Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[wesleyibruce]

Good. Detecting most terrain is long term. However when we drop bombs/ meteorites the holes  shape and size is dependent on the material. Differentiating road from dirt. I'll need to see that line of code. A non destructive version is a material detector but not fast.

Detecting water is very important for procedurally placed bridges etc. I've been diving into the water a lot around our causeway/ fords taking measurements.
If we can detect bridges on openstreetmap first then that would help.
If not a late bridge building pass might be an option though ideally it should be able to toggle off.
Doing it in world procedurally is slower. It may be fun to watch.
If we can detect the placement of a road at or just below water level with 1.5 m or more water either side (or road) then the location of a bridge segment could be chosen. Then others can be placed adjacent to the bridge if the water conditions each side is met. The bridge height could be inferred by the height of the first road tile with all adjacent tiles being dirt/ not water or bridge. Where the game places a causeway a meter or two above the water we could tag that and build a bridge too. It would turn some real world causeways into bridges though.

In all cases the road and raised causeway would need to be lowered to the average of the adjacent water river/ sea bed tile. I gather we have two basic bridges in the mods supply.

Two variations are possible:
Wharf, this is not a bridge because one road/ bridge tile has water tiles on all three sides with road/ bridge on the forth side. A wharf may have decorations on it to distinguish it from a bridge. One variant might be a boat ramp. The other a break wall.

Dam wall/ weir, Where water tiles are at different heights and the underwater ground/ river bed is also at two different heights. The dam wall would extend into the ground. Some decorative water flow at sluice gates. A variant would be water falls, paddy fields and farm dams. Water falls & Rapids may be possible if there is no road but the water slope is high enough. Is there a way to get the surface normal for a point? Levelling for a building placement and vehicle angle must be doing that.     

Two other variations would be possible with water and normal detection: more gently sloping beaches and sharp cliffs with overhangs. Some trees also will not grow on steep slope. 

There would be memory limits and graphics speed limits. We will need solutions there and that's being worked on I gather.

[fly77]

Lots of useful info...I'll certain try to implement something..for now I tried to actually use the "empty" to spawn some insects (flies) (courtesy andfly ) . Collisionless empties are rendered invisible by pressing a ctrl+z ...flies mod is spawned when approaching an empty < 15m. flies are removed from scene at distance >16 m.

The result is really cool ! (sorry for the bad video quality ) I can already feel how with this tool we can really start to create a lot of fun stuff and slowly slowly some small games ...thanks so much for sharing your deep insight into game design here !


The next thing I'll experiment with is to explore what I can do with some other animal mods of andfly and with my cows. Its a long time I was looking for a way to spawn them globally...let's see if it works out.

[wesleyibruce]

LOl deep insight. Don't let the real game designers hear they! They will lynch us both for being smart Aleks. It looks very good. Always check the frame rates. We need to keep them low and avoid heads exploding from over work in Slovakia. 

[fly77]

I was wondering how a procedural placement of "empties" would work.  In an immense space like outerra (the whole earth)
I suppose they would only be placed in some reasonable surrounding of the player, and this means for outerra that also "empties" should be spawned when needed and deleted when not needed.  In my understanding only in games with small maps ("game level" maps ?) they would be "permanent" allover the map. Right ?

[wesleyibruce]

Absolutely. but you could store and pass them to others if you had a reference latitude and longitude and a table listing an xy & z coordinate from that reference point, an orientation and a few other columns for traits.
One empty essentially triggers a group of others and allows you to build whole towns, puzzle mazes, etc, that can be placed and turned on and off and re set.

Thus we could have multiple client side but shareable games in the same big map. Removing them is essential otherwise we create a data problem for Cameni.

[wesleyibruce]

I just found your snow ball game and harvester.  These show existing engine capabilities. They are more advanced than one of Cameni statements implied.
We can do a survival game relatively easily. It would require Lots of modeling and a UI would be needed but I have bender, etc.
It would require: A UI ideally with 4 layers:
User info and world view. Day Z minimalism. health, hydration, food, comfort, held item, closest  item in interaction volume or 'what your looking at'.
Inventory and crafting. 4 by 10 grid for loot plus 5 crafting slots = three inputs, 1 out put, 1 byproduct. 3 buttons for other menu. [I already have a survival tech tree worked out.]
Character equipment and Stats/ skills screen.
Hints, setting, exit, and tutorial screen. All 4 screens would have a close button and buttons for the other three.
Would resources. Like the coins in snow ball you have bushes, logs, flowers, crop patches, rock & scrap piles. And cows, We have cows! Search these for resources.
With Tool equipped you can knock down small trees spawning a log. "Harvest" building for scrap. Bigger better tools = bigger tree, bigger log, bigger lootable piles. Resources yield a spent flag, a location tag/list to stop you looting the same place over and over.
Build structures including quarry, mine, etc. 
Build garage that spawns/ rebuilds vehicles for big harvest, bigger construction. 
Build entertainment, quarters, story and end game buildings. Re build civilization.

I already have a story arc and end game planned out that would fit in the world. Very few world changes needed.
It is called Abdicated. Humans have been uploaded to virtual worlds for ~1000+ years. Something has gone wrong. You have been sent to the real world to fix the hardware and start with nothing but a robe in a run down medical facility. Your information about the real world is somewhat out of date!

[fly77]

So finally I figured out how to procedurally spawn (and delete) worldwide  "empties" (cubes) which in turn procedurally spawn (and delete) each one his own scenery stuff (here a dino and 3 deers, just as a test) .
I still need to corrrect some bug to make the deleting part work reliably (so not to leave stuff behind)

[wesleyibruce]

Great work. Fun choice of animals. I may mention this at my creationist meeting on Tuesday.
Good clean up is the key. In some games spawn rate is tied to server load and player population. Though player population is not high in Outerra. I'll have to have a chat to you on voice we can cover stuff faster that way.
I have other things eating my time. After years trying to get my brothers to play Space Engineers with me they do and its a disaster. I never realized how cautious I was. They weren't. lol. I'll have to drag them into Outerra. What could possibly go wrong?

[aWac9]

fly77. stop bringing bugs to outerra, you will make it impossible to fly the skies of outerra.

[fly77]

yes I need to fix the clean-up. "Cover things faster" ? Well actually as you see it takes a lot of time to get even simple things working in outerra without issues. Partly its because there is no "user manual" or ready made code examples for these things in outerra, and for sure its because I am not a professional programmer. Even the engine itself probably lacks some functionality ...and crashes are happening very easily....so I consider myself fortunate if I get anything working at all.

[fly77]

fly77. stop bringing bugs to outerra, you will make it impossible to fly the skies of outerra.

Just zip back in time 150 million years and problem will be solved. No planes , no problems ! 

[fly77]

getting slightly more serious with hierarchical procedural spawning: a nice idea  as a first test could be Farms with different animals, vehicles, crop, plants, buildings...randomly picked and arranged from a list of these items
and spawning them worldwide on flat terrain. ...creating variations from farm to farm...and possibly matching to local fauna/flora/building/equipment styles...lots of challenges to be dealt with to get it done...certainly a good excersize of "many things procedural" ! 

[wesleyibruce]

By Jove I think he's got it. Good stuff. You need random rotations and they need to be indexed off latitude and longitude with the first two digits removed so people in multiplayer see the same animals, positions and facings.
I am not sure if a tractor counts as an animal?
How is the band width and load speed?
I'm thinking about understory plants. We can't swamp the world with vegetation. So we need to localize it and clean it up as someone moves on. Also if we can spawn it and de spawn it we can make it 'lootable' or harvestable.

[fly77]

By Jove I think he's got it. Good stuff. You need random rotations and they need to be indexed off latitude and longitude with the first two digits removed so people in multiplayer see the same animals, positions and facings.
I am not sure if a tractor counts as an animal?
How is the band width and load speed?
I'm thinking about understory plants. We can't swamp the world with vegetation. So we need to localize it and clean it up as someone moves on. Also if we can spawn it and de spawn it we can make it 'lootable' or harvestable.

yes need random rotations and also offsets.... and yes they're already indexed by lat lon. Unfortunately there won't be multiplayer for the next few years so there's no need to think about how to share those things among players. ..lootable/harvestable is certainly interesting....but for now I just need to get something basic done which works and which I can share as a mod ...still having problems with cleanup...load speed is good

[wesleyibruce]

The first rule of multiplayer is plan for it early and in every change. Otherwise you have to rewrite all the code from scratch to make it work. So that's why I always consider it.