A Radicaly New Approach for Viable Planets:

[MacThule]

Soo... you seem to support OB, but feel like it would also devalue planets. How do you propose countering OB genocide being the easy defacto solution and making worlds actually worth settling to begin with if they are that easy to take out? In Other words. what balance would exist to make it either worth the risk of an OB to colonise, and/or the value of a planet being intact to be worth a ground invasion at least some of the time?

Passive resource income.

Lots and lots of it.

 

[MacThule]

Also, a thought... to prevent scorched earth campaigns of bombardment and even heavy atmospheric/surface razing from ruining the long-term fun of a server (possibly forcing an early reset), it should be pretty easy to set a re-generation timer for any planetary plate/hex that loads in 1) without a faction block on it, 2) that has taken weapons damage over 20% of its total value, and 3) has not been loaded in at all for at least 1wk (ie is either middle of nowhere or is being ignored because it's been bombed worthless). Such a plate entity could simply restore its original surface, or randomly generate a new surface fit to its planet type. The timer could be achieved with a single variable and would only need to be checked once when a plate loaded in (although plate loading is already so slow, the notion of adding miliseconds to that particular disaster does make me cringe a bit on the inside).

Server adjustable timer (including to -1), of course.

 

[Winterhome]

Also, a thought... to prevent scorched earth campaigns of bombardment and even heavy atmospheric/surface razing from ruining the long-term fun of a server (possibly forcing an early reset), it should be pretty easy to set a re-generation timer for any planetary plate/hex that loads in 1) without a faction block on it, 2) that has taken weapons damage over 20% of its total value, and 3) has not been loaded in at all for at least 1wk (ie is either middle of nowhere or is being ignored because it's been bombed worthless). Such a plate entity could simply restore its original surface, or randomly generate a new surface fit to its planet type. The timer could be achieved with a single variable and would only need to be checked once when a plate loaded in (although plate loading is already so slow, the notion of adding miliseconds to that particular disaster does make me cringe a bit on the inside).

Server adjustable timer (including to -1), of course.

Partial regeneration. Generate a new surface, but not regenerate anything that's been placed by a player. But that might not be possible with how few bytes we have to work with.

Soo... you seem to support OB, but feel like it would also devalue planets. How do you propose countering OB genocide being the easy defacto solution and making worlds actually worth settling to begin with if they are that easy to take out? In Other words. what balance would exist to make it either worth the risk of an OB to colonise, and/or the value of a planet being intact to be worth a ground invasion at least some of the time?

Farming and passive income, combined with planets potentially representing an enormously powerful defense mechanism.

I'd like to see planets be able to be given orbital defense cannons, which operate sort-of like an "inverse" orbital bombardment going by the system Captain Fortius described. Link weapons computers or turret dock points to an AI module, and set targeting parameters. The AI module detects what kind of entity it's been placed on, and if it's on a planet surface, it gives the option of targeting entities in space, and saves the details of the turrets/weapons and its targeting parameters, as well as the locations of the weapons. Entities approaching a planet that fit the parameters laid out in the AI screen are automatically fired on from their recorded position and don't stop until the region they're firing from has been suitably bombarded, even without the blocks being loaded or the damage being calculated on the actual planet. Ingame explanation would be that the planet's military staff evacuated the position of the turret.

 

[Plazmatic]

[doublepost=1516099353,1516099091][/doublepost]On a more serious note there is one major problem, how would orbital bombardment work, I guess it could be removed making way for a new class of ship designed specifically for atmospheric combat or bombing

We've covered this again and again, incoming damage from lat long can be converted to 'real world' coordinates on the ground precisely. If you've solved the "how do I figure out where I land when I enter the instanced planet" you also solved the "how do I know where my lasers/ bombs/ bullets end up" on the planets surface. The question isn't "is orbital bombardment possible with instanced planets?" It is, and not very difficult, the question is whether it should even exist and how powerful it should be, as others have mentioned, because of balancing issues.

Partial regeneration. Generate a new surface, but not regenerate anything that's been placed by a player. But that might not be possible with how few bytes we have to work with.

It is possible, procedural generation needs only the seed to recreate what it initially created (hence why mine-craft has a seed option that you can use for your worlds to copy other world generation). With pure terrain generation this would be simple, since perlin/simplex noise is likely the source, any point you generated is completely independent from all other points in perlin noise, which means you can do this per hexegon/square/what ever sized chunk or specific blocks you selected. Some Minecraft servers do this as well (re-generate chunks every so often on servers with separate mining worlds), this is not unheard of.

Farming and passive income, combined with planets potentially representing an enormously powerful defense mechanism.

I'd like to see planets be able to be given orbital defense cannons, which operate sort-of like an "inverse" orbital bombardment going by the system Captain Fortius described. Link weapons computers or turret dock points to an AI module, and set targeting parameters. The AI module detects what kind of entity it's been placed on, and if it's on a planet surface, it gives the option of targeting entities in space, and saves the details of the turrets/weapons and its targeting parameters, as well as the locations of the weapons. Entities approaching a planet that fit the parameters laid out in the AI screen are automatically fired on from their recorded position and don't stop until the region they're firing from has been suitably bombarded, even without the blocks being loaded or the damage being calculated on the actual planet. Ingame explanation would be that the planet's military staff evacuated the position of the turret.

You could have a base/use faction claim block module like you have a star-base module, have it the upgrade system work similarly where it has separate rules since it is static allowing it to have different characteristics to space ships in a already-implemented-in-game way.

 

[MacThule]

Partial regeneration. Generate a new surface, but not regenerate anything that's been placed by a player. But that might not be possible

Resurfacing was my desire as well, but I thought about it and realized that it would probably require a fairly complex AI algorithm to pull off so I went with something simple. It could be upgraded to resurfacing later though, that would be awesome.

Farming and passive income,

Those, additional faction points, maybe a better mining bonus in systems with controlled planets. There's a lot they could be that would make players beg for the opportunity to get bombed out just because until it happened they would be living large.

 

[Winterhome]

Resurfacing was my desire as well, but I thought about it and realized that it would probably require a fairly complex AI algorithm to pull off so I went with something simple. It could be upgraded to resurfacing later though, that would be awesome.

I was leaning towards something along the lines of using the planet/universe seed to regenerate the planet surface the same way it's generated the first time, but I'm not entirely sure how to do that but exclude spaces specifically mined out by players and playermade structures. The generation is the easy part - excluding important stuff from the regeneration is the hard bit, I think.

Another possibility would be to go full Dorf Fortress on it and whack the surface a few times with an erosion effect until the craters and holes get filled in and smoothed out, but that's a bit more intense and may result in lag spikes.

 

[Plazmatic]

Resurfacing was my desire as well, but I thought about it and realized that it would probably require a fairly complex AI algorithm to pull off so I went with something simple. It could be upgraded to resurfacing later though, that would be awesome.

It wouldn't be too difficult, I explain a bit of this in my previous post. Because the noise algorithm used to generate planets is very likely to be coherent (given the same inputs, generates the same outputs) and likely perlin based, given the same seed used to generate it initially, you can just regenerate the chunk with out regenerating anybody's buildings (since they weren't even a part of the noise generation). This is something that minecraft servers do often to regenerate ores for mining. If you want to keep player stuff that works too, when storing information about procedurally generated terrain in a mine-craft style game, you don't store the generated data, you store the seed, re-generate, and then apply the changes the player made, which means you can just regenerate the chunk, and then apply player made results that aren't the empty block.

I was leaning towards something along the lines of using the planet/universe seed to regenerate the planet surface the same way it's generated the first time, but I'm not entirely sure how to do that but exclude spaces specifically mined out by players and playermade structures. The generation is the easy part - excluding important stuff from the regeneration is the hard bit, I think.

This is a bit confusing, if you regenerate the entire surface, but keep player made structures and mined areas, then you've effectively done nothing. I think you may mean keep ground player made modifications, and regenerate orbital damage? Since you only save player changes anyway with procedurally generated worlds, you would just segment the changes into space based changes and ground based changes, the biggest issue here is keeping orbital damage on bases if that is desired, since that would require cross referencing orbital changes with currently created structures by ground modification.

 

[Dire Venom]

Well RIP this concept since the Devs will not consider instanced planets. I do hope the devs can pull through on their vision of planets.
As long as they're useable and don't lag the server to death like the past few itterations they've got my (cautious) support.

 

[Plazmatic]

Well RIP this concept since the Devs will not consider instanced planets. I do hope the devs can pull through on their vision of planets.
As long as they're useable and don't lag the server to death like the past few itterations they've got my (cautious) support.

Instanced planets were really an implementation detail aimed at improving performance, it doesn't change the gameplay of core concepts of the ideas proposed in this thread. When we say "we want instanced planets" what we are really saying is "we want non polygon-like large planets with out lag". They said they are overhauling the planet system, or that it will be overhauled, if they manage to achieve large planets with out lag nobody is going to complain that it that they didn't use instancing to achieve that. And even the idea of "large planets non polygon planets" is really a result of "we want planets that are actually good". If the planets are implemented well, you won't have large swathes of the player base complaining about them.

 

[EricBlank]

Using a separate sector/wormhole atmosphere to instance planets sounds pretty much like the best idea ever, tbh. It takes time to actually load the chunks in to visualize things anyway. This would be much better than both a loading screen and the current dodecabedrons, which do put a huge strain on my pathetic little laptop.

The thing about collisions is youd want a reasonably good collision mesh on the ground for land vehicles like our repulsors and players/fauna to interact with, and allow craft to actually land, not to mention actually interact with individual blocks of the terrain. Perhaps if the collisions become abstracted if a craft is moving above a certain velocity or has been flagged as having been in the air for a certain length of time? We know you only need three points to define a plane in 3 dimensional space, we can use the three highest points in a given chunk/section of a chunk to define a plane defining the elevation and slope of that area, below which a ship is pushed upwards and friction to its velocity perpendicular to that is also applied, as if it had crashed into . If this causes it to reach zero velocity (say it was going straight downwards) then it is bouyed back up until it is no longer in contact with any block, at which point full game physics begin to apply and it can collide with the blocks as normal. If it is underground by more than a certain amount or for a certain time it can begin to destroy soil and other soft materials blocks around it, lowering/changing the angle of the plane and creating a hemispherical impact crater or series of craters. Whether or not the craft takes collision damage can depend on its own properties i guess.

 

[alij331]

i love that idea!

tbh, if flying parallel to the ground, i could see such a plane working a little differently (if possible).

if the entity hits a point of the plane, then its instant impact, and the ship takes velocity-based collision damage on where that point was, and the point gets a crater applied (as it will be a peak of an undulation).

if, however, the entity enters the plane away from the points, it could load the true collision mesh just for that sector - each plane would basically count as a lookup, so you would only be calling a single sector's collision mesh, which should be pretty damn quick... that way, you could fly through the valleys of tall peaks without being thrown up into open atmosphere