Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  
Pages: 1 ... 5 6 [7] 8 9 10

Author Topic: Dwarf Fortress 0.47.04 Released  (Read 257013 times)

jecowa

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #90 on: April 10, 2020, 12:09:43 am »
I tried again on a different computer and got different results. The world generated all the way to 550.

I tried generating the same world to only 424 to see if it matched the other world from the other computer. The map was different.
Spoiler: map (click to show/hide)

The history was different:
Spoiler: world history (click to show/hide)

The maps, world histories, and world sites and pops would match in 2 worlds generated with the same seeds on one computer (even though the save files themselves came out different), but they are different from what the other computer produced. I can't tell that restarting has any difference from not restarting.
Logged

Shonai_Dweller

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #91 on: April 10, 2020, 01:13:00 am »
Quote from: jecowa on April 10, 2020, 12:09:43 am
I tried again on a different computer and got different results. The world generated all the way to 550.

I tried generating the same world to only 424 to see if it matched the other world from the other computer. The map was different.
Spoiler: map (click to show/hide)

The history was different:
Spoiler: world history (click to show/hide)

The maps, world histories, and world sites and pops would match in 2 worlds generated with the same seeds on one computer (even though the save files themselves came out different), but they are different from what the other computer produced. I can't tell that restarting has any difference from not restarting.
Ht abort after the world generates and try again and you'll get the exact same world each time (based on hist figs, dead and events numbers).
Shut down and restart and you'll get a different set of numbers.
Logged

jecowa

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #92 on: April 10, 2020, 01:42:57 am »
I get the same numbers of Hist Figs, Dead, and Events even after relaunching the game. The world has the same name, and there's the same number of rejected.

The other computer has the same world name and number of rejected worlds as the first computer, but the Hist Figs, Dead, and Events are different.
« Last Edit: April 10, 2020, 01:46:58 am by jecowa »
Logged

Shonai_Dweller

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #93 on: April 10, 2020, 01:52:14 am »
Quote from: jecowa on April 10, 2020, 01:42:57 am
I get the same numbers of Hist Figs, Dead, and Events even after relaunching the game. The world has the same name, and there's the same number of rejected.

The other computer has the same world name and number of rejected worlds as the first computer, but the Hist Figs, Dead, and Events are different.
That's weird. Something about my PC? What OS are you using?
So you can recreate your crash every single time then?
Logged

jecowa

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #94 on: April 10, 2020, 02:10:14 am »
Windows 8.1 and Mac OS X Mojave. Using 64-bit builds on both.

It crashes every time on the PC. It generates every time on the Mac.
Logged

Shonai_Dweller

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #95 on: April 10, 2020, 02:17:50 am »
Quote from: jecowa on April 10, 2020, 02:10:14 am
Windows 8.1 and Mac OS X Mojave. Using 64-bit builds on both.

It crashes every time on the PC. It generates every time on the Mac.
And the Mac generates different numbers each time, right?
Guess we need someone else with Windows 8 to confirm if this is an OS specific thing then.
Logged

PatrikLundell

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #96 on: April 10, 2020, 02:37:09 am »
Remember that we have several kinds of world gen issues that need to be dealt with, and Toady would need seeds for all of them (with a preference for the ones that generate quickly):
1. Reproducible world gen crashes.
2. Divergent world gen from the same seeds using the same starting conditions (OS, DF launch state, etc.). Some code apparently uses condition dependent seeds when it should have used world gen seeds.
3. Stable world gen using the same starting conditions, but reliably divergent when conditions are changed in specified ways (DF starting condition and/or OS changes), with no rejects (as it's known that rejects can cause divergence at least between OS variants, and fixing that is probably a harder problem).
4. Crashing in a significant percentage of the cases. Ideally, 1-3 should make this class unnecessary.
Logged

jecowa

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #97 on: April 10, 2020, 02:52:03 am »
Quote from: Shonai_Dweller on April 10, 2020, 02:17:50 am
Quote from: jecowa on April 10, 2020, 02:10:14 am
Windows 8.1 and Mac OS X Mojave. Using 64-bit builds on both.

It crashes every time on the PC. It generates every time on the Mac.
And the Mac generates different numbers each time, right?
Guess we need someone else with Windows 8 to confirm if this is an OS specific thing then.

The Mac generates same numbers each time, but the Mac's numbers are different from the PC's numbers. The PC also generates the same numbers each time, but they are different from the Mac's numbers.

Are you using AMD maybe? Mine are both on different gens of Intel.
Logged

Shonai_Dweller

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #98 on: April 10, 2020, 02:58:31 am »
Quote from: jecowa on April 10, 2020, 02:52:03 am
Quote from: Shonai_Dweller on April 10, 2020, 02:17:50 am
Quote from: jecowa on April 10, 2020, 02:10:14 am
Windows 8.1 and Mac OS X Mojave. Using 64-bit builds on both.

It crashes every time on the PC. It generates every time on the Mac.
And the Mac generates different numbers each time, right?
Guess we need someone else with Windows 8 to confirm if this is an OS specific thing then.

The Mac generates same numbers each time, but the Mac's numbers are different from the PC's numbers. The PC also generates the same numbers each time, but they are different from the Mac's numbers.

Are you using AMD maybe? Mine are both on different gens of Intel.
Intel i7 laptop.
Logged

Iä! RIAKTOR!

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #99 on: April 18, 2020, 05:44:36 am »
How code RAISING_GHOST power to necromancers? I want try, even if this is bugged.
Logged

Shonai_Dweller

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #100 on: April 18, 2020, 05:47:10 am »
Quote from: Iä! RIAKTOR! on April 18, 2020, 05:44:36 am
How code RAISING_GHOST power to necromancers? I want try, even if this is bugged.
Raising ghosts wasn't added to the game this time, so presumably it's not something you can just switch on. Still, ask in the modding forum, there's probably a hacky way of getting the same result.
Logged

PatrikLundell

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #101 on: April 21, 2020, 06:11:00 am »
Theory on cause of huge lists of deity relations:
In fortress mode babies seem to inherit their mothers' set of deities (excluding megabeast and the like), plus one of from the civ's pantheon (which may be one already inherited, resulting in no increase), assuming things haven't changed since 0.40.24. If the same logic applies to historical figures outside of fortress mode, the introduction of the ability of hist figs to acquire new gods through religion would risk exploding the number of gods babies inherit as the growths can be far higher than a bit below one per generation. This assumes acquired deities are inherited in the same way as inherited/acquired at birth ones. Thus, a mechanism that was unbalanced gets pulled further out of check with religions. There is a need for mechanisms to get hist figs to abandon gods, not only pile new ones on top of the pile.

(Also saw a vampire migrant with 30 "deities", with most of those being megabeasts [19?]. It illustrates how deity piling affects long lived creatures as well).
Logged

Iä! RIAKTOR!

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #102 on: April 21, 2020, 06:19:17 am »
Quote from: PatrikLundell on April 21, 2020, 06:11:00 am
Theory on cause of huge lists of deity relations:
In fortress mode babies seem to inherit their mothers' set of deities (excluding megabeast and the like), plus one of from the civ's pantheon (which may be one already inherited, resulting in no increase), assuming things haven't changed since 0.40.24. If the same logic applies to historical figures outside of fortress mode, the introduction of the ability of hist figs to acquire new gods through religion would risk exploding the number of gods babies inherit as the growths can be far higher than a bit below one per generation. This assumes acquired deities are inherited in the same way as inherited/acquired at birth ones. Thus, a mechanism that was unbalanced gets pulled further out of check with religions. There is a need for mechanisms to get hist figs to abandon gods, not only pile new ones on top of the pile.

(Also saw a vampire migrant with 30 "deities", with most of those being megabeasts [19?]. It illustrates how deity piling affects long lived creatures as well).
Interesting! Would you so kindly, give me a screenshot.
Logged

clinodev

  • Bay Watcher
  • Embark Profile Enthusiast, Kitfox & reddit mod.
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #103 on: April 22, 2020, 03:35:56 pm »
I'm also seeing ongoing reports of very long deity lists. This user suggests it might be the result of accepting visitor petitions, (and they're definitely playing with 47.04.)
Logged
Team Bug Fix!

Shonai_Dweller

  • Bay Watcher
    • View Profile
Re: Dwarf Fortress 0.47.04 Released
« Reply #104 on: April 22, 2020, 04:35:09 pm »
Quote from: clinodev on April 22, 2020, 03:35:56 pm
I'm also seeing ongoing reports of very long deity lists. This user suggests it might be the result of accepting visitor petitions, (and they're definitely playing with 47.04.)
That's just a list of gods worshipped in your fortress. Which, yes, of course includes all the visitors so can get pretty long. You don't need a temple to every god, "No specific deity" fulfils normal critter's needs just fine.

The bug is when one dwarf worships far too many gods to ever fulfil its needs. One civilization only has a few gods now, so most dwarves will be OK. But some dwarves have roots in several civs so their potentially worshipped gods expands rapidly.
« Last Edit: April 22, 2020, 04:40:46 pm by Shonai_Dweller »
Logged
Pages: 1 ... 5 6 [7] 8 9 10
« previous next »