Print

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

[PatrikLundell]

[0.42.01] I've embarked in what I thought was a single biome savage, completely flat, deadlands. However, my fortress was wiped out by undead ravens late spring. Save scumming...
Later, towards the end of the first winter, I built a watch post above the trade entrance tunnel (only entrance so far), with the recessed courtyard and original entrance floored over. Since I started building one level above the ground, I've gotten reports of rain of dwarven blood. The strange thing here is that I've gotten blood in the building (before the roof was put up) and on top of the building, but nothing on the ground anywhere, including the floor covering my courtyard?!

I know I have evil areas nearby, so the undead could possibly have migrated, but not the weather, presumably.

Do I have some kind of vertical biome separation shear? I recall encountering an oddity way back with 0.40.09(?) where I could farm properly at ground level, but not a couple of z levels directly above (on a muddied floor) close to the boundary to a mountain biome.

Edit: Looking in the air I can find invisible pools of dwarven blood 1, 2, and 3 levels above the ground. (Invisible = nothing seen graphically, but randomly moving the cursor about in the 'l'ook mode reveals them). The don't seem to fall either...

[SimRobert2001]

Weather will spread 3 tiles out from an embark.

[PatrikLundell]

Sorry, I don't understand your description. Do you mean weather will spread up to 3 embark tiles from a nearby embark level tile, thus embarking within 3 embark tiles of evil weather will cause you to get it as well (with a 3*3 embark next to a "row" of such tiles will cover your whole embark)?

[SimRobert2001]

Correct. Weather will spread 3 embark tiles.

[PatrikLundell]

Thanks! You learn something new every day...

[arbarbonif]

Higher z levels can also be a different biome.  I've had several cases where I've had blood snowing from onto non-evil biomes.  Using probe from dfhack showed that the higher z levels had different biome setups (and different temps so the snow melted before it hit the ground but stained the trees).  I've also had cases where I could only get wind power by raising the windmill, two levels up was a different biome that had 20 windpower.

My running theory is that the biome at a given level is based on the biome of nearest ground at that level (if any).

[gchristopher]

As far as I ever saw, in 0.34 and 0.40, evil weather is strictly limited to the biome boundaries. By "evil weather", I mean the Interaction that creates evil rain or evil clouds. (and reanimation.) The boundaries are definitely vertical, aligned with the biome edge at ground level.

It'd be neat to see an example of biomes varying in the sky tiles above ground. Maybe there is a bug with sky tiles that haven't been allocated yet?

Does anyone have a posted save or bug report to point to? I've spent a lot of time at evil biome boundaries (prior to 0.42) and never saw biome weather change by z-level. Interpreting the dfhack biome numbers can be hard. What was the code used to inspect them?

[PatrikLundell]

I can upload the save of my embark if desired, but any DFHack examination will likely have to wait until DFHack has been updated to match 0.42.XX...

Different biomes above ground and at ground level would explain why I've had 3 groups of fortress killing (forcing save scumming) flying undead hordes (10 ravens, a pile of giant keas, and a large group of flying animal people) in just above one year, while all the land critters have been living ones (not sure if I've actually had any living fliers).
I've given up on that fortress/world, since my aim was to try out the new stuff, not to fight flying horrors slaughtering the migrants.

[gchristopher]

Well, it'd probably accomplish nothing more than satisfying my curiosity, but I'd be really interested to see evidence of region interactions at a distance. I've spent a lot of time trying to pack as much variety as possible into an embark.

It's possible to dfhack in biomes, geologies, and rivers, though editing interactions (evil weather) remains easier to do by editing an uncompressed save file.

Is it possible that your embark borders on any evil biome tiles? e.g. the edges of an 3x3 embark will include some terrain from the surrounding 5x5 area.

Oh, also experiment with building higher! The game does not allocate memory for tiles any higher than the current terrain/construction reach, and that might be related to the observed behavior. You can force it to allocate the tiles by building higher.

[arbarbonif]

I used probe in dfhack to see the evil/savage and biome information in 40.x.  I don't have any saves for it though.  What made me look initially was the blood snow with no evil biomes on the immediate map.  I bet you would only get rain/snow, since clouds come in at ground level.

In the cases I saw it wasn't due to allocating new tiles, the rain was before I had build anything up and the wind power was only a level or two above the ground, which is lower than the sky on my maps.

[PatrikLundell]

Uploaded my save in http://www26.zippyshare.com/v/9pvVz49K/file.html.
The save folder contains both my fortress and the save copy made at embark.

It would be interesting to hear if any results emerge.

[Livingdeath]

Well, it'd probably accomplish nothing more than satisfying my curiosity, but I'd be really interested to see evidence of region interactions at a distance. I've spent a lot of time trying to pack as much variety as possible into an embark.

It's possible to dfhack in biomes, geologies, and rivers, though editing interactions (evil weather) remains easier to do by editing an uncompressed save file.

Could you explain how to do this or at least where to look?  Would it be possible to say have a 4*4 embark with all one biome and then use DFHACK to partition it into 4 different Biomes one one each corner?  Wouldn't that break a lot of things?  I saw the dfhack changelayer command but it didn't seem general enough.