02:19:50 03CrawlOdds02 07* 0.35-a0-684-g812a268367: Give gozag gold on rimeblight kills 10(12 minutes ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/812a26836775 03:35:12 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5261-gd9800d219b 03:54:41 03CrawlOdds02 07* 0.35-a0-685-g47b49b7c1c: Add rest_wait_ignore_mp to the options list 10(76 seconds ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/47b49b7c1cc3 05:35:06 <04dracoomega> @Odds While it may have been a bug that Gozag wouldn't give gold for rimeblight kills, I don't feel like blocking the pillar effect on Gozag is the right change either. The leaves_corpse check was mostly to exclude summons and whatnot, and if you notice, it was already possible to get a zombie and a pillar at the same time. The pillar is independently growing out of the body and not transforming it. I am not entirely sure why I 05:35:06 had a leaves_corpse = false; in the pillar effect; maybe I thought it was just better visuals if one assumes that the corpse is lost inside the thing? I don't recall now. But I don't think Gozag and animate dead should have a different opinion about whether this needs a corpse or not, and the rime pillar effect is just an additional thing. 05:53:01 <11Odds> Ah yes, I had very much read this as consuming the corpse and not realised that animate dead was compatible with rimeblight 05:54:45 <11Odds> Tendrils of ice devouring a body does still seem somewhat incompatible with that body turning to gold, though? 05:54:51 <04dracoomega> (Yes, leaves_corpse = false was probably just a visual thing, if I had to guess at my past intent. To avoid saying that the corpse was 'underneath' the pillar) 05:56:30 <04dracoomega> I mean... maybe? Though at the very least, animate dead and Gozag ought to agree about this. 05:57:15 <11Odds> Yeah I agree with that 05:58:25 <04dracoomega> I think I intended 'devour' more in the sense of 'is completely subsumed' rather than 'the flesh has itself turned into ice'. Though the spell description does talk about 'leaving nothing behind'. (Mostly, I guess, I am wondering whether it justifies perhaps-somewhat-unintuitively blocking animate dead, or whether it even matters at that point whether it does so.) 06:00:29 <11Odds> Personally I’d have guessed rimeblight and animate dead couldn’t both use a corpse. But no strong feelings, happy to go either way 06:01:30 <11Odds> Maybe best to leave it compatible and fix gozag, necromancers don’t exactly need a nerf 06:04:01 <11Odds> (If you agree that’s the right way round to fix this, I’ll do that a bit later and be sure to preserve the no-corpse-visuals) 06:04:58 <04dracoomega> I'd probably favor that, yes 08:09:18 03CrawlOdds02 07* 0.35-a0-686-g2af20b573a: Allow pillars of salt from Rimeblight under Gozag 10(10 minutes ago, 6 files, 33+ 26-) 13https://github.com/crawl/crawl/commit/2af20b573a63 08:10:05 <11Odds> Turned out rimeblight kills were also not showing up in morgues, which was maybe rather good for pacifist runs 08:10:19 <08nicolae> can you make a commit to fix commit names, since it's pillars of rime not salt 😛 08:10:42 <11Odds> Oh oops. I got it right in the body 🤦 08:59:26 03CrawlOdds02 07* 0.35-a0-687-g93eb4e48af: Give LRD a no-hostiles-in-range check 10(9 minutes ago, 1 file, 29+ 1-) 13https://github.com/crawl/crawl/commit/93eb4e48af9b 10:34:36 <11Odds> There's a bug where drag attacks which cause the monster to get shafted can cause crashes when we try to apply the brand damage. Perhaps the right thing to do here is finalise attacker movement right at the end of attacks - and if so possibly we should also do this for trampling, which currently gets round this by having the attacker movement being a fineff (which predates movement finalisation - and isn't an option for dragging because 10:34:36 the defender needs the attacker's square) 10:44:26 <04dracoomega> https://ptb.discord.com/channels/735056636644687913/747522859361894521/1466553550304841889 I am reminded of the opinion I stated here, for all the many, many times this sort of thing keeps happening 10:45:06 <04dracoomega> (And a handful of follow-up messages) 10:46:43 <11Odds> Hmmmmm interesting 10:49:23 <04dracoomega> It's mildly spooky, but it also feels like just a fundamentally safer way to operate 10:50:00 <11Odds> This wouldn't move around when anything happened right? Just a blob of memory remains valid for longer pointing to a now-dead monster 10:50:16 <11Odds> And no-one should be relying on the monster to have been cleaned up, I'd think 10:50:27 <04dracoomega> Yeah. (And, as mentioned, we do already query and perform actions on dead-but-not-reset monsters in lots of places) 10:51:36 <04dracoomega> There's even argument to monster::hurt() to defer cleanup for a while, which at present has to be carefully called in certain places to avoid incorrect behavior 10:52:48 <04dracoomega> (And final_effect_monster_cache is used in a bunch of places so that we can still access some of those monsters by fineff time) 10:53:06 <04dracoomega> A lot of which seems like it could be greatly simplified by doing cleanup very-slightly-later 10:53:50 <04dracoomega> (Of course, there may be consequences I've not thought through >.>) 10:54:11 <11Odds> Not sure I understand the monster::hurt bit, is that not a core place that would have to defer cleanup? 10:54:19 <04dracoomega> But the current model makes new instances of these sorts of crashes basically inevitable 10:54:57 <11Odds> Yeah. I like this idea and will give it some more careful thought 10:55:57 <11Odds> (And then do it if I still like it afterwards 🙂 ) 10:56:13 <04dracoomega> Sounds good to me! ^^; 15:43:54 Unstable branch on underhound.eu updated to: 0.35-a0-687-g93eb4e48af (34) 16:02:53 <11Odds> (I do still like this after diving into it, and have a mostly-working version. Still a bunch of testing and thinking about any nasty cases to do) 16:34:52 <04dracoomega> Wow, that was fast 19:25:17 -!- TAS-2012v is now known as TAS_2012v 19:49:55 03gammafunk02 07* 0.35-a0-688-gec49406eb0: Add randart scarves to some vault item sets 10(31 minutes ago, 1 file, 4+ 7-) 13https://github.com/crawl/crawl/commit/ec49406eb076 22:36:05 Unstable branch on crawl.develz.org updated to: 0.35-a0-688-gec49406eb0 (34) 22:59:54 Windows builds of master branch on crawl.develz.org updated to: 0.35-a0-688-gec49406eb0 23:20:20 Unstable branch on cbro.berotato.org updated to: 0.35-a0-688-gec49406eb0 (34) 23:55:55 Monster database of master branch on crawl.develz.org updated to: 0.35-a0-688-gec49406eb0