Developer Insights #21 - Rockets' Red Glare

[Periple]

4 hours ago, Superfluous J said:

But it's not. The part must contact the water, so the base must be built near the water or there must be heat piping between the base and the water cooler. Or water must be pumped I suppose but anyway. So long as the water cooler part must touch the water, nothing tangible is gained in the gameplay by forcing it be submerged a set amount.

Plus you don’t even need a physical simulation if you want to check that the party is fully immersed, or add a cool animation that immerses it as long as water is within reach. Much like the scanner arms.

[schlosrat]

On 7/21/2023 at 5:35 PM, cocoscacao said:

TotalHeatIn, TotalHeatOut (who the hell names variables in pascal case these days???)

Microsoft for one: https://learn.microsoft.com/en-us/dotnet/csharp/fundamentals/coding-style/coding-conventions#naming-conventions

Edited July 23, 2023 by schlosrat

[Periple]

It’s a C# convention.

[PicoSpace]

33 minutes ago, Periple said:

It’s a C# convention.

Actually its even older then C#, I learnt that method of variable naming way back in Basic...

[Periple]

13 minutes ago, PicoSpace said:

Actually its even older then C#, I learnt that method of variable naming way back in Basic...

Unity games are written in C# so it’s relevant for KSP2.

[schlosrat]

17 hours ago, Pthigrivi said:

I kind of hope for the sake of simplicity both power and coolant lines can be transmitted underground within a certain range of the colony hub. Perhaps this range could be extended with upgrades?

I like this idea! Imagine that with various tech unlocked the distance between modules might be extended (like a max distance property for colony parts perhaps, where more advanced or otherwise capable parts might be able to be located further away). Whether the distance was fixed to one value for all parts or varied "booster pump" stations could be built so that you could daisy chain your way from one part of the colony to another. This could be really useful for heat transfer applications. A solar farm up on a ridge or peak, and a radiator farm down in a canyon for shadow perhaps. I think such a mechanic, regardless of how it's implemented, could lead to some interesting colony designs and site selection challenges!

On a related topic, such a mechanic might be useful also for resource transfer as well as heat transfer.

[cocoscacao]

2 hours ago, schlosrat said:

Microsoft for one: https://learn.microsoft.com/en-us/dotnet/csharp/fundamentals/coding-style/coding-conventions#naming-convention

Classes, interfaces and such. Not variables

[Scarecrow71]

12 hours ago, Pthigrivi said:

Keeping it on the subject of heat but also more broadly speaking it could be as simple as a buildable radius around your start position wherein all resources are shared and its effectively one vessel for pools of resources (power, fuel, heat-exchange). So very unlike SimCity you could make outbuildings or set up wind-farms wherever you liked within that radius and it would be considered connected to the network without having to manually place power poles or trench coolant lines. This seems more in vein with the way other systems work in KSP--you can put monoprop storage and RCS thrusters wherever you like. You're tasked with understanding how much they will consume and where they're best placed, but not with manually routing or piping fluids through your vessel. The analogy for a colony would be you might place solar panels where they get more sun or liquid heat exchangers on a coast, but you don't have to manually drag pipes and power lines to those locations so long as they're close to your main base.

But that's just SimCity.  You put down areas for population (residential, commercial, emergency), place your power/water stations, create lines to run from them to your zoned areas.  And that is NOT what we should be getting out of a SPACE EXPLORATION game.  I mean, if that's what you want, fine and dandy, and you are more than entitled to play however you want.  But this just seems to me like the developers are giving us a city management game with rocket building being an afterthought.

[schlosrat]

10 minutes ago, cocoscacao said:

Classes, interfaces and such. Not variables

True, but the important point seemed to be that there can be an equation, and the issue of the variable naming convention seemed like a distraction from that point.

I don't think the equation should be quite as simple as the one hypothesized, but that too is aside from the point. There will be some sort of equation, and it may not be a terribly difficult one if the heat generation/transfer/rejection mechanics are set up sensibly. By sensibly I mean with a reasonable balance between fidelity and gameplay so that the end result is interesting, fun, and sufficiently predictable that one can design with it in mind.

[The Aziz]

25 minutes ago, Scarecrow71 said:

SPACE EXPLORATION

Weirdly enough, setting up the basics for a working, safe base sounds to me like part of space exploration. I get that you don't want colonies at all, in any form because anything will be too much of a city builder for you.

Then maybe it's not the game for you and that's perfectly fine.

[Superfluous J]

  

On 7/22/2023 at 9:34 AM, Scarecrow71 said:

This game is about building rockets, not micro-management of colonies.  If I wanted a colony micro-management game, I'd go back to playing Civilization.  I'm here for the rockets, not the outposts.

36 minutes ago, Scarecrow71 said:

But that's just SimCity.  You put down areas for population (residential, commercial, emergency), place your power/water stations, create lines to run from them to your zoned areas.  And that is NOT what we should be getting out of a SPACE EXPLORATION game.  I mean, if that's what you want, fine and dandy, and you are more than entitled to play however you want.  But this just seems to me like the developers are giving us a city management game with rocket building being an afterthought.

You seem to be arguing both sides of the issue. One post, you don't want the micromanagement that you are afraid will come with colonies, and the next you are afraid colonies be way too simplified unless they're more complex than a game where all you do is build cities.

Personally laying out zones and letting the Kerbals build what they want sounds awesome to me. So does placing each building myself, making sure it's exactly how I want it. Hopefully both will be possible.

[cocoscacao]

38 minutes ago, schlosrat said:

True, but the important point seemed to be that there can be an equation, and the issue of the variable naming convention seemed like a distraction from that point.

Just an observation, nothing I take into account for further discussion 

[Periple]

13 minutes ago, Superfluous J said:

Personally laying out zones and letting the Kerbals build what they want sounds awesome to me. So does placing each building myself, making sure it's exactly how I want it. Hopefully both will be possible.

I think SimCity zoning-based gameplay would be a big departure from KSP-style gameplay. I’m assuming that colonies will be built much like rockets, by placing and connecting up components.

I’d also assume that resource transfers including heat/power/cooling will be pretty abstract because that’s how they are for rockets. That still leaves open questions about how specifically it works — for example, is there a radius around a power plant where buildings will be powered or do they have to be connected, if the latter do you need to draw power lines and so on.

Who knows, maybe they have some completely different ideas and we can always speculate but until then I’m not expecting drastic departures from the way things work with rockets. Basically I expect a colony will be much like a stationary spaceship but with some specific parts and a few mechanical twists.

[PDCWolf]

9 hours ago, Superfluous J said:

But it's not. The part must contact the water, so the base must be built near the water or there must be heat piping between the base and the water cooler.

Why would you want to have proper surface area based cooling on one (pyramid example), and magic "anywhere near water" heat conduction in the other? That'd be unintuitive and obviously a horrible user experience when you try to take what you learn from one part into another. If the part is intended as a heatsink, and you have half of it out of the water, you're dissipating half of the flux to air, not water.

That's the problem with oversimplifying everything. They're not making it easier, they're making it dumber, and creating these types of conflicts.

 

[mcdjfp]

I have a quick question about this thermal system that isn't dealt with in the 4 examples. 

I have 2 parts attached to each other.

Part A: Large, has a 1000 K temperature tolerance and produces a positive heat flux. 

Part B: Small, has a 500K temperate tolerance and is essentially static with no heat flux.

Which of these 2 parts will blow up first?

[Scarecrow71]

2 hours ago, The Aziz said:

Weirdly enough, setting up the basics for a working, safe base sounds to me like part of space exploration. I get that you don't want colonies at all, in any form because anything will be too much of a city builder for you.

Then maybe it's not the game for you and that's perfectly fine.

Colony management is NOT exploring space.  Getting to the location, scouting resources, dropping the first building?  That is exploration.  All the micromanagement is not.  Big difference there.

[The Aziz]

Transferring a crew between docked ships and fuel reserves in tanks could count as micromanagement with that mindset.

[Pthigrivi]

3 hours ago, Scarecrow71 said:

But that's just SimCity.  You put down areas for population (residential, commercial, emergency), place your power/water stations, create lines to run from them to your zoned areas.  And that is NOT what we should be getting out of a SPACE EXPLORATION game.  I mean, if that's what you want, fine and dandy, and you are more than entitled to play however you want.  But this just seems to me like the developers are giving us a city management game with rocket building being an afterthought.

No, its not. It's just one area. There's just one zone. You don't have to lay down roads, or pipes, or power poles. Inside the buildable area you could put things wherever you like, however you like. Everything within that area would automatically behave as if it was connected, sharing power, fuel, and heat management. You're not doing any of the things you do in SimCity: managing traffic, budgets, complex services, or even manually connecting infrastructure. You are doing all the things we already do in KSP: assembling working components that provide and consume resources like power, heat, and fuel in any way you desire. It's the same as a vessel, it just happens to be sitting on the ground.

Edited July 23, 2023 by Pthigrivi

[Superfluous J]

1 hour ago, PDCWolf said:

Why would you want to have proper surface area based cooling on one (pyramid example), and magic "anywhere near water" heat conduction in the other? That'd be unintuitive and obviously a horrible user experience when you try to take what you learn from one part into another. If the part is intended as a heatsink, and you have half of it out of the water, you're dissipating half of the flux to air, not water.

That's the problem with oversimplifying everything. They're not making it easier, they're making it dumber, and creating these types of conflicts.

We'll just have to agree to disagree.

[Periple]

1 hour ago, mcdjfp said:

I have a quick question about this thermal system that isn't dealt with in the 4 examples. 

I have 2 parts attached to each other.

Part A: Large, has a 1000 K temperature tolerance and produces a positive heat flux. 

Part B: Small, has a 500K temperate tolerance and is essentially static with no heat flux.

Which of these 2 parts will blow up first?

There is no thermal interaction between the two parts, their temperatures are independent of each other -- the hotter part won't transfer heat to the cooler part. This means that:

1. The 500K part won't ever blow up -- no flux, no temperature change (and it's clearly below 500K at the start of the experiment since it's still a part).
2. The 1000K part will heat up. As it heats up, the negative flux will increase with the temperature until it equals the positive flux and the part reaches equilibrium temperature, or the part exceeds 1000K and blows up.

 

1 hour ago, Scarecrow71 said:

Colony management is NOT exploring space.  Getting to the location, scouting resources, dropping the first building?  That is exploration.  All the micromanagement is not.  Big difference there.

I think building a base will be much like building a rocket, except with different engineering challenges. Building a spacecraft requires a lot of fine-tuning if you want to make it efficient or are doing something complicated with it, and I hope base-building will have opportunities for similar gameplay.

I do hope that the bases won't require continuous interventions to keep them running -- if they're well-designed and well-built I ought to be able to leave them alone and just enjoy using the resources they produce to build more spaceships!

Edited July 23, 2023 by Periple

[mcdjfp]

11 minutes ago, Periple said:

There is no thermal interaction between the two parts, their temperatures are independent of each other -- the hotter part won't transfer heat to the cooler part. This means that:

1. The 500K part won't ever blow up -- no flux, no temperature change (and it's clearly below 500K at the start of the experiment since it's still a part).
2. The 1000K part will heat up. As it heats up, the negative flux will increase with the temperature until it equals the positive flux and the part reaches equilibrium temperature, or the part exceeds 1000K and blows up.

That is how I interpret the situation based on the Developer Insight. I am just hoping I am wrong because this behavior completely fails the authentic test for me. 

[Periple]

19 minutes ago, mcdjfp said:

That is how I interpret the situation based on the Developer Insight. I am just hoping I am wrong because this behavior completely fails the authentic test for me. 

I don't think you're wrong.

The OP describes the system and also why they made this simplification: there are very few gameplay situations where the more accurate simulation would make a noticeable difference, as in most situations the fluxes from the environment are much higher than conductivity between parts, while the simplification is much less expensive and allows them to simulate thermals on craft that aren't in the physics bubble as well as craft with very large part counts. It also makes it easier to design/balance parts as you'll have one fewer parameter to tune since you won't need to worry about how good thermal conductors or insulators they are.

As I said earlier, in my view it's very similar to the way KSP1 and 2 treat aerodynamics -- the drag cube/occlusion based model isn't realistic, but it's predictable and close enough to the real thing that design and gameplay are meaningful. 

Edit: or, for that matter, patched conics as opposed to n-body physics. I think it's very much in the same spirit as the other physics systems in KSP -- a close-enough approximation to feel realistic while being fun!

Edited July 23, 2023 by Periple

[The Aziz]

25 minutes ago, Periple said:

 

I do hope that the bases won't require continuous interventions to keep them running -- if they're well-designed and well-built I ought to be able to leave them alone and just enjoy using the resources they produce to build more spaceships!

Pretty much that. Any sort of management will happen when, or rather if, the player decides to build/expand the colony, be it every month, year, decade, or never. The gameplay loop around colonies is most likely to look like this, based on everything I've learned about them over the last 4 years:

1. Build

2. Make sure it has everything it needs considering all available gameplay aspects (just like you would while building a rocket, so temperature, fuel, power, crew)

3. Set up delivery routes if you want to do so

4. Forget

5. Come back only when you want to do something more to it

6. Forget again

[Book]

Crafting a user story to see what kind of interesting problems the thermals system ça create in ship construction.  

>Building a nuclear rocket.
>This time I put enough radiators for the engine heat
>The rocket works great !
>Takes the rocket sunward since it has so much dV
>Boom when I speed up time
>Fridge! Forfot the sun heat
>f9
>Carefully manages to survive the mission by doing several burns that don't heat up the engine past boom point 
>Genius    

 

No idea if all that would work in game, specifically the last part, but I think it goes to show the spirit of thermal management and how it won't just be "plonk radiators until cold" because players make mistakes, and that's where the fun is.

Edited July 23, 2023 by Book

[mcdjfp]

I quite frequently ended up with uneven heat distribution (careful, often muti-pass aerobraking as one example, high speed planes for another) when playing KSP 1 and I was hoping for improvements in this area, not a removal. The ability to place low temperature parts directly on the hottest part of the craft with no consequences feels like a bad lesson to teach.  

I was also hoping that some of this stuff could be properly moved into their own threads to better make use of modern multi-core CPUs. Seeing only 1 core under any kind of strain makes me sad.