Print

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

[FantasticDorf]

Some aquifer shenanigans on this version, unsure but likely on previous ones too as detailed here 0011382: Unmarked aquifer embarks spread water from non-aquifer stones, i've had my food production on the 5th layer of chert & gabbro (where aquifers shouldnt even be present because not on material definitions) swell up with disproportionate groundwater and evaporation for whatever reason has stopped occuring even with the walls smoothed.

Ill probably be playing without aquifers until its addressed, not that light aquifers aren't fun technical challenges without this bug complicating things with fun things like easy mist generators & drainage considered. The new relationship changes feel good with firsthand experience; socialization is slow but gradual, some migrants have hyper-socialization skills mostly related to being peddlers compared to a few dwarves with little to none but i think it'll iron out eventually.

• Though i wonder if dwarves using the same person-to-person AI could be used on decorative objects in related zones to give a lift to displays, statues and zoos which are still slightly neglected for slowly guiding the katamari-ball of talking dwarves near them (self concious it sounds a little demanding)

[charming.snark]

Here's a fresh batch of "long pocket world" worldgen crashes.

Not super scientific, but it feels like the rate of crashing is about the same as 0.47.02.

I re-ran each found seed a couple times to rule out cosmic rays.

If this is not useful or if there's something I can do to be more useful, let me know.

df_47_03_win (win10 home 64-bit, intel) totally fresh install

year 170
http://vpaste.net/Qv7iO

year 210
http://vpaste.net/33gRu

year 538
http://vpaste.net/Oj5EH

year 719
http://vpaste.net/RWGeQ

[Toady One]

Are the seeds still exporting with that extra space in there, or were those copied over from 47.02?  I haven't gotten a crash yet, but I'm still trying.

[ZM5]

I copied mine directly from 47.03's gamelog.

[Ryan]

I'm now only getting crashes right at the very end of generation.

*** STARTING NEW GAME ***
Generating world using parameter set CREATE WORLD NOW = 5555333
 Seed: HZEifMcy6qP5nVpn1K3a
 History Seed: tyYLCA7AWZrwPEfo3rzW
 Name Seed: x9EThGxRVElu2nPTPLvS
 Creature Seed: WTeoYQyEQO2B4Jg6k9ta
Generating world using parameter set CREATE WORLD NOW = 5555333
 Seed: pdwkQ48EV6QlI1rSyoR3
 History Seed: LR9GkyzyFYPZmYsDb5nc
 Name Seed: mWrY7HfcfuS0pO8yBpxT
 Creature Seed: LvpxxnPUn4iUQCva0fcf

[charming.snark]

Oops, sorry, the worldgen params (but not the seeds) are from 0.47.02.

My methodology has been to:

- Create a world with "create world now"
- Save that world's worldgen params to use as a "template"

Then

- Use those worldgen params to generate more worlds with the CLI
- If there's a crash, copy the worldgen params to a new file (e.g. crash_year_712.txt) and insert the seed lines
- Copy crash_year_712.txt over world_gen.txt
- Generate the worlds with the crash seeds a few more times and verify that it is the same world and crashes on the same year
- Repeat

The idea being that you could just slap this into world_gen.txt, select GETCRASH and easily reproduce the crash.

Not sure why this doesn't seem to be working. Now that I'm thinking about it, I wonder if I'm either using the world generation CLI improperly, or if the error has just been failing to inform you about this. Although I was previously verifying the crashes manually through the advanced worldgen interface, so maybe not.

They're being generated with

Code: [Select]

'Dwarf Fortress.exe' -gen 0 RANDOM "GETCRASH"
In other news, I've been generating "default settings" worlds for 5 hours (50 worlds or so) and haven't seen any crashes.

[martinuzz]

Embarked on 47.02 on a site that had no aquifer according to the map, yet it had at least 4 layers of sandstone slow aquifer.
Maybe more but I gave up building down further when I hit the 4th layer of damp stone.
I did use the modest mod + accelerated, but I don't think that should affect aquifers.

[risusinf]

Embarked on 47.02 on a site that had no aquifer according to the map, yet it had at least 4 layers of sandstone slow aquifer.
Maybe more but I gave up building down further when I hit the 4th layer of damp stone.
I did use the modest mod + accelerated, but I don't think that should affect aquifers.

Same on vanilla .03, three layers of slowly soaking sand/whatever else, not specified on embark map.

[Toady One]

Got a w.g. crash now, but the crash year is hopping all over and I'm getting nowhere fast.  But this is normal.  Just have to push at it for some more hours and hope it pops me out somewhere meaningful.

As I recollect, the mismatching aquifers is an old issue - I looked at it some time back (maybe when I was doing these aquifers, but maybe before), and made some progress getting the frequency down, but I haven't figured it all out yet.

[jecowa]

In other news, I've been generating "default settings" worlds for 5 hours (50 worlds or so) and haven't seen any crashes.

That sounds like a good methodology. With default settings the worlds are 250 years old, but my crashes have always been at later years than that. My earliest crash was at the year 320. I think my latest crash was around year 990.

[ZM5]

Got one right as it got done placing civs and was about to start history - still using the same rawset I put up before. All default values.
CREATE WORLD NOW = 4333333
 Seed: V9c0lWKiYypPBfrFfm3A
 History Seed: beyhSg54qzzKxMxiVFJw
 Name Seed: LZuvxgzLbwRUEhhvxPLc
 Creature Seed: k1qOWemMEOgfqFUcAX7k

This one crashes at year 81
CREATE WORLD NOW = 4333333
 Seed: HzOc9GgO4MVJ7bdrXWtc
 History Seed: tgGLGULeCFtw7M3Uph9K
 Name Seed: Fb25zAnrjST42r9r191e
 Creature Seed: qziCcyeqya8d2XCwWtPE

[Untrustedlife]

Thanks for the update Toady. It would be nice to clearly identify which of those fixed bugs were long-time issues, or just bugs introduced since 0.47.

The random hist fig bug has been around since the first re;lease of the last big version
http://www.bay12games.com/dwarves/mantisbt/view.php?id=11019

So thats an old bug.

[charming.snark]

Here's a fresh one from a 0.47.03 worldgen params file. Params are for a default "create world now"

stock df_47_03_win (win10)

Adreorid, "The Past Universe"
Crashes in year 182
http://vpaste.net/WuIr1

Would some of you other folks posting seeds mind verifying this? Just paste those worldgen params into world_gen.txt and select it in the create advanced world screen.

Are these things just not consistent from machine to machine, or just not from machine -> Toady's dev environment? The only other machine I have handy to test with is a mac.

[ZM5]

Off-topic from the crashes, but I notice that intelligent wilderness creatures still aren't capable of worshipping deities.

[Shonai_Dweller]

Off-topic from the crashes, but I notice that intelligent wilderness creatures still aren't capable of worshipping deities.

Isn't that expected?
Intelligent wilderness creatures aren't from civs, you pick them up just as they join a civ so they won't have had time to integrate and become religious (and adopting a religion isn't something you can do in-game yet).