Inconsistent ruleVariant

[hackedpassword]

Is there an existing issue for this?

• I have searched the existing issues

Game Version

4.15.2

Describe the bug

Same render stack, two different results.

Steps to Reproduce

Adding some new spites to Ultima V tileset mod. I've run into similar issues, mostly fighting with base TileSets and their use of this+that png's bumping heads with ruleVariants.

Thanks to the stack revealer debug panel, we see the stack is the same for this grassland + forest + deer + camp tile. It's issues like this I wouldn't stand a chance figuring out without that debug panel, most appreciated adding it in.

Screenshots

See above.

Link to save file

Korea - 540 turns.json

That save is 10MB

Operating System

Android

Additional Information

I've untangled a number of sprite issues with that panel, but I'm confused how I'm getting two different renders. My best guess is that grassland 2 & 3 are in some kind of caching conflict.

If I had the time/priority I'd wipe out every this+that from both TileSets and rebuild them properly in ruleVariants. There's no reason to have +' s in the base, makes modding sprites way harder than it needs to be. /ranty but not really complaining

[hackedpassword]

K post made, to add, the above screeny uses -industrial era, the 2nd screenie does not. Doesn't make sense. At least not atm to me.

[yairm210]

The first tile is within borders so it gets the era modifier for that civ (would also get the style modifier for the civ)
The second one is outside of borders so it doesn't get the modifier

[hackedpassword]

Uhm, wrapping my head around this, I see.

You know, when you talk like that you sound like a coder-lawyer. Everyone walks away in disbelief now accepting that somehow the planet in fact spins half as fast and there are two suns! Incredible.

Ok then, the 1st tile at -31, -76 would you have a suggestion or solution how to apply the intend ruleVariant? A major issue I continue to face as noted is tileset fallbacks are aggressive in persistence. The Ultima V mod isn't a true tileset yet, applied in conjunction with HexaRealm, no fallback specified, yet still pulling from FantasyHex as a fallback.

Here, an extra dimension of attribute, tile ownership, makes matters more challenging... how to solve... back when I was trying to solve eras and other complex matters in Z2, Trog noted limits were being exceeded for an extention mod (SeventhM might have too), recommended moving to a baseruleset. Perhaps moving to a full tileset is the solution to the increasing tileset conflicts this mod is having.

[yairm210]

Not sure what to tell you regarding the borders... I mean, if you have two civs, one in modern era and one in industrial, you'd expect to see the improvements in each territory go by the era of the territory, no...?

Regarding this+that, it's mostly things that we could not separate into different layers, for whatever reason, it's been a few years since I've really delved into it

Regarding Camp etc, if you replace the camp with your camp then the era specific overrides shouldn't trigger, I'm like 70% sure?

[hackedpassword]

Borders are fine. Your explanation makes perfect logical sense. Do CS's consider eras?

To be real, I think the order of rendering gets weird when pushing a non-primary tileset, meaning HexaRealm is acting primary, being the selected tileset, UV is being overlaid as a HexaRealm theme, and FantasyHex is backing these up. Without UV being a primary, rendering logic is getting too difficult to predict behaviors.

So, maybe buggy, but I'm going to call the use case beyond supportability. UV must ascend to a proper tileset.

Yea?

[yairm210]

Sounds like a plan

[SomeTroglodyte]

Darn; I vaguely remember there were tricks to kill Lord British back then, or was that a different Ultima? Those were golden days...

[hackedpassword]

In IV yea, in V he was afK (kingdom lol!), in Online he got jumped during a Richard Garriott speech as lord British and killed to the disbelief of many spectators. Today, they're a dime a dozen ;)

(great generals)