wraitii

This looks excellent, congrats!

If you're developing a complete-information game, then using our architecture makes it logically/mathematically impossible to cheat and be undetected. 0 A.D. is not, and it is possible to cheat slightly, by getting more information than the game wants you to have, because of that.

I believe civilisations are rather cheap overall; templates are costly in terms of javascript memory.

I think the current idea is to refactor the "helper" Library functions so that they can be called from arbitrary places in the code, then create proper encyclopaedias and trees everywhere needed. It's basically planned, but progress is slow and steady.

I do believe I sort of fix (2) in D13. Nice catch on the first bug though.

Correct, units no longer have a per-entity obstruction but rather have a "clearance" defined by their passibility class. Pathfinding-wise, that's the only thing that matters.

Also need better data-structures in the engine to handle this kind of size or things are going to go wrong.

I don't actually think they must be trained individually. We could dynamically add units to a bataillon. But that doesn't change that they need to be treated as a single entity.

It seems you are trying to play Rome Total War.

On a quick note: Atlas currently recomputes the whole pathfinder grid (on top of water stuff) when we change the water height, and that's also extremely slow. 50% is the terrain stuff, 25% re-rasterizing entities, 25% recomputing the hierarchical pathfinder entirely. So we probably should try and speed that up a tad if we start having maps that change water height, and/or we should change atlas water height from a slider to a text-box. The slider compounds the problem.

Well obviously we support level 1. Level 2 is easy enough. Level 3 would need some new code. We'll probably get there eventually, but can't say anything on the when.

The most likely outcome, to me, right now, is that we'll ultimately support 3 things: -individual units -individual units walking more or less in formation (like in Age of Empires 2) -actual bataillons for warfare, treated as a single unit, but possibly composed of several underneath.

That's entirely dependent on what type of game you are playing/making. I do agree that it's most likely not a good idea for an AoE-like RTS though.

Ok just wanted to make sure I think your problem is that the height map is 16-bit and the usual image is 8-bit only.

@shieldwolf23: the water height is a setting in Atlas, so you can raise it or lower it.

I believe the solution was that water height is a separate setting in 0 A.D.?

Seems to me 10.8 might be a little too old to run the game, sadly :/

And why is that (in this case)? Because we do not have D270. Just saying.

Correct, CCs are available in phase 2. You'd build a CC to deny resources to your opponent and to build buildings. I think it's not too much of a nerf. It's a problem on RMs because some are just terribly balanced, so you end up without metal or stuff. It also sometimes forces you to place your CCs in a braindead way, not as strategical way. Ok on the worker elephant I guess, it might need a buff.

Hm, tbh I'd like to keep the scope of those threads somewhat limited to mostly SVN stuff so that the changes are easy to implement

Hm, I would say yes. For storehouses, I think we should consider it, but I'm honestly not sure how it would play out.

My personal preference would be for a new storehouse building, that you could build somewhat quickly and cheaply, and would allow collecting any of the 4 resources, perhaps with a special new component that would give you only 75% of resources you deposit in it or something (without tech?). I think it'd open new gameplay avenues and mitigate the issues with our current system. An alternative I would consider quite viable would be a technology, unlockable in perhaps town phase, that allows dropsites to be built in neutral territory.

Let's keep keeping on bringing back proposals for things. 0 A.D. has one issue with territories: you can't make dropsites outside of your borders. This isn't a problem on some maps, but it is on others where resources are just not good enough inside your starting area. It also makes metal-poor maps sort of really force you to place CCs near resources, and it kind of plays in into why our metal mines are generally so large. There's a few solutions, which you can vote on on the poll above and discuss below (I'll apply my "keep it on topic" policy, so may hide posts that go OT): Keep as is right now, I don't think this is a real problem Make dropsites buildable in neutral territory but not enemy territory Add a new "storehouse" building for resources (all of them?) where you can't research techs, and has limited HO and cost, but allows you to gather outside your frontiers. You'll still need dropsites for techs, possibly dropsites could be given a gather aura? I'm also accepting other proposals.

Yup, I believe I did that too Some screenshots to describe my annoyances with Random maps. Here's Ardennes RM: there are easy-to-dropsite forests, with lots of wood. The micro is actually relatively limited here, it's nice, you can focus on other stuff. Here's the Sahel skirmish map. Though Sahel has limited wood, the Skirmish map offers natural areas for a dropsite. Here's, in contrast, the Sahel RM. It has no such area. Other examples or terrible RM generation: Hyrcanian shore is not supposed to have limited wood, but there are no good dropsite zones in this area: Finally Persian Highlands, which almost-gets-it-right-but-not-quite: IMO most RMs should be reworked so that trees are clumped a little more. It wouldn't really hurt aesthetics if done carefully, and it would increase playability a lot. Note that this problem is compounded by the need to have CCs for territory, since those are horribly slow and costly so they need to be worth it.

That would be fixed by my unit motion rewrite which reintroduces a viable way for units to walk together at the same speed.