Jump to content

Nate Simpson at Space Creator Day talks about KSP 2.


RayneCloud

Recommended Posts

On 11/1/2023 at 9:58 AM, PDCWolf said:

If you don't want to get out of the forum, that's fine, but that'll never change the reality that the 1.5 million in the subreddit and the almost 500k between both hubs are more representative than the 20 to 50 users that keep posting in this forum.

Different channels, different metrics. These 20 to 50 users that keep posting here are trend setters, those opinions usually get echoed there. The inverse is also true so, nope, just Forum is not enough.

Something that I learnt over the years, Forum's heartbeat is kinda reflected on reddit et all (or vice versa, or even besides - I didn't managed to establish causality) and the number that matters the most is the anonymous readers ones. This Forums os a source of curated information, after all.

There are content creators, content replicators and content consumers. The metrics we need for the task at hands are the content consumers.

 

On 11/1/2023 at 12:02 PM, PDCWolf said:
On 11/1/2023 at 11:48 AM, LoSBoL said:

What is my view? Please share your...

...assumptions

And you think those assumptions (some of which aren't assumptions) are not grounded on reality. They are.

Ignorance ignores their own ignorance - the path to wisdom is not knowing about everything, but knowing what you don't know. This path is beyound most people, including on the Game Industry and is the very reason things are how they are.

We don't need to know exactly how many players KSP2 have, it's enough and sufficient to know if it's a growing or shrinking trend. And the knowledge we have, besides incomplete and imprecise, is good enough to the task at hands.

 

On 11/1/2023 at 11:48 AM, LoSBoL said:

You are filling in blank spots you know nothing about.

We are filling blank spots YOU know nothing about, something completely different.

As an example, I don't need to know that Gravity is exactly 9.80665 m/s2 (what's not, we used this number by definition) to know if something will break or bend if I let it fall from the table. Since my objetive is to have things over the table that will not break when falling, once I know the object's resistance I can use 10 m/s2 (or even 9, why not?) that it will serve me the same - it's a freaking table, not a suborbital trajectory.

Do you get it now?

 

23 hours ago, PDCWolf said:

The one reason this kind of stuff isn't headlights-to-the-eyes evident is because SQUAD was really afraid to put consequences for anything in the game. Normally, having a Kerbal EVA right after you re-entered a capsule means the little guy should melt. Having a nuke producing heat should degrade the quality of life in the cockpit as coolant is sent there unless you had extra. Living in a hellhole where you can't touch the walls should mean the science or resources produced should have a drop in quality or amount. Mounting 17 engines to a single plate should mean a piping hell that kills your performance with excessive mass. Flying your SSTO inside a fireball should vaporize the kerbals inside before the structure (or worse). But god forbid the game actually has any semblance of difficulty, and now in KSP2 any semblance of depth.

I couldn't agree more (but I'm still trying). :)

 

23 hours ago, Pthigrivi said:

This is a good example of why I'd actually prefer the game assumes <not important to my answer>

The problem is not the assumptions. The problem is the absolute lack of consequences.

As an example, I don't mind KSP¹ tendency to wobble on bad designs. As a matter of fact, I really like it (besides the wobbling reflecting a tree structure, instead of a more real life equivalent spars, spines and reinforcements). I don't mind even the auto-struts (except by that absolutely stupid bug happening since 1.2.2). What I mind is the lack of consequences on using auto-struts! They should add mass and cost to every part being auto-strutted, that would made things interesting.

I completely agree with @PDCWolf when he complains about how shallow the game is becoming. This game were intended to use STEM as an entertainment and learning tool, but right now is just a somewhat 3D Hollow Knight using engineering as scenario but without any engineer on the creative and development teams! Parts are nothing more than spells and powerups.

How do you plan to promote STEM without STEM professionals around? It's like trying to teach Math using teachers of Literature!

Alternatively, perhaps this is intentional - KSP2 is not a STEM entertainment and learning tool, just another glorified RPG or proto SHMUP intended to allow people to be condescending to themselves by pretending they are doing something "hard".

 

23 hours ago, Periple said:

(I've done this exercise, and the only difference I can find is the one I've been repeating: if you have radiators that only cool the part to which they're mounted, then you have to attach them directly to the part you want to cool, rather than an adjacent one. And, once more, if that really is the only difference, then in my view it is a pretty trivial one -- and well worth it, if the reduced computational cost allows simulating thermals for all craft and bases, whether they're focused or not.)

With all the due respect, I don't think you understand the game enough.

Ideally you should out the radiators on the part that most need it, but if your design doesn't allow it, you need to compromise.

One viable compromise is to attach parts with LOTS of thermal mass on the target part, and then shoving the radiators on these parts. Or in parts attached to these high thermal mass parts.

There're parts with better convection than others. There're parts with better conduction than others. There're parts with more thermal mass than others - worst, with variable thermal mass. And all of that parts have different skin and core heat capabilities. Making these work the way you want are the challenge, not merely attaching parts by aesthetics in a glorified sudoku pattern.

The whole reason to bother with all of this is, hell, to solve problems. You remove problems to be solved, you remove content from the game.

Edited by Lisias
Better phrasing
Link to comment
Share on other sites

Just now, Lisias said:

With all the due respect, I don't think you understand the game enough.

Okay? What am I missing?

Just now, Lisias said:

Ideally you should out the radiators on the part that most need it, but if your design doesn't allow it, you need to compromise.

This is the case for radiator panels, but not deployable radiators (TCSs).

The only reason I've come across where the design doesn't allow this is an artificial one -- the big ISRU and drills don't allow surface-mounting anything, so you have to attach them to an adjacent part as a workaround. I would expect that heat-producing parts in KSP2 are designed with the KSP2 thermal system in mind, and do allow mounting radiators.

2 minutes ago, Lisias said:

One viable compromise is to attach parts with LOTS of thermal mass on the target part, and then shoving the radiators on these parts. Or in parts attached to these high thermal mass parts.

Yes, I've done this with ISRUs and drills, as described above.

2 minutes ago, Lisias said:

There're parts with better convection than others. There're parts with better conduction than others. There're parts with more thermal mass than others - worst, with variable thermal mass. And all of that part have different skin and core heat capabilities. Making this work the way you want are the challenge, not merely attaching parts by aesthetics in a glorified sudoku pattern.

The whole reason to bother with all of this is, hell, to solve problems. You remove problems to be solved, you remove content from the game.

(Okay, the nitpick again -- convection isn't simulated in KSP. It doesn't do fluid dynamics.)

I have had and solved problems managing heat in KSP1.

The easy problem has been with drills and ISRUs -- the solution is just to spam enough radiators, either the deployable ones anywhere, or the panels on a fuel tank attached to the ISRU and drill. This will be affected by the change in KSP2 if you're using the panels; you'll have to attach them directly to the heat-generating part instead of the tank. It is in my opinion a trivial change.

The harder heat-management problems I've dealt with don't even have anything to do with radiators. They've been about surviving atmospheric entry. Eve, in particular, can be a tough nut to crack if you don't want to use heat shields. And I'm pretty certain that the KSP2 system will behave in practice as good as identically to the KSP1 system, because the heat flux directly on the parts from the atmosphere will be much, much higher than any heat fluxes from conduction between parts.

As to skin vs core heat, I think that's just an unnecessary complication which can be done away with without in any way affecting the outcome, at least within the constraints of the KSP1 simulation. You are simulating thermal mass, so just having a single threshold will work. You could create parts where it would make a difference, e.g. having a very high skin temperature threshold, a very low core temperature threshold, and a very low thermal mass; this would require somehow dissipating heat out of the part quickly, but I don't think there are any parts like that in the stock game. Or can you point to any?

 

Link to comment
Share on other sites

24 minutes ago, Lisias said:

We don't need to now exactly how many players KSP2 have, it's enough and sufficient to know if it's a growing or shrinking trend. And the knowledge we have, besides incomplete and imprecise, is good enough to the task at hands.

It depends at what task is at hand, and what you want to derive from that Playercount? You can't derive anything as to why the playercount is where it currently is at. Like I said before, I can name  at least 10 reasons as to why people that bought it aren't playing.  If you can't differntiate, you can't assess.

24 minutes ago, Lisias said:

We are filling blank spots YOU know nothing about, something completely different.

The sentence you quoted and are now reacting to is a reply to below;

Quote

Nah, you just can't find a sample that agrees with your view and thus are ready to think everyone else is just a groupthink. 

That does not compute, or are you able to think for me?  furthermore, I already gave you a reply to the first time you said that;

Quote

YOU don't know nothing. Don't project your ignorance over the rest of us. ;)

And the answer to that remains the same.

Edited by LoSBoL
Link to comment
Share on other sites

1 hour ago, Periple said:

Maybe? Or could it be that you're missing how unimportant it is? You did have a pretty fundamental misunderstanding about the system, if you believed that parts had been removed altogether and the entire craft is simulated as single part!

Actually, being individual parts that can't conduct to each other yet somehow can have their heat dissipated by select others is worse. That a hot thing will heat others around it is basic reality.

1 hour ago, Periple said:

As an aside, has correcting this fundamental misconception altered your view of the system at all? Because I do agree, if the entire craft was simulated as a single part, that would be a significant change for the worse!

Definitely, but it didn't make it any better.

1 hour ago, Periple said:

Exactly like in KSP1. If a part heats up past its limits, it will explode, by itself, without affecting anything else. I think we've all seen this happen when re-entering a bit too hot, as solar panels, antennas, goo containers etc. explode... or even the retrograde parts of the spacecraft itself. (It could cause a structural failure of course, if a component in the middle of the craft blows up, like when you're coming in too hot in a spaceplane.)

Even if that's true on re-entry, it's not true anywhere else. A part having lots of neighbors means its heat can dissipate before it's a problem, even if to other parts.

1 hour ago, Periple said:

Exactly like the deployable radiators in KSP1.

Contrary to KSP2, where every radiator is a magic radiator and parts don't conduct to neighbors.

1 hour ago, Periple said:

I think the skin/core system can be abstracted away with zero impact to anything. If I understood it correctly, it was a representation of a part's ability to soak heat flux, which can be represented as a single value (thermal capacity). 

It's also an abstraction of conductivity not being 100% efficient, thus a part could soak up flux, as its skin remained hot, and then after the skin cools, it'd slowly conduct to neighbors or the environment. Yes, it can be abstracted to a single value, so long as you include the efficiency factor for conductivity, which'd make no sense if all your radiators can magically conduct to themselves.

1 hour ago, Periple said:

Can you describe a vessel where this would actually happen? I can't think of any way to do this. Literally, any way, even if I tried. It's always the heat flux directly on a part that kills it.

Literally anything ISRU. Converters and drills were the biggest heat generators in the game. Of course they wouldn't explode parts, as they'd shut themselves down first with bad conductivity. Again, blame SQUAD for not wanting to include proper consequences for bad stuff. However, to see the effect at hand, you could make a craft that has drills and converters at the bottom, and static radiators at the top. You could then try putting the radiators closer to the drills and converters and see the difference as the heat transfer becomes more efficient and those parts are able to operate for longer.

53 minutes ago, Periple said:

I think my previous message was a bit long, so I'll try a tl;dr. @PDCWolf, can you describe, or show, an example of how the thermal system in KSP2 would change the way you would have to design a vessel, compared to KSP1? And by "an example," I mean an actual vessel you're designing, not describing the system in general terms. 

(I've done this exercise, and the only difference I can find is the one I've been repeating: if you have radiators that only cool the part to which they're mounted, then you have to attach them directly to the part you want to cool, rather than an adjacent one. And, once more, if that really is the only difference, then in my view it is a pretty trivial one -- and well worth it, if the reduced computational cost allows simulating thermals for all craft and bases, whether they're focused or not.)

Well, in the case of ISRU or NERVas you can't attach them directly, adding a bit of nuance to how you have to manage that, and giving a chance for the system to show its workings and consequences. In KSP2 it literally doesn't matter: radiators can suck heat from any part anywhere in the vessel, so where your heat generators and your radiators are has 0 influence on how you make your ship.

Also KSP1 would simulate heat (and ISRU) for off focus vessels, in a basic way. In fact there was a bug at some point where your whole base would save up heat and instantly explode when focused.

You also keep mentioning CPU impact, but I haven't seen any metric of it, so I'd appreciate a source. The last thing I remember is the heat overlay bug, but that was graphical.

41 minutes ago, Lisias said:

Something that I learnt over the years, Forum's heartbeat is kinda reflected on reddit et all (or vice versa, or even besides - I didn't managed to establish causality) and the number that matters the most is the anonymous readers ones. This Forums os a source of curated information, after all.

These days, not at all. Also, anonymous readers don't matter at all, given that the people we can gather an opinion from are not the anonymous readers, and them reading the forum doesn't mean they agree with anyone in particular. In fact, I'd bet most anonymous are here for mods.

Also, the reddit doesn't exist in a vacuum. The "cesspool" of hate pictured on Reddit is the same on Steam, on their twits, the youtube comments and so on. Only FS! announcement has changed that a bit, and a lot of that change is also by people giving up and going away instead of staying for 8 months to support or criticize.

 

Edited by PDCWolf
Link to comment
Share on other sites

11 minutes ago, PDCWolf said:

You could then try putting the radiators closer to the drills and converters and see the difference as the heat transfer becomes more efficient and those parts are able to operate for longer.

The you could increase time-warp, and forget about it all-together.

Link to comment
Share on other sites

Just now, cocoscacao said:

The you could increase time-warp, and forget about it all-together.

Lack of consequence is an issue that permeates every system, yet funnily enough heat does build up under timewarp for isru parts.

Now tell me, how has the perma-timewarp problem been addressed for KSP2? oh wait.

Link to comment
Share on other sites

22 minutes ago, PDCWolf said:

Contrary to KSP2, where every radiator is a magic radiator and parts don't conduct to neighbors.

How do you know that all radiators have implicit heat piping?

I just re-read Nertea’s article and it didn’t say that. On the contrary, he mentioned heat sinks and radiators on a part. 

22 minutes ago, PDCWolf said:

Actually, being individual parts that can't conduct to each other yet somehow can have their heat dissipated by select others is worse.

I take it you never use the deployable TCSs in KSP1, then?

22 minutes ago, PDCWolf said:

Even if that's true on re-entry, it's not true anywhere else. A part having lots of neighbors means its heat can dissipate before it's a problem, even if to other parts.

Anywhere else, such as…?

22 minutes ago, PDCWolf said:

Literally anything ISRU. Converters and drills were the biggest heat generators in the game. Of course they wouldn't explode parts

But the question was specifically about exploding parts! :joy:

22 minutes ago, PDCWolf said:

In KSP2 it literally doesn't matter: radiators can suck heat from any part anywhere in the vessel, so where your heat generators and your radiators are has 0 influence on how you make your ship.

Once again, how do you know this? Nertea didn’t say so, and nothing about removing conductivity between parts precludes having radiator panels that only cool the part to which they’re attached. Would you be happy if KSP2 provided such parts?

22 minutes ago, PDCWolf said:

You also keep mentioning CPU impact, but I haven't seen any metric of it, so I'd appreciate a source. The last thing I remember is the heat overlay bug, but that was graphical.

I’ll refer you to Nertea’s article again:

“We don’t have a lot of user stories that benefit from simulating conduction between parts, or intrinsic simulated thermal emission. A lot of the time this reduces to equilibrium in KSP1, particularly at high time warp factors. In addition, if we want this simulation to apply to vessels in the background, we need to simplify things. It’s hard enough computing thermodynamics for 500 parts on a vessel – imagine if we wanted to crunch fancy thermodynamics on 50000 parts on 100 vessels! This has to scale effectively and be performant.”

 

Edited by Periple
Link to comment
Share on other sites

41 minutes ago, Vl3d said:

Your do what?! That makes no sense! Bad TWR, very heavy.

On low gravity worlds, even the Mun, TWR isn't what matters. The sky cranes are reusable and dropping whole bases so its long-term fuel efficiency Im after.  This is niche in KSP1 but it sounds like we'll be doing a lot of similar things to get colonies established in KSP2.

Sky crane is in the background there next to the vac lander. 
TEHf0Ar.png

1 hour ago, Lisias said:

The problem is not the assumptions. The problem is the absolute lack of consequences.

In this case it is the assumptions. Just like with RCS the game assumes that coolant is being properly routed from the heat generating part to the radiators and isn't leaking into the parts next to it. Thats a completely reasonable assumption and opens up much more interesting design and engineering possibilities. You still have to match your loads to your radiators, it just makes it easier to understand, predict, and plan for so players have fewer confusing failures and can get deeper into more complex mission profiles. For instance in the image above the mass of the rover in the foreground is centered through both the docking port on the coupla/ cockpit and the docking port on its back. That allows you to push it into orbit and transfer it to Munar orbit using a tug from the front, then land it with the sky crane straddling its back. You couldn't do that if the engines needed to be attached directly to a huge fuel tank. 

Edited by Pthigrivi
Link to comment
Share on other sites

3 minutes ago, Periple said:

How do you know that all radiators have implicit heat piping?

I just re-read Nertea’s article and it didn’t say that. On the contrary, he mentioned heat sinks and radiators on a part.

Check the "complex" examples, the one below the 3 part ship orbiting kerbin specifically. The radiators deployed near the engine are used to cool the (i guess) reactor at the other side of the ship.

4 minutes ago, Periple said:

I take it you never use the deployable TCSs in KSP1, then?

I did ISRU a couple times until I learned to manage the heat and design good looking stuff, then like with literally everything else in the game I figured there's no point to it. One of my biggest gripes of KSP1 is that every gameplay "loop" is cut by simply asking "why". Why would I add ISRU to a Jool5 ship and bother with the extra designing and mass? for literally nothing other than saying I did.

Sadly the rest of the game doesn't really require much heat management (again, blame SQUAD for being afraid of consequences), and KSP2 will make that even lesser it seems.

7 minutes ago, Periple said:

Anywhere else, such as…?

ISRU, rockets with too many engines (like nukes), high speed atmospheric flying, flying near the Sun. Again, you can't blame the system for SQUAD not having made enough scenarios for it to be used. KSP2 will include those scenarios apparently but not the in depth system, which is really funny and sad.

8 minutes ago, Periple said:

But the question was specifically about exploding parts! :joy:

I don't get this, you want a situation where heat buildup would explode a part but it could be stopped by conducting to its neighbors? That's literally anything I mentioned above. If there wasn't an adjacent conduction between parts, those heat producing parts would just be isolated and build up heat till they explode. Again, the only ones that don't explode are ISRU and that's because they were programmed specifically to not explode.

11 minutes ago, Periple said:

Once again, how do you know this? Nertea didn’t say so, and nothing about removing conductivity between parts precludes having radiator panels that only cool the part to which they’re attached. Would you be happy if KSP2 provided such parts?

It's funny because every example he gave points to literally that: You have part with negative flux somewhere, it counteracts parts with positive flux anywhere else. Check the "complex" ship example, and the colony example:

image.png.bb15b9ce411aa66e692509d83aa26f

^Radiator on the bottom cools reactor at the top even though (if piping is not assumed) the thermal mass of the truss in the middle would be negligible.

image.png.e6f2a469db6b7af5995f4576b2d795

^Steam turbine looking thing and water piping cool down the whole base.

As for what I would be happy with: An expanded rather than reduced heat system, with actual consequences and that required some thought be put into ship design. Hopefully well tied into life support. But hey, they're clearly making a cartoon instead of a proper sequel.

12 minutes ago, Periple said:

I’ll refer you to Nertea’s article again:

“We don’t have a lot of user stories that benefit from simulating conduction between parts, or intrinsic simulated thermal emission. A lot of the time this reduces to equilibrium in KSP1, particularly at high time warp factors. In addition, if we want this simulation to apply to vessels in the background, we need to simplify things. It’s hard enough computing thermodynamics for 500 parts on a vessel – imagine if we wanted to crunch fancy thermodynamics on 50000 parts on 100 vessels! This has to scale effectively and be performant.”

This part is really funny, and also disingenuous for many reasons:

  1. That a hot thing makes things around it hot too is reality. If you think you "can't benefit from simulating that" then might as well not simulate anything, as you've thrown reality out the window. This is exactly what trivializes the whole heating system and reduces it to nothing more than soulless a mass tax.
  2. In KSP1 it reduces to equilibrium because it was safeguarded that way to not have parts explode on timewarp without warning. That safeguard is really dumb and trivializes the whole game.
  3. They need to simplify everything because they've build an overly bloated serialization system that needs to save every part of every vessel at every point of data they have. He talks about 50000 parts in 100 vessels when you can't even do that in the current game, let alone when heating and colonies arrive.

 

Link to comment
Share on other sites

1 minute ago, PDCWolf said:

heck the "complex" examples, the one below the 3 part ship orbiting kerbin specifically. The radiators deployed near the engine are used to cool the (i guess) reactor at the other side of the ship.

Yes, that demonstrates that a deployable radiator (TCS) analog part exists in KSP2. It does not demonstrate that a radiator panel analog part does not exist in KSP2.

3 minutes ago, PDCWolf said:

As for what I would be happy with: An expanded rather than reduced heat system, with actual consequences and that required some thought be put into ship design. Hopefully well tied into life support. But hey, they're clearly making a cartoon instead of a proper sequel.

A quite a few vocal people would like KSP2 to be more of a hardcore simulation than KSP1, yes. But that’s neither here nor there. The question we’re discussing here is whether KSP2’s thermal system is a “droolfest” compared to KSP1’s system.

I contend that the changes have minimal impact on design (radiator panel analogs need to be attached directly to the part that needs cooling) and significant impact on the overall game (loaded and unloaded craft/colonies are simulated the same way at all warp levels). In my view this is a worthwhile trade-off. 

I understand that you disagree, although I still don’t understand why you feel so strongly about a relatively minor design impact. I am also a little disappointed that you don’t appear to have shifted your position at all even after you realized that you had completely misunderstood the system by assuming that part-level simulation had been eliminated altogether.

Link to comment
Share on other sites

5 minutes ago, Periple said:

I understand that you disagree, although I still don’t understand why you feel so strongly about a relatively minor design impact. I am also a little disappointed that you don’t appear to have shifted your position at all even after you realized that you had completely misunderstood the system by assuming that part-level simulation had been eliminated altogether.

I am definitely just as confused. I was hoping to get an answer to this by the time I got to the end of this thread I am but still waiting.

Link to comment
Share on other sites

38 minutes ago, Periple said:

Yes, that demonstrates that a deployable radiator (TCS) analog part exists in KSP2.

It demonstrates that every part will act as the magic DTCS in KSP2. There are no examples provided of parts that don't magically sap flux from anywhere in the vessel.

38 minutes ago, Periple said:

I contend that the changes have minimal impact on design (radiator panel analogs need to be attached directly to the part that needs cooling) and significant impact on the overall game (loaded and unloaded craft/colonies are simulated the same way at all warp levels). In my view this is a worthwhile trade-off. 

I mean, for someone that completely overlooked how heat worked on the original, yeah, it's obvious to me you'd see it as minor. Instead of taking the good system from one, maybe making it better and actually having any mismanagement of it end up in proper consequences, they stripped it down to a mass tax and called it a day.

There's still no source for the "significant impact".

40 minutes ago, Periple said:

I understand that you disagree, although I still don’t understand why you feel so strongly about a relatively minor design impact. I am also a little disappointed that you don’t appear to have shifted your position at all even after you realized that you had completely misunderstood the system by assuming that part-level simulation had been eliminated altogether.

Having the whole vessel be a single total flux would kinda at least account for heat propagating and normalizing along the whole craft, as if it had internal cooling systems that move it around, even with the pitfall that it'd be instantaneous and magical. Now that you made me see that  every part will be its own little heat thing in isolation, without conducting anywhere, yet with magical heatsinks that can be anywhere on the vessel, actually makes it worse. 

Link to comment
Share on other sites

3 hours ago, Pthigrivi said:

In this case it is the assumptions. Just like with RCS the game assumes that coolant is being properly routed from the heat generating part to the radiators and isn't leaking into the parts next to it. Thats a completely reasonable assumption and opens up much more interesting design and engineering possibilities

Interesting designs? Perhaps. Engineering possibilities? Not at all - I'm afraid you have a dissonance between design and implementation.

It's exactly that plumbing routing what makes engineering very hard.

Ok, we can abstract part of that engineering challenges so players don't need a PhD on Aerospace to make something reach orbit, but too much abstraction and we have an RPG with spells and powerups instead of STEM.

Now, people need to take a decision and stick with it. EXACTLY what kind of game is KSP2? No wrong answer here, but it need to be properly answered because right now I'm seeing a trend of bait and switch on the marketing, positioning KSP2 as a STEM game but delivering a fantasy RPG with KSP¹ thematics and orbital mechanics… 

 

3 hours ago, Periple said:

This is the case for radiator panels, but not deployable radiators (TCSs).

As a matter of fact, you gave me an idea for a mod...

 

3 hours ago, Periple said:

(Okay, the nitpick again -- convection isn't simulated in KSP. It doesn't do fluid dynamics.)

Well, I'm calling them by the name I found in the Physics.cfg - granted, they may had misnamed it but, anyways, KSP¹ have something called convection on its guts. (link is for a 2015 post, but I just confirmed at least one of the attributes is still found on 1.12.5).

 

3 hours ago, Periple said:

I’ll refer you to Nertea’s article again:

<article>

And, yet, they got abyssal performance until recently. Not saying Nertea is wrong, but they may be doing shortcuts on the wrong places - sometimes, it worths to kill some FPS for a feature.

Quote

“The real problem is that programmers have spent far too much time worrying about efficiency in the wrong places and at the wrong times; premature optimization is the root of all evil (or at least most of it) in programming.”

 Donald Knuth

That said, something that I would love to have simulated is heat transfer on RESOURCES. I would really enjoy pumping cold fuel into hot tanks to cold them, and then pumping back into a cooler and so goes on. Having every part between the fuel tank and the cooler influencing and being influenced by such resource's temperature would, by itself, a feature that would surely make things interesting - we would have to manage heat in the same way we manage fuel.

 

3 hours ago, LoSBoL said:

And the answer to that remains the same.

Of course, since the question is the same. The failure on associating the answer to the question is yours, there's nothing we can do about!!

 

3 hours ago, PDCWolf said:

These days, not at all. Also, anonymous readers don't matter at all, given that the people we can gather an opinion from are not the anonymous readers, and them reading the forum doesn't mean they agree with anyone in particular. In fact, I'd bet most anonymous are here for mods.

You are not understanding. People are getting here to "get the opinion", and then spread it out there, not vice versa. And a good part of the anonymous readers are here to get support, as only the very basic and known problems are usually tackled down outside this Forum.

About most anonymous being here for the mods, it's yet another example of how Forum can be seen as a good heartbeat :) exactly by the reasons you gave: anonymous readers can't do anything but to read Forum, so if they are here is because they need to read something about the game, and so we have interest in the game. You can even infer how KSP is doing around the World by checking these numbers over the timezones - China appears to have a very significant user base… 

 

3 hours ago, PDCWolf said:

Also, the reddit doesn't exist in a vacuum. The "cesspool" of hate pictured on Reddit is the same on Steam, on their twits, the youtube comments and so on.

Being the reason I said Forum by itself is not enough. The hate pictured on lousily or not moderated at all channels is not reflected here because such hate is actively removed here.

However, I had noticed that things are changing on reddit - at least on the most populated reddit about KSP, I had noticed a lot of vitriol (including some old ones) being removed from the channel 2 days before the launch of the last patch. If such posts were being systematically removed, I would hadn't noticed it, but this happening specifically about 2 days before the last patch launch surely caught my attention.

Moderation on reddit is getting a bit overreached, as it appears - I wonder what will happen if Reddit fulfils the promise of allowing voting for demoting moderators there...

 

3 hours ago, PDCWolf said:

Only FS! announcement has changed that a bit, and a lot of that change is also by people giving up and going away instead of staying for 8 months to support or criticize.

Only one in twenty six customers complains, the others twenty five just goes away. Silently.

 

2 hours ago, Periple said:

A quite a few vocal people would like KSP2 to be more of a hardcore simulation than KSP1, yes. But that’s neither here nor there. 

Things would improve considerably if we would have a definitive answer for my question above: KSP2 is a fantasy RPG game with some orbital mechanics, or a de jure et de facto KSP¹ sucessor?

 

2 hours ago, Periple said:

I understand that you disagree, although I still don’t understand why you feel so strongly about a relatively minor design impact. I am also a little disappointed that you don’t appear to have shifted your position at all even after you realized that you had completely misunderstood the system by assuming that part-level simulation had been eliminated altogether.

Because it's not a minor detail on the whole system: managing heat is one of the very most important problems of space fare!

I'm disappointed that you are failing to address the importance of the second most important engineering problem on every space craft ever built by man, second only to the propulsion itself.

@PDCWolf had explicitly detailed exactly what worries him on the new thermal abstraction, explaining exactly what was being removed from the simulation and why it makes it shallow and meaningless, essentially removing it as an interesting challenge from the game. He's complaining that the new thermal system essentially does to heat what removing mass from the fuel would do to craft designs.

So, hey… Let's remove mass from the fuels too! Why bother with it? This would also "opens up much more interesting design and engineering possibilities". :/ 

Edited by Lisias
Tyops, as usulla...
Link to comment
Share on other sites

41 minutes ago, MechBFP said:

I think I'll take the word of the person who actually developed the most in-depth heat management system mod in KSP1 as to the value of such a system.

How many hours did you wait for the discussion to develop just to add a basic appeal to authority type fallacy?

Edited by PDCWolf
Link to comment
Share on other sites

2 hours ago, PDCWolf said:

How many hours did you wait for the discussion to develop just to add a basic appeal to authority type fallacy?

In regards to Chris Adderley I think it's more of an appeal to expertise. He's both seen KSP1's heat system, understood its flaws, and developed a system that was much more robust and in-depth. I've also used that mod pretty extensively and as fun as it is I wouldn't recommend it for stock. It took many, many hours of trial and error to understand how everything was supposed to work. But I have to understand how heating and cooling and solar gain and thermal mass work in buildings for my job so it was worth it and appealing to me. To draw an analogy it would be a bit like making engine design, engine cycles, turbopumps, and fuel ratios stock. Im sure a small subset of real rocket nerds would have fun with that but for post players who just want to snap on an engine and have it go it would be a huge source of confusion and frustration. The devs need to consider the entire player base and make judgement calls on how much is too much for most players, and how much emphasis to put on one system vs another so that the whole experience doesn't become an unwieldy slog. KSP already has a bit of a reputation for having an incredibly steep learning curve. Heat is a really cool problem in space so Im really happy to see them digging into it, but in doing so they need values and behaviors that are pretty clean and easy to understand and predict. If I'm generating x heat, then I need -x cooling to stay at equilibrium. If radiators are already assumed to have proper plumbing to and from heat sources there's no reason those sources should also then bleed heat into adjacent parts. It's just an opaque overcomplication. Worse than that its assuming the parts you're using aren't designed or functioning properly and then imposing flawed design constraints as a result. 

Edited by Pthigrivi
Link to comment
Share on other sites

Just now, Pthigrivi said:

In regards to Chris Adderley I think it's more of an appeal to expertise. He's both seen KSP1's heat system, understood its flaws, and developed a system that was much more robust and in-depth. I've also used that mod pretty extensively and as fun as it is I wouldn't recommend it for stock. It took many, many hours of trial and error to understand how everything was supposed to work. But I have to understand how heating and cooling and solar gain and thermal mass works in buildings for my job so it was appealing to me. To draw an analogy it would be a bit like making engine design, engine cycles, turbopumps, and fuel ratios stock. Im sure a small subset of real rocket nerds would have fun with that but for post players who just want to snap on an engine and have it go it would be a huge source of confusion and frustration. The devs need to consider the entire player base and make judgement calls on how much is too much for most players, and how much emphasis to put on one system vs another so that the whole experience doesn't become an unwieldy slog. KSP already has a bit of a reputation for having an incredibly steep learning curve. Heat is a really cool problem in space so Im really happy to see them digging into it, but in doing so they need values and behaviors that are pretty clean and easy to understand and predict. If I'm generating x heat, then I need -x cooling to stay at equilibrium. If radiators are already assumed to have proper plumbing to and from heat sources there's no reason to those sources should also then bleed heat into adjacent parts. It's just an opaque overcomplication. 

Oh no, don't take it as me questioning Chris' authority on the issue at hand. I wouldn't dare. I would dare, however, to point out that appealing to his authority as the ultimate decider is a fallacy, and I'm not even asking, but outright stating it.

In regards to the system itself, sure, it doesn't need to go as deep as to be a good tool for actual spacecraft planners, we're all on the same side on that. KSP1 does have a steep learning curve, but there's noting to do once you're at the top, neither in the form of other learning curves, or anywhere to apply what you learned (bar roleplay, or mods).

Link to comment
Share on other sites

57 minutes ago, PDCWolf said:

Oh no, don't take it as me questioning Chris' authority on the issue at hand. I wouldn't dare. I would dare, however, to point out that appealing to his authority as the ultimate decider is a fallacy, and I'm not even asking, but outright stating it.

In regards to the system itself, sure, it doesn't need to go as deep as to be a good tool for actual spacecraft planners, we're all on the same side on that. KSP1 does have a steep learning curve, but there's noting to do once you're at the top, neither in the form of other learning curves, or anywhere to apply what you learned (bar roleplay, or mods).

Its more like if I had cancer I'd weigh the opinion of an oncologist higher than I'd weigh the opinion of some random person off the street. Its not because of the title and the lab coat, its because they've spent their lives understanding and treating the disease that I have. 

On the latter I think thats kind of the point of colonies and interstellar, taking what we've learned making fairly simple flags and footprints missions and applying it to build a much more diverse set of vehicles to explore surfaces, deliver outpost modules, harvest and transport resources, etc. I've experienced that with modded KSP1 but not yet with KSP2. Every time a player goes into the VAB they should be tackling a really new problem in terms of design and engineering, flight profile, and overall mission architecture. We'll see if FS! expands on that. 

Link to comment
Share on other sites

19 minutes ago, Pthigrivi said:

Its more like if I had cancer I'd weigh the opinion of an oncologist higher than I'd weigh the opinion of some random person off the street. Its not because of the title and the lab coat, its because they've spent their lives understanding and treating the disease that I have. 

On the latter I think thats kind of the point of colonies and interstellar, taking what we've learned making fairly simple flags and footprints missions and applying it to build a much more diverse set of vehicles to explore surfaces, deliver outpost modules, harvest and transport resources, etc. I've experienced that with modded KSP1 but not yet with KSP2. Every time a player goes into the VAB they should be tackling a really new problem in terms of design and engineering, flight profile, and overall mission architecture. We'll see if FS! expands on that. 

Good thing this is not about cancer, good lord.

The problem with appeal to authority is that they failed to present an argument and instead just bring an authoritarian figure that shares their viewpoint into the conversation. The pitfall of appeal to authority is precisely that not all authorities might agree, thus the discussion turns into a pokemon card match, to see who's got the biggest or most famous authority, instead of discussing the matter at hand.

Even oncologists can have differing opinions.

Link to comment
Share on other sites

Another thing about Chris Adderly, out of all of us, he's the one who has the most hands-on experience with the new system.

It's really easy to find flaws in something but it's really hard to see the entire picture when all you're given is a small snapshot of a little piece.

We're the blind people trying to determine if we're encountering a tree, a snake, or a rope while Chris is sitting atop the elephant.

Link to comment
Share on other sites

10 minutes ago, Superfluous J said:

Another thing about Chris Adderly, out of all of us, he's the one who has the most hands-on experience with the new system.

It's really easy to find flaws in something but it's really hard to see the entire picture when all you're given is a small snapshot of a little piece.

We're the blind people trying to determine if we're encountering a tree, a snake, or a rope while Chris is sitting atop the elephant.

I also disagree pretty vehemently with Chris on LS and RP parts on a conceptual basis. But, y'know, Im just a guy playing. It's different from his experience and perspective. I can respect his accomplishments and disagree with his opinions. 

Link to comment
Share on other sites

2 hours ago, Pthigrivi said:

In regards to Chris Adderley I think it's more of an appeal to expertise. He's both seen KSP1's heat system, understood its flaws, and developed a system that was much more robust and in-depth. I've also used that mod pretty extensively and as fun as it is I wouldn't recommend it for stock. It took many, many hours of trial and error to understand how everything was supposed to work. But I have to understand how heating and cooling and solar gain and thermal mass work in buildings for my job so it was worth it and appealing to me. To draw an analogy it would be a bit like making engine design, engine cycles, turbopumps, and fuel ratios stock. Im sure a small subset of real rocket nerds would have fun with that but for post players who just want to snap on an engine and have it go it would be a huge source of confusion and frustration. The devs need to consider the entire player base and make judgement calls on how much is too much for most players, and how much emphasis to put on one system vs another so that the whole experience doesn't become an unwieldy slog. KSP already has a bit of a reputation for having an incredibly steep learning curve. Heat is a really cool problem in space so Im really happy to see them digging into it, but in doing so they need values and behaviors that are pretty clean and easy to understand and predict. If I'm generating x heat, then I need -x cooling to stay at equilibrium. If radiators are already assumed to have proper plumbing to and from heat sources there's no reason those sources should also then bleed heat into adjacent parts. It's just an opaque overcomplication. Worse than that its assuming the parts you're using aren't designed or functioning properly and then imposing flawed design constraints as a result. 

Yup exactly. This is similar to engines having limited number of ignitions, not being able to throttle engines from 0 to 100%, fuel ullage when in orbit, random failures, and many other things that while all being more realistic definitely aren't a good fit for accessible game play, especially in a game like KSP where the innate challenges of getting to space with a rocket are already pushing that envelope for new players.

Link to comment
Share on other sites

I am sympathetic to the wish that KSP2 would be a deeper and more detailed simulation rather than effectively the same game as KSP1, only broader and with multiplayer (eventually). I was a little disappointed that they didn’t choose to add life support, I think it could have had some very engaging gameplay consequences if done right. 

Ultimately though none of us are getting exactly the KSP successor we would want. It’s their vision, their game, and made within their material and human constraints. If you don’t like their choices, you can always mod — or make your own space game, like Juno!

 

Link to comment
Share on other sites

1 hour ago, Periple said:

I am sympathetic to the wish that KSP2 would be a deeper and more detailed simulation rather than effectively the same game as KSP1, only broader and with multiplayer (eventually). I was a little disappointed that they didn’t choose to add life support, I think it could have had some very engaging gameplay consequences if done right. 

Sending a kerbal on a years long mission around the sun with just a capsule, solar panels, a fuel tank and an engine is not realistic or engaging enough? (I hate not having something that makes me feel the difference in distance between a flight to low orbit and a flight to Kerbal orbit, it just feels incomplete, I can't play like this.)

Link to comment
Share on other sites

This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...