Building

I like the idea of more building materials. Actually, I like the idea of vastly more (orders of magnitude more) "stuff" in the game in general, so that we can have enough diversity in items to build a reasonable in-game economy around. Resource gathering, crafting intermediate and final products, and trade are all huge aspects of successful MMOs. We have resource gathering, but crafting intermediates is little more than a trivial annoyance, and trade is a joke at best. To this end, I believe that the simplification of raw food items is exactly the wrong way to move, but that is starting to get off onto too much of a tangent.

I also like Ian's general outline of the different properties of materials, and I think it's a good starting point. Other factors could eventually be worked in, like the difficulty in machining titanium could be reflected in longer time to construct places from ingots. I might also consider the manufacture of alloys.

But before the strength of materials can be meaningful for anything other than armor/HP, this game would need an improvement to the structural integrity mechanics. The one that is place now, besides only working for BAs, is a total joke. It doesn't even remotely approximate anything realistic. A big problem, however, is that physics simulation is computationally expensive, and not easy to do over the internet in real time. And all signs that I have seen so far suggest that the devs on this game (despite the numerous nice things they have done) have very little physics knowledge.

So, tl;dr, I like the idea of more building materials, but I don't have any confidence that the devs can pull it off in a way that would be meaningful and satisfying to me.

 

You're right that the physics isn't that realistic. I don't think it needs to be though, in order to just be "better". Games should be fun, but also influenced by reality. They don't need to make the density and effect of alloys 100% accurate. In real life, steel alloys actually contain only a small percentage of manganese, but in a game like this, to say hardened steel is 50% iron and 50% manganese/vanadium/sathium/whateverium is OK. Most people playing don't want to do math; hell I have a math degree and I don't want to do math while playing Empyrion, I want to build funky bases and blow stuff up. So to distill it down to simple concepts like "an alloy is half iron and half other stuff" is a dumbing down that I think we can accept for playability.

The benefits of other materials, as I proposed, are basically:

• Density - matters only for ships, at the moment.
• Hit points - matters for stuff you take into combat.
• Availability - different ores may be more/less common on different worlds, also some are expensive (plain steel requires only iron, whereas alloys require sathium or potentially other metals)

You've proposed basing structural integrity on it - I am 100% on board with that. It doesn't have to be "accurate" though. Merely that the number of contiguous blocks of concrete that can overhang without support should be less than steel, and harder alloys should have even greater structural integrity. That would help.

Better would be if there was a limit to how HIGH you could build something before it would fall over - again this is made better by using tougher materials. So a skyscraper could be higher if it had a frame of steel, than if it was wood or pure concrete.

Improving that would be possible without implementing realistic physics, basically by just changing the decrease in SI across different types of blocks depending on the material.

Ships are tougher, at least in space the issue is acceleration (due to thrust or gravity) and higher acceleration puts more strain on things. You don't need completely realistic physics to make something that would sort of work, there's plenty of ideas out there and I'm sure they've thought about it since it is on the roadmap.

 

In general, I agree with what you're saying. Perfect realism/simulation is not the goal for most games. Tynan Sylvester (of Rimworld fame) wrote an interesting post related to that.

So yes, I think it would be silly to have to worry about different alloy composition down to a fraction of a percent. I'd maybe like to see carbon be added to iron to make steel, but I'm more or less ok with putting iron into the constructor and getting steel plates out. And sure, titanium is often used in alloys containing a little aluminum, but I'm fine just calling titanium titanium. Making every alloy from a 50/50 mix of ingots I think would be a bit crude, so having some other "easy" ratios like 1:2 or 2:3, etc. would be nice.

You do make points about density and availability that I agree with. Those could be used in a more meaningful way now, especially for ships. Right now you basically choose between building a light ship with few HP, or a heavy ship with more HP. Using xeno steel, IIRC, was kind of like aluminum, but it was impractical to obtain in significant quantities. I think that is a step in the right direction. The biggest problem that I see is that there is not a big enough penalty to mass and the resulting flight characteristics to make anyone think twice about building a ship out of solid (or mostly) combat steel. Just slap on a few more RCS and thrusters, and you're good to go.

I disagree with the idea that structural integrity can be adequately modeled by something as simple as maximum overhang distance, even for a BA. Sure, you could improve what we have right now by varying that number for different materials, but it would still be incredibly crude and ugly and hard for me to like. If you think of a beam, the thicker it is, the more load it can carry. Of course the actual calculations are much more complicated than that, but even the simplified concept is not represented.

And you are right, ships are tougher. The physics is exactly the same, of course, but you can use simplifications for static structures that just won't fly for dynamic ships. The 'overhang' simplification is a perfect example; if you are only modeling a fixed structure's resistance to gravity, it's obvious what an overhang is. But if the direction of force applied can change, then what? And you already mentioned thrust/acceleration, which suddenly changes everything from a static analysis to a dynamic one, which is entirely a different can of worms. Dynamic simulation over a laggy network is a hard problem to solve; just look at Space Engineers as an example. Clanging rotors and pistons, anyone?

I would like to see enough simulation, however, so that you might be able to build a CV that has enough structural integrity to survive in space, but no withstand gravity sitting on a planet surface. You could potentially build a super-heavy combat CV out of cheap materials that is slow and unable to land. But the way people build ships right now usually is so much thrust that acceleration is much larger than gravity, so maybe this is unrealistic to expect.

Maximum stable height of something is also a pretty tricky thing to calculate realistically, but would be nice to see approximated to some extent in game (as you suggest). I would like to at least have some modeling of the differences in tensile strength vs compression as part of that.

Something else I'd like to see the effect of landing a ship on a structure. That also brings the complication of dynamic modeling to what were previously static structures, but could probably be approximated at least to some extent. When you land that 75m combat steel CV on the roof of your wooden starter shack by the lake, you should get toothpicks. Lots of toothpicks.

 

That's starting to sound an awful lot like Space Engineers. It's a very accurate and realistic game, but I find it extremely frustrating and un-fun to play. I don't want Space Engineers. I already have it, and I don't play it. I want something to relax with, have fun, and play.

 

There are a handful of things that SE takes into account that EGS doesn't, like mass and volume of cargo affecting overall vehicle mass and performance. But for the most part, it isn't even remotely realistic. It does do a little bit with damage in collisions (in a superficial way), and when grids break apart they become separate grids. But that's really about it. For a game that started out claiming realistic physics, it is a dreadful failure IMO.

I played SE for a long time, and enjoyed many aspects of it, but the things that ultimately drove me away from that game were mostly related to the complete and total performance nightmare that it was in multiplayer, along with the constant desync problems - especially the situations where a little glitch resulted in all your stuff blowing up.

Meanwhile, I started playing EGS. Even though it has even weaker physics simulation, similar multiplayer problems (especially desync issues), and flat rectangular planets, it doesn't let those things get in the way of it being a (for the most part) playable game. SE became unfun for me because it fell way short in what it was attempting to simulate and became unplayable in the process.

The blog post I linked earlier from Tynan explains to an extent how too much simulation can be a problem. I also saw an article recently from a bit of the opposite point of view that was discussing how driving games have evolved, and the serious amount of physics simulation that goes into some of those. I wish I had the link handy, but now I can't seem to find it. But I do believe that there room for more realistic simulation in a fun game in the style of SE or EGS (as opposed to KSP).

Anyhow, I'm not especially thrilled about either game at the moment, but I'm still here offering suggestions and rooting for EGS to evolve into something great. I wrote off SE a while ago.

 

No, it can't be accurately modeled, but I think there's things they can do, using the system that they have right now, to "fake" it in a way that is more realistic and brings more meaning to materials.

First, let's look at the system they have right now. It presumably works something like this:

• Each block that touches the ground has 100% SI.
• Each block that is stacked directly on top of another block will inherit the SI of the block below it.
• Each block that is not stacked on another block will have its SI decreased by some amount based on the distance to the closest block that has 100% SI. (There are several possible algorithms to do this and I don't know what they did but it is not important for this discussion.)

Here are some ways that can be improved that do not require any type of realistic physics simulation:

• The rate that the SI decreases as it gets away from a support should vary based on the material. This can vary based on material strength and weight. The SI should decrease more across 4 unsupported concrete blocks than across 4 unsupported combat steel blocks.
• Adding blocks on top of an unsupported block should decrease the SI of both blocks due to the added weight. (I've seen evidence it might already do this but am not sure if it is really working or just a bug.)
• Adding a small pillar should not give 100% SI to the stuff above it, it should only grant something like 80% SI. To get 100% SI you need a solid block.
• The decrease in SI needs to be proportional to the force applied. (In this case, the force is gravity.) You should be able to make structures larger on lower gravity worlds.
• If SI is exceeded, building should not collapse right away. Instead you should get a warning that SI has been exceeded, building starts making groaning noises, and damage starts being applied to blocks. You have time to remove offending blocks or shore up the structure with pillars, but if you don't deal with it soon building will eventually collapse.

This still does not solve the problem of building a thin tower that is 200 blocks high. In reality it would fall over. Why? Because wind can apply forces in a horizontal direction. So you do this:

• Come up with a number for wind strength. This will be much less than the gravitational force. It could vary by planet and weather or you could just pick a fixed number which we'll consider the "max wind strength" for this planet, and if you exceed it your building collapses.
• Existing SI calculation is for the Z axis. So we're going to calculate it for the X and Y axes now. Building collapses if SI values drop to 0 along any axis.
• For X and Y SI calculation, blocks that touch the ground have 100% SI. Blocks going upward lose SI according to the normal formula where it decreases by distance from the 100% SI block.

Now, if you build too tall the building can fall over. This is not accurate physics but would work better than what we have now. Materials now matter, gravity now matters, height/wind now matter.

How do you apply that to ships?

• When thruster is engaged, calculate SI along that axis, such that the direction of the thruster's force is "down"
• Blocks that have a thruster directly pushing on them have 100% SI
• SI decreases with distance from block supported directly by a thruster using a formula similar to SI for bases
• Exceeding SI of ship will damage structure much like I described for bases, it doesn't immediately crumble but you'll get a warning and ship takes damage while excessive force is applied. If you ignore that ship can break apart.

Now you can build large ships as long as you put multiple thrusters in different areas to "support" moving the structure.

How to deal with gravity? Well, it adds to the force applied in the opposite direction as thrust (so if you are flying upward the SI has to withstand gravity plus the thrust force). That's how a ship that can work in space might break up if you bring it into atmosphere.

I propose this because I believe it is a system that just extends what they already have done and can be implemented by programmers with basic algebra rather than requiring accurate physics using 3D vectors. It is just a game and it really is possible to fake a lot of things without making the math super-realistic. These types of rough approximations are also far quicker to compute than a real 3D physics simulation. Shortcuts like this will keep the game playable.

I also agree with your other comments about materials. I wish there were more alloys like titanium and aluminum but no one should worry about needing 10.2% carbon or whatever, dumbing things down to simple ratios is good enough. If they allow more substitute materials, then they can leave the more advanced materials to ship designers and the like, and people that don't can just make everything out of steel and the stuff will at least still fly.

I think the depth should be there though, because this is what keeps the game so interesting for so many different types of people.

 

The problem with Space Engineers isn't that it is "too realistic", it is that when the developers were given the choice between making something realistic, and making it fun, they chose realism. And this seemed to be their stated intent. But they sell it as a game, not as a simulator that universities and NASA can use to help do research. So I think they had their priorities shifted.

People don't want 100% realism, but the game is more immersive when it is closer to real experience. It is not believable that you can build an almost unlimited height tower on a planet and it will not fall over. It is more realistic that stronger materials or lighter materials could be used to build bigger structures. It is not that realistic that a thin wall has half as many hitpoints as a solid block. They can tweak this stuff to make it seem more real, without needing to introduce actual percentages of alloys or density of materials or dynamic force vectors to simulate wind resistance. Doing that sounds cool on the surface but would be complex to implement, a drag on performance, and unlikely to make the game more fun.

The game would be more realistic if you have a maximum weight for your inventory too. Problem: in reality you could not even carry one 2m block, even if it was made out of styrofoam because it is just too big, and a single solid block of steel that size would be over 60 metric tons. Would the game be fun if you really needed to drive a forklift or use a crane for every single block one at a time? No, EGS is a game and should stay a game, not a realistic construction simulator.

But having the weight affect acceleration of a ship is a place where it helps realism - a more massive ship should accelerate more slowly - while not negatively impacting the fun you can have playing the game.

I think improving SI, and applying it to ships, could be done in a way that makes it more realistic but still keeps the game fun. (And to some of us, building a ship that falls apart when you launch it because it was a poor design would potentially increase the fun of a survival-type game.)

 

I could never find the supposed realism in Space Engineers, building a smallish frigate with 2.5m thick walls and ceilings is not very realistic at all, it can even grow to 7.5m thick if you include conveyor systems. Not to mention that the entire game world is completely barren and devoid of life (apart from one single viable lifeform and another completely oddball one). I like Space Engineers for building complex and wacky machinery and all the logistics involved but, in the overall scope, Empyrion has practically rendered it obsolete, for me at least.

As for building in Empyrion, it would be nice to be able lift up a build and "levitate" it in mid-air like you can in From the Depths, it would make building an HV on a planet so much easier, only for creative of course. The custom weapons and engines of FtD would also be cool in Empyrion but, I think that's a bit too much to ask. FtD uses smaller voxel blocks so custom weapons may be too bulky in Empyrion.

 

Let me first start by saying that I do think your suggestions would probably be an improvement. I think they would add a little bit to the game, and they might be easy enough to be worth doing.

I would guess that it's probably something like that, yes.

This is where I start to have a bigger problem. Yes, you could do the things you describe, and yes I think it would make the game more interesting. I just don't think it's enough.

So here's the first thing I think needs to get straightened out: what the game calls structural integrity is not really SI. Consider a cantilever beam. In game, blocks farther from the support have less "SI". In reality, the blocks having the most stress, and the blocks that will break first, are the blocks closest to the supported end! The blocks hanging out in space have very little stress. The simulation isn't just over simplified, it's completely backwards!

Except maybe it's actually not. Maybe just the terminology is the problem. If you look at the current system, and instead of looking at the red blocks as having lower SI, you think of them as having higher deflection, things start to make more sense. Of course, high deflection by itself doesn't cause things to break, but for now let's just simplify and assume that high deflection in one block at least means that something is under stress, somewhere.

But then also consider your proposal that increased weight from stacking blocks should cause a decrease in SI. Let's think of a simple vertical column. The blocks at the bottom are under the most stress. But they have the least deflection. So if red means higher deflection, then you would show the "SI" as green on the bottom changing to red at the top.

Ok, I like the idea that wind could knock over buildings. But even without that, you still can't build a super thin, super tall tower. Or a super tall, thick tower. In the former case, you will probably exceed the critical load first (the tower starts to bend, and all hell breaks loose). In the latter, you crush the bottom and everything comes tumbling down. Probably it would be fine to ignore the first effect for now. The second is something that I would like to see.

Now for wind, things get interesting. You could do a very crude approximation of the static load by looking at the surface area of the building and square of the wind speed. And then, yes, you could approximate the horizontal deflection in the direction of the wind. Again, the blocks at the top would deflect the most, and stress would be applied to the structure at the bottom.

In reality, wind speed varies with altitude, and the shape of the building matters. Those things could be taken into account easily enough. But where things really get interesting is when you get into aeroelastic effects. That would be cool to see, but is more than even I think is reasonable to ask for in this game.

In general, I can go along with this, but with revisions to consider deflection vs SI. Of course, if you have a non-accelerating ship and you kick on the thrusters, it makes more sense to me to think of the maximum deflection as being the parts closest to the thrusters, but it's really all just relative. The biggest problem I see is that the thrust provided by different size thrusters is obviously quite different, and this will somehow need to be taken into account. Maybe it is as simple as a small thruster reduces displacement by only a fraction instead of turning the attached block back to full green. It really should account for multiple small thrusters in a row, also.

I like the idea of some sort of warning before breaking up. (Of course, right now, ships don't really "break up", and that's a whole separate issue to deal with. I guess you could disappear a section like the game does with BAs, but I really don't like that. Either way, what to do upon structural failure can wait until after we have decided how to determine that structural failure has occurred.) I think this should occur in at least two phases, however. And this could be applied to BAs as well. The first phase could be a warning and damage taken, as you suggest. But if you exceed the limit by way too much, then it should just be a big crunch/rip/etc.

Something else this brings up is the possibility of variable thrust. Right now, it's on/off, at least as far as the player can control directly. Maybe there should be a way to either directly control variable thrust, or maybe the ability to set a thrust limit for "on". The specific case i'm thinking of is a combat ship with lots of extra thrusters so you can still maneuver after having a few blown off. Firing them fully all at once might tend to break your ship. There are lots of possibilities for how to work this into the game, although the general feeling seems to be that anything less than near-instant max speed isn't fun, so.... maybe this would not be popular.

This is wrong. If you are flying in the atmosphere, and we are ignoring aerodynamic forces (for now, at least), then the force of gravity acts uniformly on the ship, and nothing will break from that. (At least for our purposes. The strength of gravity is a field that varies, but not by enough that we care in this game. And the same applies to BAs, for that matter.) The force that matters is the force from the thrusters, in whatever direction it is applied. I'd like to add aerodynamic forces as well, but let's skip that for now.

This all goes back to what I said previously about ships already sustaining large acceleration from thrust compared to gravity. If it can survive that thrust, it can survive gravity - at least if it is applied at the same points. That doesn't happen until the ship lands, however, and you don't land on your thrusters. Ok, so maybe you do, but maybe we make landing gear mandatory, and if you try to land on the hull or a thruster or turret or something else, you break things. But that probably isn't realistic, and the ground isn't usually flat, so... maybe not. Landing is really a problem to solve with dynamics, anyhow, but maybe as a first step the game can run a static analysis on your ship when it comes to rest.

I don't disagree that crude simulations will require less computation than more detailed ones, I just don't think that the ambition of the developers should be limited to what's easy. Of course the ROI on dev time has to be considered, but until someone shows what is possible, there is just a lot of guesswork in speculation in that. There are a lot of very clever ways to do things, and I would still really like to see something more realistic.

 

Yes, that's basically true for a small object, what I wasn't thinking of when I wrote that is that the reason gravity can break up a structure when it is on the ground is because the planet stops part of the structure (the part touching the ground) and not the rest of it.

The general philosophy behind a game like this is continuous improvement - to keep making it incrementally better. You can crudely simulate certain things, and then when issues are identified, you can come up with a better crude simulation.

It isn't impossible to accurately model it. The problem is that to accurately model it would make the game unplayable on current hardware. An accurate simulation would not be scaleable. If you start doing realistic physics on all the buildings and ships, a server would start having severe limitations on the number of structures in it.

The number of blocks in a structure, with the number of structures on a playfield, would make the performance terrible. You have to come up with an approximation, and if that approximation has things that seem unrealistic, then you kludge around them. This is just being practical. In order for the game to have the large scope it tries to have, it needs to take shortcuts in a lot of areas. A single-person shooter can be more accurate because the playfield is so much smaller. An MMO can potentially do it because it could scale computations across a cloud of servers. This game cannot accurately simulate the physics of its playfield on current generation hardware. It has to make compromises in many areas, and this isn't due to the limitations of the developers, it is simply due to the nature of the problem.

It could do better than it is doing, though.

 

@Hummel-o-War

Hopefully this is the right place to post this and get the devs' attention.

For reference: https://empyriononline.com/threads/...-release-candidate-2.38093/page-5#post-248178

The following videos would be an Empyrion Builder's dream. You can do it!





Please?

 

didn't they shut down that game?

 

I have no clue. Unless it was the micro-voxels that caused the shutdown (if there was one) it isn't relevant. I've never played that game.

Cloud Party was micro-voxel based but it wasn't shut down because of the micro-voxels.

 

I really like all these suggestions, it would be awesome to be able to see a capital ship break in half during a fight due to hits at the center that destroy the structural integrity .

The only thing I might want slightly different is the wind shear. The way you described checking the SI on the x and y axis is great but it seems like it would limit the total height fairly short for buildings even with better materials. There needs to be a way to support higher for example instead of a one block wall make the wall 2 or 3 thick would give more support or a 2x2 column, attaching a roof or beam connecting 2 towers of blocks adds rigidity allowing them to support more shear forces. I am not sure how to implement this in game in a simple way, but there needs to be some basics there for it. maybe its just a supported width vs height ratio that cannot exceed the wind force applied.

I would love the wind to not be an average number but have a calm weather number and a storm number for the various weather effect so something you build in calm might be iffy during a storm if not build well. It would be great to be in your nice all glass skyscraper and a storm comes and you start to hear creaking and see cracks in the glass etc.

Lastly when something does collapse it should not leave behind nothing, the collapsed blocks currently just disappear, I even get frustrated with this while taking apart things to reuse mats for an upgrade, or taking apart a poi. Please let the blocks fall to the ground or fall as rubble which can then be mined for at least part of those materials with either drill or multi tool. Having a building collapse is never fun, but having it collapse immediately with 1 block over when trying to build is the main reason most people build in creative mode and spawn blueprints, having the creaking and damage is a good way to fix this, having some resources salvageable when a collapse happens is another as you wont loose hundreds of even thousands of ignots in a half a second from one multitool miss click.

 

Wind shear: what you may have missed is that under this proposal SI is affected by the force. Wind shear is a force, and it is far less than the force of gravity (unless it is a tornado which can lift stuff off the ground precisely because it does exceed the force of gravity). So while you might be able to only stick out unsupported blocks horizontally by 10 blocks or so, going upward would be much easier, maybe it could go up 80 blocks. Depends on wind force.

I like the idea of wind shear being weather dependent but suspect it would frustrate people to build a base and have it get blown over the first windy day. That can actually happen in real life (see "Galloping Gertie") but some people might not care for it. Maybe if someday they have catastrophic weather (hurricane, tornado, flood) and other disasters (earthquakes, volcanoes, meteor strikes) it could add a cool element to the game but there's too many basic things it needs first.

 

I understood what you were saying about the force of the wind, just work around too many engineers not to use the term the structural engineers use all the time by default.

I guess my concern was even with low wind, say you can build out 5 squares and up 40, because the shear force is less, it still limits buildings to make 40 high, unless there is a way to add more support by making thicker columns, having side walls supported by being connected together to make a rigid frame etc. That would allow you to extend that distance and get a building that is 80 or 120 blocks high if built correctly with even a basic structural design in mind. Doesn't need anything crazy but a 10ft wide wall is much stronger vs those sheer forces then a 5ft wall, and 2 walls connected by a beam between them are stronger than either wall alone. Really just needs a ratio of width vs height.

As for the wind being able to knock over buildings it might piss off some people but people would learn quick to over engineer the structure part if you go with the minimum to make something stand then any little change can make it fall. I would enjoy finding where that risk vs reward spot is, so long as i didnt loose all the materials whenever something fell.

 

I don't really have anything to add beyond my previous post as far as the wind and SI discussion goes, but specifically wanted to call this out for a +1.

This is also related to your mention of splitting a ship in half that is destroyed in the middle, in that there is a general problem that should be addressed where one structure becomes broken. The difficulty in handling this realistically is that you suddenly have more objects to compute physics for. Space Engineers (at least the last time I played it, which was quite some time ago) doesn't really model SI very well either, but it DOES have grids break into multiple pieces when there is destruction. This is cool, but also causes substantial performance issues even with aggressive removal of small free-floating objects.

So, an easy way to avoid the performance problem of adding more bodies is to just make things go away. Or not break them apart in the first place (which is where we are still with ships in EGS). I've always hoped that this was a "just make it work for now, and we'll revamp it later" type of thing.

 

I can see the problem with performance impact of splitting off many parts even just splitting a large CV in half could be a major cause for lag, it would just make combat so much better if you could compromise the structure and end up silencing half an enemies guns because they were no longer attached to the core and power.

Maybe we will see something in the future, even if it is as basic as it has to be the size of a class 1 ship or more to separate otherwise that part just distengrates.

 

I currently dont see a matching corner block for the "Ramp Wedge Bottom" block, am I blind or is there no corresponding ramp corner for that specific block?


Also, We should have railings that can be put on the same block or a 1 block railing. Just a suggestion.

 

The craetivity of building is very limited ! There should be the same shapes of Windows as there are for blocks just without the mass of corse.
And the Basic blocks should be available in thin and half to! By Right klick the window offers 64 places, should not be a Problem to ad a second page to it like the texturing device has. Some of the Windows need visual fixes of there Corners, that all match togehter. And a few more rounded blocks to be able to built nice Towers. And at last, the rotating of blocks by 45° steps (of corse placable to), that would give us plenty room for creativity.
Thanks!