Let's talk about Zlevels

Started by Daman453, September 01, 2015, 10:27:36 AM

Previous topic - Next topic

Toggle

Yes, it would be hard, as you apparently have no idea what you're talking about. When there's no support, it just drops rubble on the ground. Having z-levels is entirely different then simply having roofs collapse.
Selling broken colonist souls for two thousand gold. Accepting cash or credit.

blub01

I meant the bit about having buildings collapse when they are too high/badly built from a structural point of view.
Quote from: Zobaken on September 02, 2015, 12:37:37 AM
1. Please make people unable to move through deep waters. I don't like raiders cosplaying Jesus.

Goo Poni

Quote from: MarvinKosh on September 02, 2015, 12:30:08 AM
I'd much rather just have a pawn go into a rabbit hole with wooden/steel supports to mine deeper.

Ah, I remember the joy of ATTEMPTING to build a house in Towns. 10 hours in that failed game, I haven't successfully built a simple square box, let alone a castle or anything more elaborate. I've also lost towns to the townies quarrying the ground and then dying because they dug themselves into a pit they couldn't climb out of, which spawned ghosts that then totally wrecked my shit.

akiceabear

I think z-levels would greatly enhance the game. A few thoughts:

Z-levels impact on overall gameplay enormously positive. I think the payoff between developing z-levels and populating other content in existing systems is significant, because z-levels offers a much more flexible way to play the game creatively. This type of story is impossible in RimWorld - in fact, most of them for DF are, and z-levels/related mechanics are a big reason. Adding z-levels can really enrich the variety of emergent stories driven by creative base planning and player-set goals, something which RimWorld is lacking in at the moment. I also list a few examples how z-levels could get great for emergent gameplay in this thread.

Z-levels is a major feature - but not impossible. I think some here tend to over- or under-emphasize the development cost (time) for the feature. Yes, it would be a huge challenge, taking perhaps 1-2 years to get everything right. But would it be a completely re-write of all systems? I doubt it - most mechanics would rely on a new pathfinding system. For example, assume relationships are implemented in A13 - do they need to be rewritten for z-levels? I doubt it. Moreover, the process could be compartmentalized so that backers for the new system could provide feedback along the (admittedly longer) alpha cycle. For example:
A1 - basic system to build up/mine down, UI to navigate
A2 - pathfinding - which are called on by other systems like work/relationships
A3 - combat - how height changes probability to hit/damage with ranged weapons, how AI pawns use that plus the new pathfind system to wage war
A4 - construction - beyond the simple building up and down, now structural soundness is a consideration (like with roof collapse), also considered in combat AI
A5 and beyond (?) - water dynamics for underground lairs, new event types specific to elevated/sunken z-levels, new tech for them also.

Some of the mechanics preferred over z-levels are going to be developed anyway, as part of the current RimWorld. One of the most often requested features is a relationships system - but Tynan has said before that he may focus on that for A13 when he is back from his break. It seems silly to request something which is already likely part of the development pipeline.

Some of the content preferred over z-levels will be developed by the community or in a relatively short time. Tynan previously stated (in 2013) that he could develop a dozen new raider types or buildings with 6-days time. But here we are in 2015 without hundreds of buildings and raider types. I think this reflects his development philosophy, which is to focus on systems that interact in emergent ways, rather than spamming content. As result, I think its unrealistic to expect a huge amount of new content from him - it seems clear to me he is going to leave that to modders using the systems he's developed. Rather, he's probably likely to continue at the pace he's kept content-wise - a few new bits and bobs each alpha for old systems, but focusing more on new systems and their content.

Z-levels would clearly require a new development round via a paid expansion or sequel. I think its very clear that z-levels is unlikely as part of development before 1.0. More likely is that it would be a feature in an entirely new development round, either an expansion or a sequel. Then it becomes a case of funding what only Tynan/Ludeon can develop - I'd happily pay $50 for fleshed-out z-levels implementation, maybe $15 for a solid fog of war implementation, but probably only $5-10 for a content-for-existing-systems expansion. Why? It's clear that the small but great modding community is already very capable of filling old systems with new content. Abrexus' mods basically scratch that itch for me entirely with regards to tech/buildings. Hell, there have even been some amazing event mods like TTM and Crash Landing. I think this modding will only get richer as the game goes to Steam and gains a wider audience, which will make Tynan shifting his focus from new systems to old content expansions even less meaningful and worth my (additional) money.

Dawa1147

I understand the advantages of z-levels, but I think true z-levels would be too much (e.g.: a 200x200 Map needs to simulate 40000 Squares plus the content of said Squares. Even only 5 z-levels would put that to 200000+content, ), and that is a lot (especially since many Computers struggle already on that map if it burns or you have many colonists). What might work is fake z-levels (e.g. trenches technically are below the surfaces, but would be treated like a counduit that slows you but makes you harder to hit), and at most one z-level up.

akiceabear

Hi Dawa - that's a common suggestion. Tynan has said if he does zlevels, he'll do it for real.

blub01

also, it could maybe be an optional thing, so weak computers can run 2d rimworld, while better PCs provide fancy 3d rimworld.
Quote from: Zobaken on September 02, 2015, 12:37:37 AM
1. Please make people unable to move through deep waters. I don't like raiders cosplaying Jesus.

b0rsuk

#22
I understand that it can be very complicated. Even after all those years, Dwarf Fortress still didn't get it 100% correct! Tamed flyers can't fly. Hatches are invulnerable from below, monsters can't see what's above them - does it mean bowmen can't shoot creatures on towers ? I don't play DF, I just read wiki and stories sometimes. I dabbled in it and found the interface/micromanagement insufferable.

But it's also a very lucrative feature. Many fun things become possible. As I hinted above, it doesn't have to mean copying DF. Rather than digging down, Z-levels in Rimworld might focus on building UP. Flying vehicles, flying creatures, jetpacks, orbital strikes, drop pods on your roof, meteorites, parachutes, cats falling 10 stories with just minor injuries, collapsing structures...

It's also very hard to have convincingly running water without Z-levels. How are you going to implement rivers ? Heightmaps ? How to visualize it on a single plane ? Does it sound exciting to play a game like this without rivers ?

As far as I remember Z-levels is one of reasons Tynan hasn't implemented embrasures. I think he said if he implements something like that, it will be via Z-levels.

Quote from: blub01 on September 06, 2015, 03:09:13 AM
also, it could maybe be an optional thing, so weak computers can run 2d rimworld, while better PCs provide fancy 3d rimworld.
And it's harder to program something to be optional rather than either on or off. And I'm not sure you understand the increase in complexity. Just consider the number of tiles. On a 200 * 200 map you currently have 40.000 tiles, which track state, what floor does it currently have, what floor is there when deconstructed, furniture, creatures, items. 200 * 200 * 200 map (200 zlevels) would be 8.000.000 tiles.

Suppose a modern computer can just barely run Rimworld (not true). Do you think 200x more powerful processors exist on consumer market ? How many years would we have to wait ?

Maybe it would be more reasonable to limit Z-levels to a low number, like 5 or 10.

blub01

Quote from: b0rsuk on September 06, 2015, 04:09:55 AM
I understand that it can be very complicated. Even after all those years, Dwarf Fortress still didn't get it 100% correct! Tamed flyers can't fly. Hatches are invulnerable from below, monsters can't see what's above them - does it mean bowmen can't shoot creatures on towers ? I don't play DF, I just read wiki and stories sometimes. I dabbled in it and found the interface/micromanagement insufferable.

But it's also a very lucrative feature. Many fun things become possible. As I hinted above, it doesn't have to mean copying DF. Rather than digging down, Z-levels in Rimworld might focus on building UP. Flying vehicles, flying creatures, jetpacks, orbital strikes, drop pods on your roof, meteorites, parachutes, cats falling 10 stories with just minor injuries, collapsing structures...

It's also very hard to have convincingly running water without Z-levels. How are you going to implement rivers ? Heightmaps ? How to visualize it on a single plane ? Does it sound exciting to play a game like this without rivers ?

As far as I remember Z-levels is one of reasons Tynan hasn't implemented embrasures. I think he said if he implements something like that, it will be via Z-levels.

Quote from: blub01 on September 06, 2015, 03:09:13 AM
also, it could maybe be an optional thing, so weak computers can run 2d rimworld, while better PCs provide fancy 3d rimworld.
And it's harder to program something to be optional rather than either on or off. And I'm not sure you understand the increase in complexity. Just consider the number of tiles. On a 200 * 200 map you currently have 40.000 tiles, which track state, what floor does it currently have, what floor is there when deconstructed, furniture, creatures, items. 200 * 200 * 200 map (200 zlevels) would be 8.000.000 tiles.

Suppose a modern computer can just barely run Rimworld (not true). Do you think 200x more powerful processors exist on consumer market ? How many years would we have to wait ?

Maybe it would be more reasonable to limit Z-levels to a low number, like 5 or 10.

I suggested a solution to get more "complete" z-levels without killing the game on weaker computers. Also, I'm kinda curious what is the difference in the code that allows minecraft to load up 16^4*12²*Pi(about 29740000) blocks at a time while rimworld struggles with 300²(900000)? is it because the jobs? because you usually have quite a big number of entities in minecraft too. the temperature system can't really be it because there aren't enough rooms for it to truly matter. maybe rimworld isn't optimized in some important regard?
Quote from: Zobaken on September 02, 2015, 12:37:37 AM
1. Please make people unable to move through deep waters. I don't like raiders cosplaying Jesus.

zandadoum

i didn't read the comments, but i wanted to say that before zlevels are implemented, please implement a sprinkler fire extinguisher... otherwise it will be impossible to fight fire undergound with 1500 degree rooms

also: maybe oxygen should/could be a factor (even for fire extinguishing undergound: turn off oxygen supply, fire dies)

b0rsuk

Maybe Rimworld isn't optimized... but Rimworld has many colonists, raiders and animals. That's a lot of pathfinding. 3D pathfinding would be even more expensive.

Minecraft is mostly static, isn't it ? It also uses different data structures. I'm not sure voxel octrees or whateer minecraft uses were around when DF was developed.

I'm not competent enough in game programming to evaluate, but this article sounds informative. It describes the difference between flat arrays and octrees, cites Minecraft etc... http://0fps.net/2012/01/14/an-analysis-of-minecraft-like-engines/

blub01

Quote from: b0rsuk on September 06, 2015, 08:38:51 AM
Maybe Rimworld isn't optimized... but Rimworld has many colonists, raiders and animals. That's a lot of pathfinding. 3D pathfinding would be even more expensive.

Minecraft is mostly static, isn't it ? It also uses different data structures. I'm not sure voxel octrees or whateer minecraft uses were around when DF was developed.

I'm not competent enough in game programming to evaluate, but this article sounds informative. It describes the difference between flat arrays and octrees, cites Minecraft etc... http://0fps.net/2012/01/14/an-analysis-of-minecraft-like-engines/

I'm not good at programming in general, and as a result I understood almost nothing, but I think the article basically says that code that runs again and again and again(like physics updates, lighting) is far more important to optimize than random access(placing blocks, and i think player movement counts), by a factor of the radius of the map squared, if i understood it correctly. It(the article) then proceeded to go into detail about how to optimize iteration, at which point I ceased to udnerstand a word of it.
Quote from: Zobaken on September 02, 2015, 12:37:37 AM
1. Please make people unable to move through deep waters. I don't like raiders cosplaying Jesus.

b0rsuk

Yup, it's about optimizing the right thing, solving the right problem. And people argue what actually IS the most frequently running code.
But it does give you an idea how complex it can get! You better eat higher math for breakfast.

More about Z-levels - one of Rimworld features that disgust me is automatic roof creation. No, I know there's a button to disable it. I mean how cheap, easy and exploitable it is. You can roof/unroof something remotely, at will. Got a fire in an open colony ? Unroof, hold doors open, problem solved. The only time you might avoid that is in a greenhouse. It insta-kills your plants. It's nuking from orbit!

Imagine what would be possible with Z-levels! Solid stone roof for extra protection ? Or maybe a translucent roof that lets sunlight in and saves power ? What if cargo pods fall on your roof ? How expensive is reinforced glass ? Z-levels would make roof building interesting. They would also make roofing easier to understand, less magical. Will that mortar round pierce my roof ? Yes ? No ? Why ? Now you have an idea.

Toggle

Yeah, but if we didn't have the ability to unroof, we'd need fire extinguisers, as currently any fire in your home area for me would burn my bloody colonists alive from heat since there's no good way to get rid of it.
Selling broken colonist souls for two thousand gold. Accepting cash or credit.

blub01

or water/sand buckets, for early game, and for the base, a sprinkler system would be better anyway.
Quote from: Zobaken on September 02, 2015, 12:37:37 AM
1. Please make people unable to move through deep waters. I don't like raiders cosplaying Jesus.