Spitball time: Procedural scientifically plausible plant life?

[NeilF]

Let assume this year sees atmospheric landings making it into the game say with the following sort of tools available to it:-

• Atmosphere: Clouds/vapour/fog
• Weather: Rain/Snow/Hail/Winds
• Surface: Rock/Ice/mud
• Liquids: Seas/Lakes/Rivers

Ultimately, what I expect a lot of explorers are after is diversity and variety, and that chance of finding something different/unique, and clearly if the Stellar Forge's information on each planet could be used to not only feed into the generation of the assets above, but also feed into procedurally generated plant life (where applicable) - that has some nod to science - then surely that could help give them that?

Now the reason I suggest plant life is it would seem far far far simpler than generating animal life. ie: Static plants built up on some procedural rules must surely be simpler than animals requiring no doubt far more complex rules for their generation, movement, behaviour, reactions and the food chain. So for the moment let's assume just plants are the goal...

So what do people see as the probability - from what they've seen in other games (eg: No Mans Sky) and applications (eg: Space Engine) - of this being possible in ED? What are the hurdles in doing this (eg: the engine potentially showing a huge mass of plant life from space, SRVs and ships reacting to hitting larger plants)? Do people even envisage it worth the effort? Might the assets/tech used in FDs other titles be of use?

 

[Vardaugas]

Well, let's see what a plant is made out of basically: Trunk roots, leaves, flower/fruit. This in a seasonal cycle. These can be rolled from tables individually, each with limiters to make them compliant with environmental conditions and a priority order that limits rolls on the rest to prevent silly combinations. Environmental conditions could be read out by biomes on the planet which also include data on gravity and atmospheric density. You could have a set of common species which can be found everywhere on the planet in variants and intersperse those with locally unique species for variety.

Most work would be to fill in the tables and design them in such a way that they come up with plausible plants reliably. An engine which can modify simple plant models and textures to increase visual variety without requiring more unique models to be made would be of advantage. Creating the models for the parts in such a way that they combine together seamlessly even when modified would probably also be a big challenge.

Procedurally generated plant life that looks unique for every planet requires a large amount of variables applied in layers, especially when you need to cut away variables with limiters to maintain plausibility. It's a rather complex system that would need a lot of planning and work to get right, but not completely impossible.

 

[Shiro Akai]

Bit of a side note but I wonder how our ships would react to those plants. Constant "Terrain unsuitable"? And you could just land on some barren rock, never be able to get close to admire? For it's totally true for a 700t ship (stock Clipper, for ex. sake) not to be able to land because there is a small rock or bush standing on the way.

 

[Bortas]

For it's totally true for a 700t ship (stock Clipper, for ex. sake) not to be able to land because there is a small rock or bush standing on the way.

We should definetely be able to blow our own clearings into the forests... a few trees should be nothing a Plasma Accelerator can't vaporize

 

[NeilF]

Well, let's see what a plant is made out of basically: Trunk roots, leaves, flower/fruit. This in a seasonal cycle. These can be rolled from tables individually, each with limiters to make them compliant with environmental conditions and a priority order that limits rolls on the rest to prevent silly combinations. Environmental conditions could be read out by biomes on the planet which also include data on gravity and atmospheric density. You could have a set of common species which can be found everywhere on the planet in variants and intersperse those with locally unique species for variety.

Most work would be to fill in the tables and design them in such a way that they come up with plausible plants reliably. An engine which can modify simple plant models and textures to increase visual variety without requiring more unique models to be made would be of advantage. Creating the models for the parts in such a way that they combine together seamlessly even when modified would probably also be a big challenge.

Procedurally generated plant life that looks unique for every planet requires a large amount of variables applied in layers, especially when you need to cut away variables with limiters to maintain plausibility. It's a rather complex system that would need a lot of planning and work to get right, but not completely impossible.

Indeed... Layers and layers of procedural generation rules based on the Stellar Forge data for the planet, and the resultant conditions (biomes) on the planet. Then with plenty of graphical mechanics/tools to give use to render the final product...

My head is telling me - like you suggest - as long as the thought and knowledge is put into it up front, is should be achievabe without rediculous effort?

 

[NeilF]

We should definetely be able to blow our own clearings into the forests... a few trees should be nothing a Plasma Accelerator can't vaporize

That's an interesting side issue... Making what ever is generated, interactable. ie: As you say being able to damaged/break/destroy plants...

 

[moorison]

There are many tools and algorithms for generating familiar and unfamiliar plant-like structures already, Speedtree being a middleware in some well known games already. But most games are set in an environment cultivated by humans, requiring only familiar plant designs. Check out L(indenmayer)-systems for example.
As far as I can tell, NMS goes for a (somewhat limited) alien but imaginative approach, not interested in the question if a plant in a volcanic environment should resemble one in a sub-zero one.

There is no chance Frontier can include every known and unknown aspect of expression of features in lifeforms with regards to the environment alone, even less so for realistic formation of species through expansion in different environments and removal of those that would not survive.
Only a tiny minority of players would appreciate the amount of work required approximating this "ideal" simulation, although it would be cool if the Devs add tools to reseach these things. Keep in mind that it is easy to get bored driving on planets even if they are unique because you have seen any (class of) variation possible in the engine at some time.

 

[Solsticide]

Bit of a side note but I wonder how our ships would react to those plants. Constant "Terrain unsuitable"? And you could just land on some barren rock, never be able to get close to admire? For it's totally true for a 700t ship (stock Clipper, for ex. sake) not to be able to land because there is a small rock or bush standing on the way.

We'll probably just land anyway and squash and destroy whatever is below us. We're humans, it's what we do.

 

[NeilF]

Keep in mind that it is easy to get bored driving on planets even if they are unique because you have seen any (class of) variation possible in the engine at some time.

Oh absolutely, but if it's simply another layer of variety and "feedback" to the experience. ie: Interesting looking plants, only found along river banks, next to a huge waterfall, in a ravine, with a huge electrical storm lighting up the scene...

 

[moorison]

Oh absolutely, but if it's simply another layer of variety and "feedback" to the experience. ie: Interesting looking plants, only found along river banks, next to a huge waterfall, in a ravine, with a huge electrical storm lighting up the scene...

Certainly. I have the feeling the water physics (not just a certain height around the whole planet - waterfalls etc.) might be more of a challenge than the plants.

 

[Vardaugas]

Certainly. I have the feeling the water physics (not just a certain height around the whole planet - waterfalls etc.) might be more of a challenge than the plants.

Hah. A proper surf seems to be an impossibility for many games still. Now imagine realistic tides depending on gravity of nearby objects, their position and orbital periods. Flowing water can be procedurally generated, at least on textures. Path of Exile added that to their engine. Then there's the question wether to use the linear wave function or the quantum mechanics one that can predict rogue waves.[haha]

 

[Micha]

To be blunt; the amount of effort required to make something -reasonable- would *far* outweigh the benefits for a *space* game.

People seriously underestimate the amount of effort required. Even something like NMS which has already been derided in this very thread as not even trying to be plausible took *years* to achieve what it has managed to create. NMS arguably has a smaller dev team, however, that is the *core* selling point for their game which explains the amount of effort they put into this aspect of it, to the detriment of other aspects (eg flight model).

In Elite, plants on planets would be a tiny niche, very briefly explored by the masses before they go back to their pew-pew or trading, and explored in depth (and complained about) by a few interested people.

And this is without even trying to get interactivity into it.

And finally I have to ask: what actual *gameplay* would this bring to *Elite*? In NMS, it's core gameplay to discover, tag, and name new animals and plants. Do you want the same in Elite? Or is it just for pretty backgrounds?

 

[Vardaugas]

It wouldn't just be developed for Elite. Fdev would create the engine prerequisites for such a system to use in future projects, too, if they are smart. They have their very own game engine, developing it would be an investment into the future.

 

[dannysquid]

Certainly. I have the feeling the water physics (not just a certain height around the whole planet - waterfalls etc.) might be more of a challenge than the plants.

Yes, when I've pondered rivers and stuff it seems like a huge headache. I'm not sure you can really be calculating river systems on the fly from space down to the ground. I think you'd almost have to start with the rivers and calculate the terrain from that.

 

[Rubbernuke]

FD have the geological considerations sorted, and it would be an interesting formula to see in action regarding planet age, star type etc impact flora.

Get some Aldiss action in there:

 

[B1rdy]

To be blunt; the amount of effort required to make something -reasonable- would *far* outweigh the benefits for a *space* game.

People seriously underestimate the amount of effort required. Even something like NMS which has already been derided in this very thread as not even trying to be plausible took *years* to achieve what it has managed to create. NMS arguably has a smaller dev team, however, that is the *core* selling point for their game which explains the amount of effort they put into this aspect of it, to the detriment of other aspects (eg flight model).

In Elite, plants on planets would be a tiny niche, very briefly explored by the masses before they go back to their pew-pew or trading, and explored in depth (and complained about) by a few interested people.

And this is without even trying to get interactivity into it.

And finally I have to ask: what actual *gameplay* would this bring to *Elite*? In NMS, it's core gameplay to discover, tag, and name new animals and plants. Do you want the same in Elite? Or is it just for pretty backgrounds?

This is my concern, too. No game ever applied procedurally generated content at the scope needed to fit into ED's game world. Even CIG relies on manual touch up on the planets in SC. Regarding the history of ED's development, I find it difficult to see, how they can archive that. I mean, compare that task to Multi Crew. And generating environments is just on thing, the other is the associated game mechanics. If they want to add involving mechanics, it is like creating a new game (even more actually). If they add some simple, cheap mechanics, it seems like a whole lot of effort to only create a nice background for that, like shooting rocks, but now in a Jungle.

 

[NeilF]

To be blunt; the amount of effort required to make something -reasonable- would *far* outweigh the benefits for a *space* game.

That's very vague, and therefore rather unfounded...

In Elite, plants on planets would be a tiny niche, very briefly explored by the masses before they go back to their pew-pew or trading, and explored in depth (and complained about) by a few interested people.

You realise a huge number of folks in ED are interested in exploration, which is quite a lot to do with... well exploration and seeing stuff

If we consider how popular DW2 is, does that put it into maybe some context?

And finally I have to ask: what actual *gameplay* would this bring to *Elite*? In NMS, it's core gameplay to discover, tag, and name new animals and plants. Do you want the same in Elite? Or is it just for pretty backgrounds?

What gameplay wise is there in rendering a nebula? Or the myriad of surface details/phenomena we already have?

Ultimately, some folks might like to find a rather unique looking plant, only found under extreme gravity on a very specific world types, etc etc...

 

[NeilF]

This is my concern, too. No game ever applied procedurally generated content at the scope needed to fit into ED's game world. Even CIG relies on manual touch up on the planets in SC. Regarding the history of ED's development, I find it difficult to see, how they can archive that. I mean, compare that task to Multi Crew. And generating environments is just on thing, the other is the associated game mechanics. If they want to add involving mechanics, it is like creating a new game (even more actually). If they add some simple, cheap mechanics, it seems like a whole lot of effort to only create a nice background for that, like shooting rocks, but now in a Jungle.

Agreed, but I cannot help think that a few very clever technical heads could produce some clever procedural generation fit for purpose...

 

[aRJay]

I don't think we should get too rigid in our definitions of plant and animal life, just because a life form is attached to one place doesn't make it a plant there are plenty of animals on Earth that don't move (and not just gamers) there are also so mobile plants.

Another consideration is that just because our ships can squash the life forms on a planet doesn't mean that they couldn't damage the ships in turn, we know our ships are vulnerable to corrosives for example and who knows what moulds could do to the electronics if they got inside.

 

[Latenoid]

Save the nature. You would destroy the planets... That's the reason why we cannot land....