Incarnation of Chaos

Gosh, I wish this would be stickied somewhere conspicuous in massive bold red text. You actually get it.

Is this done on the CPU or GPU? If the latter, what's the implications for AMD GPU which generally struggle with tessellation in games? Also, does this also mean that it might be possible to use a "tiled" approach to rendering KSP2 in the far future? Between this and the way your team is handling extrasolar systems with their own floating origins, there's a lot of potential for very interesting things to increase performance. Amazing writeup though, this is the stuff that I want to see!

The only one who has shown a lack of understanding here so far I'm afraid is you. KSP uses a 1/10'th scale system, because it's a game. Therefore if the engines produced "Realistic" thrust with "Realistic" mass ratios, then the game's balance would be thrown out of wack. Don't believe me? Go install S.M.U.R.F.F on a install and see how much more DV you'd get, since it also adjusts engines for higher TWR's. It's insane in a stock system until you're at 5M and beyond parts where those kind of TWR's are required. The Rocket Equation is the rocket equation I'm also afraid, and it doesn't matter how you feel. The math works out the same way regardless. You demand Realism. but ask for entirely unrealistic things as a result. Ion drives shouldn't work in atmosphere, except for the tiniest model planes (Which is what BARELY manages to fly on earth with Ion propulsion). People aren't leaving "Uninformed" comments, they're not disinterested in improving the game. In fact it's quite the opposite, they're critiquing your arguments because guess what? That's how you determine what suggestions will be actual improvements, if you took every suggestion at face value KSP wouldn't even resemble the current game. And honestly, i find the tone you've taken rather remarkable. If all of this was this much of an issue for you, then it wouldn't be too hard to roll some Module Manager patches to do it for you. And nobody would fault you for it, it's your game. Who cares how you play it? There's even an entire section of the forum where you could ask for help making them, but instead you came here and assumed you were right and everyone else who might disagree is wrong.

It's mostly because of wafer supply, which is booked 2+ years in advance. But yeah, I'm running a Ryzen 1700 with 2X Vega 56 and perfectly content. It handles everything i can throw at it from compile jobs to rendering, and even gaming. I could get ~30% more performance with a 3700 or similar, but then i'm still waiting for a decent GPU. And all the latest just reek of stopgaps, even the 3000 series still has terrible RTX performance and not that much better raster D:

F=MA is a thing isn't it? Rocket equation isn't that, but in the end all it's doing is taking into account the change in mass from the depleted fuel by using the natural log.... This is why Hydrolox has fantastic ISP, but poor thrust. It has such low mass per particle that even though it can be accelerated to decent speeds, the force produced can't approach heavier hydrocarbons. KSP's fuel is fine, the tanks are too heavy. SMURFF would make OP happy in all likelyhood, Sorry, iv'e been watching this for a while and just had to pop in.

Oh how the mighty have fallen.... I remember not too long ago a time where Intel was unbeatable, AMD teetering on the brink of bankruptcy. And Faildozer a massive embarrassment, and now Intel seems to be in the same shoe. As delicious as seeing this is, i do hope Intel eventually gets their house in order. These are For-Profit Companies, not our friends. So the only way to prevent AMD from being Intel 2.0 is for them to fight, brutally and constantly for our money.

High frame rate and low resolution stress CPU, so looking at benchmarks of games tested at 1080p and with the processors in question will be fine. Also most games batter a single thread, multithreaded games are the exception not the rule. Also I seriously doubt DDR5 is going to be available at anything resembling a reasonable price for consumers in 2022. When DDR4 was first released a 16gb kit was well over 330 USD and it was 2133 with crap timings. But you do you, my recommendation would be to plan around hardware you know exists now.

Let's just replicate Balkanor Cosmodrome, you'll be able to set Azimuths for days with that launch mount.

You can do the same thing by just setting a controlled environment to conduct your tests in, so basically pick whatever section of KSP2 ends up being most demanding and standardize a run through it. This is what most good reviewers and review sites (Should) be doing with any game. Also KSP2 is coming out in 2022, so it's unlikely that CPU's especially will be OOS....actually what CPU are you thinking of that's even close to a grand?!? Nah; he has a legitimate request. It's just a in-game canned benchmark is still not the way to go about it. If you want a canned benchmark to see this....look at physics scores in Time Spy or FireStrike ultimate for the processor you're thinking about. That'll give you the BEST CASE performance gains, and then realize that whatever you'll see will be decreased from there. Make note of the RAM Speed paired with them as well, both Intel and AMD Ryzen benefit from RAM => 3200mhz with decent timings.

To be fair though, it doesn't burn the fuel. It just uses the heat of the reactor core to expand it, but if you want a "hybrid" NTR there's a fantastic mod chock full of them either way xD

Then remove the heavy and useless blimp and replace with more rocket fuel for max efficiency. Oops, we just built a proper rocket it seems.

Pretty sure this was a confirmed no.

Built in benchmarks rarely reflect even the least demanding scenes in the game, especially after DLC and updates. So nah, you'll have plenty of benchmarks of the actual game to look at and judge if your machine is capable when KSP2 comes out.

Time warp until the data is used up in the lab, transmit. Then feed it the experiments you don't want to return, repeat until you only have the experiments you want to return to kerbin left over. Then return those to kerbin.

Ah, alright. Then review and save what you want to return to kerbin, send the rest to the lab. Sorry for the misunderstanding.

Don't store those experiments as Data in the lab in the first place, recover them with a Kerbal and land them. Once it's Data in the Lab, it can't be transformed back into the initial experiment as far as i know.

KSP2 is using the same physics as KSP, so there's no reason for them to make them stable in N-body. As far as we know, the only place where KSP2 diverges from KSP1's physics is Rask && Rusk which might implement some form of limited N-body. Does that mean they still couldn't test the systems and make sure they're stable in a N-body physics system? Absolutely not, and considering how much they've been touting "Improved modding" they might've considered doing so to make the lives of whatever people end up developing Principia II easier. But, the answer is we don't know. I'd lean more on the side of No, because they're already delayed. Things like this would've hit the chopping block floor almost instantly.

Even if you had them, the reason why they don't go full power immediately is because the compression stage has to reach the proper RPM. At least IRL, in KSP this is just emulated by a spool time. My suggestion? Just engage your brakes on the runway for the few seconds it takes to spool and then let it rip.

And I did though, look at my previous post where I literally said that I believed you wanted a separate game. But you missed the overall point, which was that even assuming that the changes are minor ( let's say making FO:NV from Fallout 3.) There's absolutely no way to actually know the results will actually be a savings of time and money in the end. And you cannot just handwave the possibility it becomes an absolute boondoggle instead of a success. Because first, and I can't believe I'm saying this. KSP 2 isn't even out yet, this technology you're thinking of leveraging isn't even complete let alone finalized and stable. And second, well K^2s above post summarized it pretty well. Your proposal is something that is easy to say, but the moment you actually get into the weeds of the implementation it rapidly becomes anything but.

You're making the assumption that they have the people ready and waiting to build a full N-body physics system from scratch, make it perform well and they wouldn't run into any issues while doing so. That's a very, very dangerous assumption to make. Especially in software. Bethesda wasn't even able to use the same engine with "minimal time and resources" as you claim. The construction kit for FO3 while outwardly similar to the one for Oblivion is massively different under the hood. Ditto for skyrim vs special edition, and they even admitted that the work required to bring it up to modern standards for TES VI was exponentially more than they anticipated. This is why bethesda has 6 year development cycles, because of the sheer amount of work and labor required to keep the underlying engine current. Also FO4 was practically sacrificed at the altar for what eventually became F076. Much of the lack of direction story wise, and the lack of even franchise staples such as skill points (lmfao) came from them having to scramble to get it into a shippable state in around a year after separating it from the multiplayer component. All of this isn't an airtight case against the idea, but it should serve as a very stern warning about assuming how easy or what resources wouldn't be affected by a project.

But back to the topic you do bring up a fair point in this, i keep forgetting KSP2 will have multiplayer because the feature doesn't mean anything to me. But if someone is able to make arbitrary code that can run elevated (Or escalate itself to Admin/Root), then you have the potential for people to unleash all manner of nasty things on others. Heck, might even be able to make a botnet with the infected machines.... ...Which would probably have GPU's.... Yeah, sandboxing it is then.

You literally had someone offer you a custom engine solution/semi-custom for this hypothetical game, conditional on getting a team of experienced individuals willing to develop it. Why does Private Division have to make it? Have you seen how much money people have been throwing at Star Citizen? If you truly believe there's a demand for such a game, then instead of asking a existing series of studios with their own obligations/deadlines and funding conditions/milestones maybe instead you should be looking at what it would take to develop a proof of concept. A minimum viable product, the people required, and then throw out a kickstarter or w/e after enough progress has been made to demonstrate the features. Or just make a bunch of pretty screenshots and promises and get the money anyway, because it's legitimately hilarious how easy it is to legally scam people out of their money because it's considered a "Donation". (Pls don't do this) Either you realize along the way there's a pretty good reason this hypothetical game hasn't been developed, or you secure a niche in a already niche market of games and become a very wealthy individual if everything works out. Because i don't see these people being against the concept, they're against saturating a studio already burdened with having to push out a project and support it with yet another more complex one. That likely won't have the same interest, and likely could be easily approached in it's scope and capabilities with mods in KSP2. Plus, many seem to think you're asking for these features in KSP2. Which from your post it was fairly clear to me you wanted a separate game, but by the same development team. But maybe some clarification in the title would help smooth over that, idk. I find it highly unlikely they'd give two flips about another independent studio making a similar game to what you've described. Since it wouldn't affect the amount of resources available to KSP2, nor KSP2's features itself.

*Rubs eyes* Is that....a Quadruple Integral? And i thought Triple Integration was a pain. Back to the topic at hand though, would it be possible to null out the majority of an object's mass (Say 98-99% of it). And then accelerate it, gaining momentum the entire way. At which point it would have so much energy just from the velocity component, that it could form a singularity/black hole if you used the warp field to increase it's mass instead at that point. Basically, my thought goes like this. We know Wormholes can do what we want, and that the quantities of Negative Energy needed for a full Warp Drive might be downright unobtainable. So instead of making the entire ship warp capable, why not use the same techniques on a much smaller scale to induce local tears in Space-Time with something akin to a relativistic missile. With the only issue being that you can't really control the endpoint (That's a rather large one mind you). And if i remember correctly, you would still need some energy input to stabilize the singularity. But because you only want it around just long enough to traverse it, i figured an intense pulse of radiation (A laser or similar) might be able to do it.

Yep, this is what people really need to realize. KSP2 looks good, but it's no Cyberpunk or Crysis 3 graphically. It's going to hammer a single thread, because the developers haven't gone away from rigid-body arrays. They are optimizing around it, so the performance will likely be much better than KSP. But still, Ryzen 3000/5000 or Intel i5/i7 8000 series and up is really what you want. Ideally ofc, you could probably get away with something like a i3 OC'd to the limit (I forget which one was Overclockable tbh). But why do that except for the lels

It's in scope, but in order of priorities it's so far down the queue that it's effectively approaching 0 importance Think about it, before tides we'd want Wind, Clouds, Weather, oh and N-Body because you'd have to have an actual simulation of gravitation at this point. Then you might as well implement a more accurate approximation of Aerodynamics al la FAR, and so on. That's what he means by feature creep, this isn't a feature that would come on it's own. It's packaged with a whole assortment of other things that would be required to make it convincing, and many of them the community would be heavily split on as well. So why spend development time on it? In the best case you get a interesting gimmick that if you REALLY wanted for a specific world you could probably hack your way around it, while isolating a large part of your customer base. In the worst case it becomes a massive parasitic drain on development time, resources, and takes away from those people working on more important features or squashing bugs, all while the timeline keeps getting kicked back and back until years later it's realized they can't deliver and have to scrap it and clean up the mess.