I agree. Though the reason I think resources are interesting, is because they serve as a basis for everything else.
While oxygen, water and food are important, it's not something that people will have a shortage off most of the time.

Metals (any kind really) serve as a basis for so many commodities as well as ships that they make a great thing to trade in. I suppose Josh will have certain commodities and split those up into components. That way you get an economy that's based on supply and demand, while using real fabrication.

But then you start getting into a procedurally generated tech tree. That could be a nightmare in and of itself.

However, if you end up just using the same tech tree and using a different skin, it kind of defeats the purpose of procedurally generated building materials and most of this thread.

The logistics of it alone seem quite frightening to say the least.

Simpler system that still retains procedural content: keep crafting and basic materials as what they are, but allow planet specific/unique resources to be generated. (If you've ever played Escape Velocity, you know what a unique resource is).

Proposed system:

Each procedurally generated system should have some probability (medium-low) of being a supplier for some unique resource related to the planet type (continental planet - Grummox ("a delectable species of rodent"), lava planet - Becullite Ore ("ultra-hard ore for ship armor plating") ... you get what I mean). Both the name and the description is procedurally generated.

Prices for the unique resource increase as a function of distance from the planet (neighbors will buy it at a slightly higher price, distant worlds will buy it at a much higher price)

With some lower probability, generate a demand for that particular unique resource on another world. This is not affected by distance (it can be literally next door, or halfway across the galaxy). The selling price would be astronomical here. Of course, you won't know what world it is... or guaranteed to have any world which demands it.

Taking into account all that, supply and demand works as normal. (Though the astronomical world might consume the resource significantly faster than normal...)

This keeps crafting and trading KISS, while encouraging mapping of worlds to find superbly awesome trade routes. Best of both worlds?

Will this create a demand for super seeds with two of these worlds trading the same resources literally 1 jump apart? Possibly. But the same thing happens in minecraft - the infamous emeralds/diamonds in villager chests thingy. It's your decision to use those seeds ... or not.

I like the idea of unique commodities, but at the same time I don't...

I would like them if they would only be known in a limited space around the source, perhaps slowly spreading out to sectors with a very limited demand.
Other commodities should be limited to only a few sectors, because there would be similar other commodities that feed the need of those people.

Bringing me back to need... I believe that if factions need some attitudes to define where it's good to trade certain items.
They'd have a "need"-, "boredom"- and "contentment"- factor. The need would for instance be, entertainment. Boredom would stand for how easy this faction gets bored with something. The contentment factor is how fast this culture grows discontent. Suppose a faction with the need for entertainment has a high boredom and low contentment factor, this would mean that this culture will jump on everything that gives entertainment but lose interest quickly. To stay happy this culture will seek after more and more new ways of entertainment.
Translated to economics, this would mean that this culture has a high export value of high tech entertainment, but is also a good trading opportunity when you come across entertainment commodities unknown to them.

I love the thought of this idea... heck... I love the idea of a procedurally generated tech tree that changes from game to game based on what materials are available for certain technological breakthroughs... buuuut... I tried playing "Beyond Protocol" back in the day... and worry that a system with procedurally generated building materials has the chance of loading a player down with "junk" building materials that either aren't very useful, just don't mix well or don't meet the needs of the tech tree.

Love the idea, fear the difficulty of implementing the idea in a fun or meaningful way.

P.S. any "starter" ships would need to be "built" out of those procedurally generated building materials (using in game logic) or it ruins the whole idea of the system.

P.P.S. Finding a new material that is better in some way than a previous material can make you dance with joy. Just saying.

Preface: I am a noob on here, but this seems really cool and I just thought I would drop my two cents in the bucket and hopefully some of this might be helpful.. this was supposed to be a simple response but then I got on a roll sorry.. that being said the point of all this is to have a system complex and variable enough to satisfy the more intense players but hidden enough that the casual player would not get daunted by needing to find some rare material to build a necessary component and systems that do that are hard to describe

Why not make the materials standard from seed to seed or only slightly varying *clarified below* so that the learning curve is easier. However, let scarcity be the main procedural element and let the npcs determine the value of said materials based on a simulation of economics. (i.e. you start off the game and the price of iron is 5 currency/unit and then you need some later and luckily the npcs have mined *Note at the end about mining* a lot of iron so the price has dropped to 2 currency/unit, later you go to buy iron again and the price it 20 currency/unit because no one has found any iron since the last time you tried to buy it).. I don't think it would be to tough (relatively speaking - I mean this coding etc. is crazy to me) to have venders raise the price of a good every time you or an npc sell to them and raise the price when you or an npc buys while having mercantile npcs programed to watch for price discrepancies and try to buy low and sell high somewhere else.

I think all things necessary for living in space should potentially be commodities. Someone above mentioned that water should not be a commodity.. really? do you know what it would be like on your spaceship if you ran out of water (your reactor leaked or your life support was cheap and in efficient).. not fun.. you would have roughly three days to find some or you would die.. Now say your seed universe does not have much water, sounds to me like you will be paying 1000 currency/unit. You think that is stupid, I remember when I realized that my buddy paid more by volume for bottled water then the price of gas for his car.

*clarifying above comment*
I think while your standard while your standard elements should stay the same, crafted materials (like alloys for stronger hulls and compounds for fuels) could be procedural and refinable i.e. you hire some geek npc to analyze a samples of carbon and iron to find a better steel than your previous steel. And this could be done repeatedly to diminishing returns. Then in turn the components created from this refined alloy could be improved *(note about tech tree below, sorry I am ranting, I am just letting myself connect dots as I write)* Then the material stats would just be simplified things that changed the stats of the components that they could be used in i.e. copper would have stats important to electrical components density and electrical conductivity while steel would have stats important to hulls and other structural components like density, strength(for kinetic damage resistance), melting point (for thermal damage), reflectivity (for beam damage), atomic stability (for radiation damage) and durability (for health)

*Note about tech tree*
If you make refinable materials then you can decrease the complexity of the tech tree while making game play richer. I mean that if you could have one discoverable tech that could be crafted out of a variety of material that would determine the quality of said material. your hull tech branch could have 'self repairing ablative thermal panels' as one of the discoverable ship hull components and when you went to craft it you would need a certain amount of a ceramic, a robotics component, and a structural component

*Note about mining*
I just watch ed the March update and thought it made sense that you would mine unrefined ores (ie iron ore) that would need to be refined into materials i.e. iron

bach wrote: Why not make the materials standard from seed to seed or only slightly varying *clarified below* so that the learning curve is easier.

...
I think all things necessary for living in space should potentially be commodities. Someone above mentioned that water should not be a commodity.. really? do you know what it would be like on your spaceship if you ran out of water (your reactor leaked or your life support was cheap and in efficient).. not fun.. you would have roughly three days to find some or you would die.. Now say your seed universe does not have much water, sounds to me like you will be paying 1000 currency/unit. You think that is stupid, I remember when I realized that my buddy paid more by volume for bottled water then the price of gas for his car.

*clarifying above comment*
I think while your standard while your standard elements should stay the same, crafted materials (like alloys for stronger hulls and compounds for fuels) could be procedural and refinable i.e. you hire some geek npc to analyze a samples of carbon and iron to find a better steel than your previous steel. And this could be done repeatedly to diminishing returns. Then in turn the components created from this refined alloy could be improved *(note about tech tree below, sorry I am ranting, I am just letting myself connect dots as I write)* Then the material stats would just be simplified things that changed the stats of the components that they could be used in i.e. copper would have stats important to electrical components density and electrical conductivity while steel would have stats important to hulls and other structural components like density, strength(for kinetic damage resistance), melting point (for thermal damage), reflectivity (for beam damage), atomic stability (for radiation damage) and durability (for health)

*Note about tech tree*
If you make refinable materials then you can decrease the complexity of the tech tree while making game play richer. I mean that if you could have one discoverable tech that could be crafted out of a variety of material that would determine the quality of said material. your hull tech branch could have 'self repairing ablative thermal panels' as one of the discoverable ship hull components and when you went to craft it you would need a certain amount of a ceramic, a robotics component, and a structural component

*Note about mining*
I just watch ed the March update and thought it made sense that you would mine unrefined ores (ie iron ore) that would need to be refined into materials i.e. iron

Nice post, and welcome to the forums!

I agree that certain ores and liquids should be associated with certain processes no matter the seed. You talk about alloys in the tech tree and I agree that the quality of the materials would increase if formed into an alloy. I haven't heard much about the tech tree in LT and not a lot about crafting so I'm unsure how much of this we will actually see in the final product.

Minor issue, iron refined becomes steel.

Why is that an issue? Steel is an alloy of iron and carbon which sometimes includes trace amounts of other elements.

BFett wrote:
Minor issue, iron refined becomes steel.

[/quote]

Refined iron is more pure iron.

Steel is iron+carbon

My mistake.

so, i had an idea for PCG'ing elementary material and mixed compound characteristics.

the characteristics of a material are numericals that describe how the material influences devices built using it or how the material behaves when in your cargo bay.

those characteristics are constant for pure materials, logically.

but what happens when the material is less pure?
the fraction of the material in the batch gets lower, which alters the characteristics of the material.
this fraction is somewhere between 0% and 100% (0.0 to 1.0) and i'll call the fraction simply x as of now, because its easier to type

the characteristics change along some simple formula f(x), this is something comparatively simple with few factors.
something along the lines of 1/(x+1) or e^-(x) or anything of similar complexity.

every material has independent formulas for each characteristic, so there may be independent f(x) for density, structural strenght, whichever characteristics materials end up having

for mixed compounds and alloys the functions of the individual ingredients get added together (or multiplied together, im not sure what would give the best gameplay)


example: material A has the function 1/(x+1) and material B has the function e^-(x).

we want to alloy it for the minimal value of the sum function (because its useful for our application and because its the more interesting example for the formulas i chose :V)

so the complete formula for the aloys behaviour is 1/(A+1) + e^-(B)

using 100% of component A (and 0% of B) gives a value of 1.366
using 100% of B gives 1.5, thats worse, so the value may be somewhere inbetween.

50% A and 50% B gives a value of 1.27, thats better

a bit more messing around gives a value of 1.26 for 37% A and 63% B

with more components it gets ever more complex and can give better results, but it also leads to more complexity when trying to design ones dream alloy.
which also gets more interesting the more materials one has available to mix into the alloy...


just some thought that recently appeared in my head as i was bored

I really like this idea, but can see why it could go badly. I too played Beyond Protocol and it was probably the worst game I have ever played while still being one I wanted to love. The main problem I see with the idea is the potential for it to become way too complicated for anybody to feasibly use given the UI and be too hard to learn to justify learning. In Beyond Protocol, it took a really long time to research anything- and even once one got some basic materials it was too slow to trial and error one's way to get something usable. It made failing early and often impossible rendering any material engineering super tedious and frustrating. On the other hand, many of Beyond Protocol's problems could have been solved by lowering the opportunity cost of learning the systems and improving the UI- so I am hesitant to say the actual implementation of the material engineering mechanic was bad outright. It was just hard to learn the system. Most of the time people complain about something being too complicated is when it is nigh impossible to learn, which is why Battle Cruiser 3000 gets so much crap but Dwarf Fortress doesn't. BC300 is buggy, unpredictable, difficult, and DENSE. DF is reliable in its behavior for the most part, but is still really hard, sometimes buggy, and has a terrible UI. DF is infinitely more popular, though, because it is well documented and so is way easier to learn and has a community that provides external software to make the UI clearer. I am convinced that if we improve the UI to make the information required to use a mechanic easy to understand and make the mechanic itself easy to learn, it doesn't matter how complicated the mechanic is and "complicated" becomes "deep".

With that in mind, if this mechanic were to be implemented it would have to make experimenting with different materials easy and fast, as well as make all information about materials developed clear as day. Beyond Protocol suffered because it hid important information- it was very hard to tell what a material was good for or even if it was good for anything in the first place. I don't think that this mechanic needs to suffer from being too complex- as long as the information is presented in an easy to digest format and the inventory is handled intelligently I don't think it would be a huge problem. The AI should get some pretty heavily optimized materials anyway, so I don't think every player would be forced to mess with the mechanic if they don't want to- it just provides another way to customize.

Having hundreds of different materials laying around is solved by working with materials like blueprints. We could have a long list of elements providing different properties, and the basic metals traded could be made from those. When you want to make a new metal, you use an element from the list, but instead of having to go and find 20 units of Zenon just to try out stuff, you would be allowed to generate any metal out of currently known elements in a simulation. That way trying out new combinations is both free and fast making it easy to customize metals for a specific job. When done, you save that metal to your library, label it for later, and then you go and run around trying to obtain large amounts of the materials required for building that component on a large scale, as you would with any other component. This lets you have a relatively deep material engineering mechanic but also have that depth easy to explore and learn about- without making it too easy to actually go about harnessing the mechanic for gain. Trading the new metal becomes as easy as listing the manufacturing chips for that metal on the market. This prevents the normal commodity market getting flooded with random AI generated metals, and new metals are treated just like the components and manufacturing chips already being listed. Inventories are kept clutter free because the player handles what metals they want to save in their library, and they only have to carry around the stuff for what they are actually building as they don't have to worry about carrying everything all the time just to experiment.

Combine that base with stuff like what cornflakes just suggested and you get a mechanic that is nuanced and interesting while retaining a low opportunity cost to experiment: all without cluttering everyone's inventory with a bunch of crap.

To make cornflake's idea work with this, you let the player determine the required purity level of the metal when you are designing it. You design that into the metal itself so that you know what to expect, and don't try to make a component and have it behave unexpectedly because RNG said your metals are less pure than normal. The higher the purity, the more materials are required to guarantee that the purity level is being reached, naturally making purer, more reliable metals more expensive. The lower the purity of the metal, the more random property changes happen to a component made from that metal during manufacturing. Perhaps we make the method of designing metals use the method cornflakes suggested as well, as it seems solid and intuitive.

Of course, this whole mechanic boils down to being an extra step in the supply chain in making components and the only way we can justify adding in an extra step is by having it provide some benefit beyond just adding more things to trade. Changing the material in a component would have to meaningfully affect how that component's behaviour. I fear the research system already handles anything we might want to affect, and adding in an extra system with different rules to solve the same problem the research system is supposed to solve (how will the player improve and customise things?) only shows that the research system isn't versatile enough to handle everything it needs to. While a special set of rules specifically for material subcomponents is realistic and cool, I am not sure if we couldn't get the same benefit by improving the reach of the research and tech systems. That gets us the same result without the player learning a new set of rules.

I do like deep crafting systems. The more options the more fun.

Sorry for wall of text and I hope this makes sense- it is late over here and I am tired. Thoughts? I also apologise in advance for any proofing mistakes, I didn't go over this very thoroughly.