks, Sepratron)
Advanced Flight Control - (RCS, reaction wheels, 1.25m clampotron, MK1-3, MK2 Lander can)
Advanced Exploration - (Science Lab, Hitchhiker, 2.5m service bay + heat shield, Mk25 + MK-16-XL chute, LT-2 landing struts, LES)
Heavy Construction - (2.5m decoupler, stack separator, nosecones, adapters, + fairing)
Micro Engineering - (.625m reaction wheel, heat shield, decoupler, stack separator, clampotron, + small girders + LT-05 struts) 
Advanced Probes - (Probodyne QBE, HECS2, OKTO2)
Communications - (DTS-M1, RA-2, HG-55, RA-15)
Advanced Aerodynamics - (Medium wings, elevons, fuselage, crew cabin, landing gear)
Air-Breathing Engines - (Wheesley, Panther, nacelles, air intakes)

Crewed landings + returns from Mun + Minmus unlock 1st R+D upgrade > Radiation + Atmospheric Scanners, Atmospheric Analyzer

Advanced Fuel Systems - (Mainsail, Twin Boar, Vector, Large RCS tanks, C7 + fuel laden adapters)
Precision Propulsion - (Spider, Ant, Puff, Vernor, .625m tanks)
Station Tech - (Shielded, In-line + Clampotron SR, Cupola, First station parts, LS)
Surface Exploration - (TR-2L, Medium rover parts, storage parts)
Specialized Construction - (2.5m couplers, I-beams, panels, girders)
High Powered Electronics - (Gigantor, Variants, Z-1k, Fuel Cells, PB-NUK)
Deep Space Probes - (RC-001S, 88-88, RA-100)
Supersonic Flight - (Whiplash, advanced intakes, Mk2 parts)
Heavy Aerodynamics - (Goliath, large plane wings, Mk3 parts)

First Interplanetary landing > Anomaly Detector, Robot Arm

Very Large Rockets - (Rhino, Mammoth, 3.5m + 5m tanks)
Nuclear Propulsion - (LV-N, Stubby 1.25m variant, LH2 tanks)
Ion Propulsion - (Dawn, 1.25m variant, small + medium xenon tanks)
Large Stations - (3.5m modules + station parts)
Colonization - (First surface colony parts, upgraded science lab)
ISRU - (First small drills + converters)
Advanced Alloys - (3.5m Fairing, decoupler, adapters + engine plates, large trusses)
Nuclear Reactors - (.625m + 1.25m Nuclear reactors, Uranium tanks)
Specialized Electronics - (RC-LO1, Z-4k, Large PV arrays)
Radiators - (First radiators + thermal control)
Spaceplanes - (Rapier, Aerospike, Spaceplane wings)

we need a mod that implements this! 

[SOXBLOX]

1 hour ago, Pthigrivi said:

This is a common ask, and its not crazy in principle, if you can solve the time-warp problem. I think you have to solve that problem anyway through because of colony scale ISRU. After all, why build up mining capacity or make things more efficient if you can just time warp to infinite resources? You need some manageable disincentive to endless time-warp that doesn’t pinch players later in the game when they’re on long transfers. Life support works well for this despite some of the controversy, but there are other solutions as well.

I didn't even think about time warp playing in to this...

Hmmm... Maybe it takes time + funds/science points/milestones. Then the time dimension still plays a role. Otherwise, you're right; it becomes superfluous.

Maybe the time-sensitivity of contracts should be bumped up?

Isn't there actually a mod that does something kinda like this? Now that I think about it, I seem to remember something...

[desert]

I really like how Kerbalism did science, by distinctly separating samples and data. Stock is not intuitive in these terms, it shows you science which is green but then it is also blue and if you measure something  twice you get bit less blue. And if you put a single thermometer measurement into lab you can farm a lot of science from it but if you move this poor thermometer measurement to other planet you can get squeeze even more science from it with the lab. It makes no sense. A simple and understandable connection of samples, data and science points is needed. Kerbalism does that, but it could use better UI.

Experiments taking time and electricity is a must-have too. Energy balance is an important part of any spacecraft and currently it is almost not a thing. Probes can be set to standby, antennae can send data partially and most stock experiments don't require energy at all and few use it as a single large portion. Again, Kerbalism did this pretty good, but it needs better UI and there should be consistency of science production for sensors, scanners, reports, surface experiments and other stuff.

As for the tree, it should probably be unlocked with science only. And milestones, resources or funds (well,  we won't have them, but maybe at least some funding API for modders?) should be tied to actual production of crafts, colony modules and their upgrades. Such separation of science and production would be easier for players to understand.

[SciMan]

I really like the idea of the tech tree being split up into sub-trees by a number of things:

1. What is the thing doing

• Propulsion (engines, maybe rover wheels)
• Structural (Girders and I-Beams and fuselages and multi-couplers and size/profile adapters, all that stuff)
• Aero (any and everything to do with aerodynamics, including parachutes, wings, control surfaces, nosecones, airbrakes (hopefully we get a few more sizes of those to work with, and a heat-tolerant version to use for Falcon-9 style reusable first stages))
• Command (pods and probe cores)
• Attitude Adjustment (Reaction wheels, RCS, and potentially guidance system add-ons (like the guidance nose-cone but with other features and form factors available))
• Power supply (solar panels, fuel cells, RTGs, probably nuclear fission and/or fusion reactors)
• Power storage (batteries, maybe capacitors if modeling the finite discharge rate of batteries is put in the game)
• Ground interaction (landing legs, landing gear with wheels, maybe crawler treads, maybe other means of getting around on the ground)
• Robotics (servos and motors and pistons)
• Habitation (parts that hold crew but aren't command parts, could also be things like the science lab or other crewed modules that do more than just hold crew)

2. What is the shape of the thing (This is best summed up by the various "bulkhead profiles" you can select parts by in the VAB in KSP 1, but I can think of more categories):

• Cylindrical profiles (in KSP 1 we have 0.625m, 1.25m, 1.875m, 2.5m, 3.75m, and 5m, but I'm pretty sure we'll be getting larger steps in KSP 2)
• Spaceplane fuselage profiles (in KSP 1 we have just the Mk2 and Mk3 spaceplane profiles, but there might (hopefully) be more added in KSP 2 (and hopefully without the unintuitive drag penalty from having a lifting surface module in them that sees the flat bulkhead profile of the end of the part square-on to the airflow "Yep this should create a lot of drag and a very tiny amount of lift, no matter the speed"))
• Structural beam profiles (in KSP 1 we have cubic octagonal strut, I-Beam, and modular girder profiles, but I expect this is where "mothership central truss" type pieces would go in KSP 2)

3. Where is the thing expected to be used:

• Rocket Parts
• Colony/Base/Outpost Parts
• Mothership Parts (if they're not just mixed into the "Rocket Parts" tree)

 

And the key feature of all this subdivision is that you'd be able to buy "just one part" without getting a bunch of useless other parts (or at least parts you can't use right now, or needed 5 tech nodes ago, or that kind of thing). Basically a way to get what you need, when you need it, without paying for any extra things you don't want (exactly the opposite of what's happening with the "Newegg shuffle" ).

Edited November 20, 2021 by SciMan

[MagicCuboid]

On 11/18/2021 at 8:41 AM, desert said:

I really like how Kerbalism did science ... Experiments taking time and electricity is a must-have too

Ditto to that! For me, the Science overhaul is the reason I can never get rid of Kerbalism. The primary reason is simple: it takes Science and integrates it much further into the engineering process of ships.

In Vanilla, Science is pretty simple: you pick whatever experiments you want and clamp them onto the side of your rocket. For probes, you need an antenna strong enough to send the signal, and that's about it.

In Kerbalism, Science is integrated much further into the electricity and communications systems, as well as probe core selection. Every experiment is passive - so long as it's on, it will passively log Science in the ship's memory. Once the ship's hard drive is full, it can't log anymore science! But, it will happily transmit the Science back to Kerbin. This is limited a) by connection to the DSN, and b) by the bitrate capabilities of your antenna. 

This way, power generation, antenna signal strength, transmission speed, and hard drive storage space are ALL potential bottlenecks for your Science. It's fun to design crafts that can manage multiple experiments at once, and it realistically can take months to harvest all the Science due to slow transmission speeds. The upside? No clicking! Kerbalism manages all of this on its own, and you can manage each craft off-screen without having to load it in.

I don't think KSP2 should directly copy this system by any means. I'm sure there are ways to slightly simplify it and make it a bit more user-friendly. But I hope they keep the core ideas that communications, power generation, and probe core tech should matter to Science retrieval, and that crafts should be manageable from a control menu rather than having to load them in each time.

[SciMan]

If they do make it work like Kerbalism, where you have multiple factors that each could be bottlenecks, I hope that they have extra parts that can make up for shortfalls in any of those departments.

The power generation and antenna signal strength (and partially the transmission speed) are already handled well by KSP 1, but if it was me I'd separate antennas into the physical Antennas themselves, and Transceivers which are the electronics that make the antennas actually do something.

That way they would be better able to tune the individual parameters without having to make antenna parts that have mostly the same ranges but differ in speed. All probe cores would include a basic antenna and transceiver module by default.

So for instance, if you were designing a satellite to map the Mun by using a radar altimeter, you wouldn't need a particularly directional antenna, but you would need a rather high bandwidth transceiver, and maybe a smallish hard drive to store data from the far side of the Mun.

However, for something like taking temperature measurements of the surface of Eeloo, the situation would be mostly reversed, because you would need a gigantic antenna, but you could probably get away with using the probe core's internal low-bandwidth transceiver, and again a tiny hard drive (perhaps just the one included in the probe core).

Of course, you might also be able to have special tranceivers and/or antennas that would be able to combine their efforts with other parts, therefore increasing either signal strength or bandwidth.

You might also even have RF amplifier parts, which would be non-antenna parts that still boost signal strength by means of amplifying the signal from whatever connected antennas are available (best for SENDING data, not receiving it, and might have a bandwidth limit of their own).

[Pthigrivi]

Yeah the whole science system risks being overly finicky, but people have enjoyed the transmission system. I still think it has problems, mainly graphically, in how well a player can understand dish ranges and how many batteries they'll need before its too late. I think what's fun about it is the spatial problem: thinking about LOS and bouncing signals off relay probes to stay in communication. I'd be inclined to simplify data transmission even more, and I definitely agree on treating surface samples differently from other kinds of data. It seems like we're talking about 3 kinds of experiments: 1) Instantaneous data like thermometer readings and radiation levels. There isn't a huge amount of data here so I would think these could upload instantly and automatically when connected to Kerbin or a nearby science lab. No clicking or fussing, it just happens. Any probe core should be able to handle this. 2) Continuous data like orbital scans: I think these should be able to upload as they scan so long as they're connected to the network. Maybe some probe cores are rated to collect and store information when not in communication and update when they regain a signal. This would give players the choice of using smaller, cheaper probes in a relay-network or using one big one in a polar orbit. 3) Samples. These can't be transmitted without being brought to Kerbin or analyzed off-world in a science lab. Maybe deep into the tech tree there's a way to autonomously analyze samples with fancy probe landers. The important thing is that players can tell how their equipment is going to perform in the VAB before they send it all the way to another planet. This could be as simple as rating probe cores as "Requires signal for data collection" and "Can collect and store data without signal". I don't personally like that you have to have enough power to transmit 100% of the data at once or you're blocked, because it means guessing how valuable the data is going to be in order to know how many batteries to bring. Just leave transmission speed and cost/bit out of it. Instead the experiments, probes, and dishes should have flat EC requirements while running so that you can just add up the total draw and provide solar panels and batteries to cover it. I'd also personally make all transmitters capable of relaying. That way all you're worried about is that each probe has enough range to reach another that's connected back to Kerbin or to a local science lab or probe control point. 

Edited November 29, 2021 by Pthigrivi

[shdwlrd]

43 minutes ago, Pthigrivi said:

I don't personally like that you have to have enough power to transmit 100% of the data at once or you're blocked, because it means guessing how valuable the data is going to be in order to know how many batteries to bring. Just leave transmission speed and cost/bit out of it.

That was one of my biggest pet peeves with the science and career modes in KSP1. Even the early satellites could figure out when it was time to transmit data and was able to pick up where it left off. The other is that there was no way in vanilla to determine how much EC you needed until you were already on mission. None of the power conversations made sense, nor was there any explanation of the units used.

[SciMan]

Since you can store it, the closest thing I can think of for "a unit of the Electric Charge resource" is a Kilowatt hour (KWh), or 3.6 MJ. That's the only kind of energy value that makes the power consumption of the ion engine make sense (they use REALLY QUITE A LOT OF POWER).

Tho in general, the power consumption of everything in KSP is off in some way if you actually pin down a conversion factor between joules and electric charge. (it can't be watts, because it can be stored in batteries, and batteries don't store watts aka power, they store watt HOURS, aka power over time, aka energy).

In other words, either the power consumption of ion engines (and the ISRU parts, so drills and converters and maybe even radiators) should go way UP, and then make Electric Charge worth less energy per unit, or the power consumption of most everything else (except the stuff I just mentioned) should go way DOWN, and Electric Charge can keep it's value of 1 KWH per unit.

Probes are just a bunch of electronics, that doesn't take much power. Same with science experiments, they don't need much power compared to the real power hogs of ISRU and ion engines.
Reaction wheels should probably take more power than they do, maybe 2.5-5x what they do now.
Antennas, actually their power usage should in general be reduced, but the longest-ranged ones should likely have their power requirements reduced less than the basic ones you start off with.
Rover wheels and robotics are I guess in a good place as far as power consumption, it's "enough to matter" but it's not "you need to make your craft use batteries as structural parts" high.

All probe cores should be able to survive at least 5 orbits around Kerbin on internal batteries (75km circular orbit, equatorial). Probe cores designed for deep space should have more internal power reserves.

 

Oh and ISRU converters should be able to automatically restart upon getting enough power if the vessel can't make enough power to run everything for a short period (like if the sun sets and they're being powered by solar panels), but I guess that's a separate thing. That right there is why I always use either lots of fuel cell arrays to power my mining bases, or if I have the mods for it a nuclear reactor. It's just too much trouble (and too heavy) to pack enough batteries to support a 2.5m ISRU converter and a bunch of drills overnight, even on Minmus where the night isn't as long as it is on the Mun.

[sciencecompliance]

I've seen a lot of great comments about how the science system should be improved for KSP2, but I have another issue with the science system / tech tree from KSP1.

Having gone through almost the entire tech tree on hard mode, I've found airplanes to be almost entirely useless for gaining science and progressing through the tree.  Why would I spend hours flying a jet airplane around Kerbin and collecting data from all the different biomes and then go through the difficulty of piloting an airplane safely to a runway / horizontal landing  if I can fly an endless number of ballistic missiles on different trajectories around Kerbin much more quickly and use reliable drogue / parachute landings in a capsule with a payload bay containing experiments to quickly and easily farm science?  There are also no achievements tied to aviation that will hold you back in the tree progression, so there is basically no reason to build airplanes if you are trying to optimize your speed of progression through the game.  This also ties into another issue with science mode in KSP1 that people have touched on but I think underscores the point I'm trying to make about building airplanes:  budget.

I understand that budget is a consideration in career mode, but it should be a constraint in science mode as well.  In reality, launching a lot of rockets around a planet to study all the different biomes / regions would be much more expensive than having science labs in a single airplane or a few airplanes that are flown around the planet to study the different biomes.  Having some game mechanic that incentivized building airplanes rather than just rockets due to airplanes' relative cost effectiveness in certain applications would both be more realistic and force the player to explore the depth of aerospace technology as they are advancing through the game.

[Pthigrivi]

21 hours ago, sciencecompliance said:

I've seen a lot of great comments about how the science system should be improved for KSP2, but I have another issue with the science system / tech tree from KSP1.

Yeah the plane thing is tricky. I've made a fair few planes for fun in sandbox and messed around with spaceplanes quite a bit, but it never felt like a core game experience. Part of it is that landing planes in KSP is notoriously tricky. It's certainly doable with practice but getting to orbit and even landing on the Mun is much easier. I know a lot of that has to do with players not being great at designing planes with high stability at low speeds, so that could be a tutorial. But still given that the focus tends to be on space exploration and now colonization planes feel like a niche use. They have a devoted following of players who love them and probably some applicability on Laythe and maybe Duna and Eve, but the overall architecture of a program isn't built around them the way it will be built around ISRU and propulsion tech. 

The other issue, which isn't exclusive to planes, is just how vast these worlds are. Going any real distance is an exercise in extreme patience. I think both for planes and for rovers some kind of autopilot that can at least cruise on a heading would be really nice. No word on that yet from Intercept. Without it I think the reasonable range players could be expected to drive or fly is pretty limited. One really nice bit of news from episode 4 is just how much work they're putting into crafting landscapes to make them more varied, and it sounds like they wont just be window dressing, that they're thinking about how that can be integrated with gameplay as well. Im really hoping for science hot spots that require more precise exploration warranting things like planes and rovers. I think even in the early stages on Kerbin having a lot of mostly nondescript biomes, all with equal science value, doesn't make for super compelling gameplay. If however you had specific locations you were seeking out and landing/traversal puzzles to reach them it might be more interesting. I think you could also include some earlyish boom event rewards for speed and altitude records, but those tend to get ticked off by rockets in the first few launches. 

I guess my feeling is plane exploration could be an interesting side-hustle for pulling in extra science but I don't think it should be central or required to advance. 

Edited November 30, 2021 by Pthigrivi

[sciencecompliance]

4 hours ago, Pthigrivi said:

Part of it is that landing planes in KSP is notoriously tricky. It's certainly doable with practice but getting to orbit and even landing on the Mun is much easier. I know a lot of that has to do with players not being great at designing planes with high stability at low speeds, so that could be a tutorial. But still given that the focus tends to be on space exploration and now colonization planes feel like a niche use. They have a devoted following of players who love them and probably some applicability on Laythe and maybe Duna and Eve, but the overall architecture of a program isn't built around them the way it will be built around ISRU and propulsion...

I think you could also include some earlyish boom event rewards for speed and altitude records, but those tend to get ticked off by rockets in the first few launches...

I mostly agree with you and don't think the core game experience should prioritize the construction of airplanes, however, I still maintain that there should be science points in early game that would be easier to gain with an airplane / jet-powered craft and that there should be some tech tree nodes that can't be unlocked without accomplishments made with jets / air-breathing engines and possibly runway / horizontal landings.

Think about it: does it make sense to make the RAPIER engines available to the player if they have never flown a jet engine at hypersonic speeds or in rarefied atmosphere?  Not really.  Does it make sense to make heavy landing gear available if the player has never made a runway / horizontal landing?  Again, not really.  In reality, accomplishments in rocketry and spaceflight will beget wisdom and further accomplishments along that technological trajectory.  There may be some cross-pollination, say, if you've landed landers with landing struts and driven rovers around with pressurized tires and brakes, but the most optimized landing gear is going to be designed with data from that application specifically.

If you think about rocket engines, too, many of the more efficient ones use turbopumps, which share a lot of commonality with jet engine technology.  It would make sense to require experience with high pressure/temperature turbines to unlock those tech nodes or at least get you there more quickly for having made accomplishments with jet-powered craft.

[SciMan]

If you had a set of speed records that were organized by altitude, with rewards given for each altitude bracket, that would encourage the use of aircraft, because then you'd be encouraged to maximize speed at all altitudes, whereas rockets are designed to maximize speed mostly at higher altitudes if not just outright entirely above the atmosphere.

This would also naturally lead into teaching the player how to pilot a spaceplane if the greatest rewards were given for setting a high (say 1500m/s or greater) speed record at a "middle" altitude, so not skimming the wavetops but instead at an altitude of say 5000-7000m, which seems to be the altitude where air resistance is low enough but jet engine thrust is still high enough that you accelerate the fastest, given that your air-breathing engines can keep up of course.

And you don't "need" the RAPIER engines to do that either, in fact I find that I get better results with 2 or more whiplashes for gaining speed and an aerospike engine or Poodle for actually making orbit. This is due to the curious "subsonic-to-supersonic thrust kink" in the RAPIER's air-breathing mode, which confuses me as to why it exists since shock cone intakes should easily dispose of such a problem.

[Pthigrivi]

Great points both^. Have either of you used planes on Eve or Duna? Ive used spaceplanes on Laythe but all these years haven’t done much without air intakes.

Edited December 3, 2021 by Pthigrivi

[SciMan]

To be honest, Eve is a planet I've never really considered doing much of anything at as far as landing. If I did do something, it was just probes that had parachutes on them.
This is because I have a personal limitation of never sending Kerbals anywhere without already having a plan to get them back, and making a spaceplane that can get out of Eve's atmosphere is just not an appealing prospect (I'd rather use a rocket, and with Eve being Eve, I'd need more than one stage anyways).

Laythe on the other hand, I've landed on Laythe with a spaceplane before, and I've even made it back to Kerbin with that same spaceplane. However, that was before the aerodynamics overhaul (aka it was back when the "souposphere" was still a thing), so now I'm practically hopeless with spaceplanes in general.

In general, I just have better luck with rockets anyways, because the type of spaceplane I like to build needs large radiators for its propulsion systems because of modded engines, and apparently according to KSP, radiators might as well be wings with the way they make such spaceplanes always flip out no matter what I do, tho now I do have B9 PWings installed so maybe I can avoid that.

Additionally, now that we have the robotics parts, if I were to make a spaceplane for Eve usage, I'd be using the ducted fan parts to ensure that all I need is a steady supply of electricity to allow it to climb to an altitude sufficient to enable me to make efficient usage of the rocket engines I put on it, instead of having to spam Vectors and/or Mammoths (with their high mass) because those are the only engines that really work well at all down low in Eve's atmosphere.

[Newgame space program]

What I want science to do is to allow you to learn about stuff more in depth then the original game did,

I want the exact composition of Duna's atmosphere dammit!

And the history of the Kerbol system!

 

[sunnypunny]

The way science drives exploration in career is good but it's also too repetitive and restrictive in being a hard requirement for accessing parts, and also unrealistic in the sense that you need temperature scans from the mun to unlock bigger engines.

I would move more towards enabling different playstyles - like you can get anywhere either through science, milestones or contracts and specialise in these areas. As for the tech tree I think less complexity is better otherwise it becomes a real nightmare to work through. Simply sorting parts into categories (e.g. by bulkhead size, crew count, etc) is much simpler and more widely compatible.

 

[K^2]

1 hour ago, maj. M.AL. funqt10n said:

you need temperature scans from the mun to unlock bigger engines

Technically, you can get all that science from KSC, by using the fact that various facilities and individual buildings, and in some cases their upgraded variants, count as unique biomes. So if you focus on getting all the scientific equipment first, do the upgrades, then collect all the science again, you can get a very significant chunk of the tech tree unlocked. It's slow, it's annoying, and it has nothing to do with the game, and yet, it's the path of least resistance to gather the science you need to have fun with the game.

Yeah, that system definitely needs changes.  

[Jefftheguyperson]

I think the science format of Kerbalism would fit nicely. It requires you to spend time in each situation in order to collect science. You can leave your experiments on, and they consume power while running, but they will automatically collect science when they enter a new situation.

[sunnypunny]

On 12/23/2021 at 6:02 AM, Jefftheguyperson said:

I think the science format of Kerbalism would fit nicely. It requires you to spend time in each situation in order to collect science. You can leave your experiments on, and they consume power while running, but they will automatically collect science when they enter a new situation.

the issue i have with this mechanic (and anything else time-based) comes from timewarp - if simply hitting warp gets you the same result you might as well just click one button ("run experiment") and get it instantly. the power requirement is also broken in space because solar power (and the RTG) is basically free - though again based on time, depending on orbital position or daylight hours, so you can just warp to get more power. timewarp x1000000 breaks a lot of mechanics like this and makes the solutions awkward. if they add things like interactible objects, collectibles etc in each area then that would add a reason to stay and go do the things - you could also flesh out the experiments to be sort of "mini-games" with graphs and data that require analysis and decision making by the player.

of course you have to design around these elements (mass, power output etc) when creating a vessel, so they still add something to the game - just in terms of science gathering i think a new approach is needed.

[Pthigrivi]

1 hour ago, sunnypunny said:

the issue i have with this mechanic (and anything else time-based) comes from timewarp - if simply hitting warp gets you the same result you might as well just click one button ("run experiment") and get it instantly. the power requirement is also broken in space because solar power (and the RTG) is basically free - though again based on time, depending on orbital position or daylight hours, so you can just warp to get more power. timewarp x1000000 breaks a lot of mechanics like this and makes the solutions awkward. if they add things like interactible objects, collectibles etc in each area then that would add a reason to stay and go do the things - you could also flesh out the experiments to be sort of "mini-games" with graphs and data that require analysis and decision making by the player.

of course you have to design around these elements (mass, power output etc) when creating a vessel, so they still add something to the game - just in terms of science gathering i think a new approach is needed.

I think what allows time-based science to work is to have it run at a steady rate for a relatively short period of time and then halt. For instance a thermometer could record for a full orbit or a full local day and then stop, or an orbital scanner could harvest a full surface scan a la SCANsat and then stop. That way you can't warp to infini-science. I do love some of the elements BG added with surface features though and I think they could definitely bring locating and studying unique locations more into the forefront. 

Edited December 29, 2021 by Pthigrivi

[DerkyJerkyWreaksHavoc]

i have a big brain idea...

Something i noticed about KSP2 is that there is going to be an extreme amount of technology ranging from primitive tech to advanced tech and part of me feels like there should be a second, more advanced tech tree for interstellar technologies. i'll call it "Quantum Science". it would work with different, more advanced science experiments that could only be accessed at its start by finishing the first tech tree. here would be all the anti engines, orion propulsion, more advanced electrics, "Quantum science" experiments, you name it. @Nate Simpson will definitely not see this so i am ok with publishing this. its a good suggestion that i hoped that would be implemented into KSP2's tech tree. i mean the first one is long enough; do we really wanna pile it up more? YES?! NO!!! i really think the 2nd tech tree would be a good implementation to the game. 

Edited December 29, 2021 by DAFATRONALDO2007 IN SPACE
and please add faster warp...

[Jefftheguyperson]

3 hours ago, sunnypunny said:

the issue i have with this mechanic (and anything else time-based) comes from timewarp - if simply hitting warp gets you the same result you might as well just click one button ("run experiment") and get it instantly.

The idea behind time-based experiments is that you can't collect every experiment instantly - a temperature scan may take only a second or two, but a materials study would require an in-depth analysis and perhaps long-term exposure, something that can't be done instantly. It balances the stock science formula this way - you can't just run through a biome and pick up every piece of science there, you may need to stay for a while and run experiments.

3 hours ago, sunnypunny said:

the power requirement is also broken in space because solar power (and the RTG) is basically free - though again based on time, depending on orbital position or daylight hours, so you can just warp to get more power. timewarp x1000000 breaks a lot of mechanics like this and makes the solutions awkward.

Hopefully, KSP2 solves a lot of the issues in this regard, because time warp certainly is absurdly buggy.

3 hours ago, sunnypunny said:

if they add things like interactible objects, collectibles etc in each area then that would add a reason to stay and go do the things - you could also flesh out the experiments to be sort of "mini-games" with graphs and data that require analysis and decision making by the player.

Hmm, that sounds like an interesting proposition. I would love to see how that would work!

[Pthigrivi]

So this might be controversial, but I think bodies in KSP have way too many general biomes, and rewarding players for hopping to each one is a huge source of grind. Honestly they don’t need more than 3 for any body. Id handle continued science differently, with BG style scatter-type surface features that could be clustered in certain regions, and a few geological hotspot mini-biomes like volcanoes, hot-springs, and signs of life. The latter two categories would require scans from orbit to locate and either precise landings or rovers to reach and perhaps special equipment to study. 

So for instance on Minmus you might have:

Biomes: Flats, Midlands, Poles. You can plonk down anywhere in one of these with a probe even and get data. 

Special Terrain Scatter - ‘Greenglass shards’ clustered in flats, can be picked up by a kerbal and returned to a lab for study, and Olivine formations, found only near peaks, and require a core-drill for study. 

Anomaly Mini-Biomes - the “Ice Crown”: a tall formation of sharp, icey-blue glass towers, the “Crystal Cave”: a large hollow with glassy stalactites and stalagmites, and “the Raisin”, a blackish red meteor fragment in a deep crater with steep walls. Each of these require orbital scans to locate and either a local science lab or plantable long-term experiment to study. 

So you might have 6 total science missions on Minmus if you were feeling thorough, but the last 3 would probably have to wait until you had the equipment to properly observe them. It would also give these bodies a few really special locations and make them feel more alive. 

Edited January 9, 2022 by Pthigrivi