Fixed two world gen crashes from earlier in the thread -- Ryan's and charming.snark's ended up being the same cause, and ZM5's were the same one in the three I checked.
Re: linux w.g. crash, dunno how to use a core dump, and I can't do much with Linux directly currently in any case (cannot run the seed, for example), and it'd be difficult to dive into a community effort to construct a more usable virtual computer with the deadlines approaching. Hopefully the upcoming two w.g. crash fixes cover it, though of course that's optimistic.
A few concerns I'm seeing emerge from reddit:
* Serious irregularities in choosing migrant professions with large numbers of particular professions, and I've seen this in my games and on Twitch as well.
Spoiler (click to show/hide)
I wonder if you made an adjustment to migrant professions to highlight the new guilds? It might be excessive, and breaks with the previous fairly reliable expectation that we'd preferentially get migrants from professions we need.
* Some sort of loyalty cascade or other form of internal fighting seems to have become far more common, reported most commonly for the end of werebeast fights and occasionally whole training squads slaughtering each other. It doesn't sound like the "historical spy identities" problem. I haven't personally experienced this one (probably because I'm slow to do military,) but this bug report with save might be promising.
* Gosh, there are a lot of deities/sects in the temple lists now!
Spoiler (click to show/hide)
I'm seeing multiple reports of 50-80 entries. This may be working as intended, but it's definitely going to increase the burden of temple building in medium to large forts.
* Okay, I admit I haven't seen anyone comment about this yet, but since you've just touched this new code, I wonder if we could get stone altars as well as metal ones? Leave the wood ones to the elves, of course.
Not sure about the migrants. The profession attraction code still has desired professions up top, so it'd have to be really out of whack (ie, no available migrants of the right type), which is probably is. Even without those, it should be creating non-historical migrants to meet need before it pulls random historical figures. Maybe those are the increased disconnected people mentioned in the thread.
Yeah, I know there's at least one (and probably more) cascade issue that wasn't a spy. Hopefully I can figure it out.
I should also resolve the god issue this time. It's not intended - an unfortunate consequence of it inefficiently fishing around for death gods in mortal civs.
I'll check the stone altars as well. I agree it's the most sensible material.
Toady, can "civilized" animal people and experiments adopt religions at all? (On a scale large enough to offer an adventurer)?
Presumably if there's a population that's moved to a city, they should just pick up religions slowly like other groups do, in the periodic shrine/culture update. I don't think they have special rules. So yeah, if that's not happening, and the population isn't newly moved, it should be checked out.
I've found that creating a copy of the generated necromancy secrets (with changed names), but designed to raise a modded creature with an experiment interaction, does not raise the modded creature, but does allow them to raise an unrelated generated creature.
Hmm, no immediate idea. It's pulling the e.g. ghoul effect from their usable interaction profile and not a global list, so it should come from the secrets they've learned (or from their metaphysical profile if they have spheres -- in which case it will be random.) Will have to check out the specific setup when it gets on the tracker.