00:44:05 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-5014-gf9597c7f93 03:22:49 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5014-gf9597c7f93 05:29:32 Unstable branch on crawl.akrasiac.org updated to: 0.31-a0-700-gf135852 (34) 07:28:06 noghettoblastrs (L12 DsNe) Crash caused by signal #6: Aborted (Lair:3) 07:43:49 <05icoson> hm for some reason cao is not generating milestones for sprint (logfile looks ok) 07:50:31 <05icoson> ah I see 07:50:35 <05icoson> very confusing 07:50:44 <05icoson> why do we write milestones to the save directory, not the shared directory? 07:50:53 <05icoson> then make servers generate a zillion symlinks 07:57:48 <05icoson> possibly the answer is "someone did it this way 13 years ago when they implemented shared directories and it hasn't changed since" 08:29:13 <05icoson> very tempting to change it to be consistent with logfile locations, I wonder how many servers it risks breaking 08:43:42 <05icoson> milestones for early descent games will probably remain a bit broken unless I go to the trouble of concatenating a bunch of different version-specific milestone files 08:44:04 <05icoson> (also milestones for seeded games unless someone took action specifically to fix this on their server -- but I don't think much is reading these) 08:45:21 <05icoson> definitely gotta do something about this before descent rolls out to other servers though 08:45:46 <05icoson> tempted to just change it, I don't see how this could break any existing setups I'm aware of (cao and cbr2 should be ok) 09:31:57 Palms (L3 BaHs) ERROR in 'mon-util.cc' at line 681: bogus mc (no monster data): invalid monster_type 1000 (1000) (D:2) 09:42:40 Palms (L6 BaHs) ERROR in 'mon-util.cc' at line 681: bogus mc (no monster data): invalid monster_type 1000 (1000) (D:4) 10:03:48 <06pleasingfungus> !crashlog Palms 10:03:52 <04Cerebot> 5. Palms, XL6 BaHs, T:3027 (milestone): http://crawl.akrasiac.org/rawdata/Palms/crash-Palms-20230907-164238.txt 10:04:21 <06pleasingfungus> death... 10:05:24 <06pleasingfungus> wow, jinxbite triggers on block?! 10:06:32 <06pleasingfungus> yeah i don't like anything happening here 10:16:12 <04dracoomega> I... hmm... I admit I didn't test as thoroughly for melee jinxbite. I wonder if different ordering here is letting jinxbite kill something before the melee attack is happening? At least, the stack trace there maybe seems that way? 10:16:40 <04dracoomega> (Also, jinxbite has a chance to trigger on any attack attempt - whether successful or not) 10:17:03 <04dracoomega> I'll take a look 10:17:07 <05icoson> it's not entirely impossible that it's natasha-specific death 10:18:05 <04dracoomega> Nah, it looks like jinxbite specifically is placed before weapon brand triggering and I don't know why 10:18:54 <04dracoomega> Probably just as simple as moving it later on, though I just want to check something with ranged triggers while I'm at it 10:19:21 <06pleasingfungus> i'm fixing it 10:19:46 <04dracoomega> (It seems it's before ranged weapon brand too, actually) 10:19:55 <06pleasingfungus> hold please 10:20:08 <04dracoomega> Sure, I'm not going to push anything if you say you are 10:23:48 03PleasingFungus02 07* 0.31-a0-701-gaea1e32358: Move jinxbite into a fineff 10(2 minutes ago, 6 files, 50+ 44-) 13https://github.com/crawl/crawl/commit/aea1e3235830 10:23:55 <06pleasingfungus> oughta do it 10:24:25 <06pleasingfungus> anyway sorry i was a bit rude earlier 10:24:30 <06pleasingfungus> i am surprised that jinxbite triggers on block 10:24:57 <06pleasingfungus> does it also trigger on miss? 10:25:04 <04dracoomega> Yes, very deliberately 10:25:12 <06pleasingfungus> interesting 10:25:27 <04dracoomega> Since it's an MR check here and is specifically supposed to help more early than late (and early ranged weapons often have pretty piddly accuracy) 10:25:45 <06pleasingfungus> i mean, very few monsters using shields early 10:25:50 <06pleasingfungus> or late, really 10:26:04 <04dracoomega> Thematically, it's not something your weapon hit is doing to them, anyway, so I figured it shouldn't matter what your weapon does exactly 10:26:05 <06pleasingfungus> so the on-block trigger is still a little surprising 10:26:10 <06pleasingfungus> mm 10:26:20 <06pleasingfungus> anyway, excited to see folks play with it 🙂 10:26:59 <04dracoomega> Your code is definitely fancier than I was going to do (which was just moving the statement to a better place). Are there any mechanical consequences to it being a fineff? 10:27:39 <04dracoomega> (I admit I don't really understand them all that well) 10:27:53 <04dracoomega> Mostly through lack of doing anything with them, I suppose, but... 10:35:19 03advil02 07* 0.31-a0-702-gf82c85558f: fix: write milestones to SHAREDDIR, not the save directory 10(6 minutes ago, 1 file, 5+ 4-) 13https://github.com/crawl/crawl/commit/f82c85558fc6 10:37:22 <06pleasingfungus> fineffs trigger after everything else is done, basically 10:37:41 <06pleasingfungus> so the rest of the attack will finish going through, and then, once we're in a tidy "neutral" state, fineffs will start triggering 10:37:53 <06pleasingfungus> the messaging will be in a somewhat different order as a result 10:38:49 <05icoson> who will fineff the fineffs 10:39:21 <06pleasingfungus> who sentinels the hell sentinels 10:39:32 <04dracoomega> I was about to say "Oh, it doesn't trigger if the attack kills the target" but while part of my brain felt it mattered that it still did, that part of my brain is evidently back in past world where it spawned a sprite that could hit something else ^^; 10:39:49 04Build failed for 08master @ aea1e323 06https://github.com/crawl/crawl/actions/runs/6112883235 10:40:37 <06pleasingfungus> Error: Restoring cache failed: Error: The process '/bin/bash' failed with exit code 1 10:40:45 <06pleasingfungus> ci... 10:41:06 <06pleasingfungus> anyway the pre-fineff code didn't trigger on enemies killed by the attack either 10:41:18 <06pleasingfungus> (otherwise this would've crashed quite a bit more 😛 ) 10:42:15 <05icoson> it looks like that ci error is a homebrew failure on updating php 10:42:24 <05icoson> I will ... maybe just hope it goes away soon 10:42:28 <06pleasingfungus> into it 10:43:38 <04dracoomega> No, I'm sure it didn't. It was just a blip back to when it deliberately did, since it, well, didn't actually do anything to the dead enemy anyway, so it was fine 10:43:45 <04dracoomega> At that point in time 10:44:33 Unstable branch on crawl.kelbi.org updated to: 0.31-a0-701-gaea1e32358 (34) 10:49:25 <05icoson> it went away 10:49:58 <05icoson> !lm cko x=vlong 10:49:59 <04Cerebot> No milestones for cko. 10:50:09 <05icoson> !lm * cko x=vlong 10:50:15 <04Cerebot> 8557227. 2023-09-07 17:49:11 MrMatt the Fighter (L15 GnFE of Sif Muna) left the Elven Halls on turn 26101. (Elf:1) 10:50:45 Unstable branch on crawl.kelbi.org updated to: 0.31-a0-702-gf82c85558f (34) 10:51:50 <05icoson> !lm adviltest cko x=vlong 10:51:51 <04Cerebot> 50. 2021-12-10 20:16:09 adviltest the Skirmisher (L3 HOFi) killed Jessica on turn 1423. (D:2) 10:54:30 <05icoson> !lm adviltest cko x=vlong 10:54:31 <04Cerebot> 51. 2023-09-07 17:51:27 adviltest the Trooper (L1 HOFi) began the quest for the Orb on turn 0. (D:1) 10:54:32 <05icoson> there we go 11:25:54 Palms (L6 BaHs) ERROR in 'mon-util.cc' at line 681: bogus mc (no monster data): invalid monster_type 1000 (1000) (D:5) 11:49:44 <04dracoomega> I'm not very familiar with vault code myself, but should it be considered a bug that monsters with a weight of 0 will generate as if they have a weight of 10 instead? (For situations where, say, the weight is dynamically calculated based on depth) 11:58:10 <06pleasingfungus> sounds dubious 12:01:57 <09gammafunk> I don't quite understand the situation, is the vault supplying the weight? And via what des statement? 12:02:39 <04dracoomega> Like, the code is pretty clear about this, even if I'm not sure I agree with the reasoning: cpp mspec.genweight = find_weight(mon_str); if (mspec.genweight == TAG_UNFOUND || mspec.genweight <= 0) mspec.genweight = 10; 12:03:01 <04dracoomega> Whereas I would assume the intuitive thing is that a monster set to weight 0 would not spawn at all, instead of spawn at weight 10 12:03:15 <04dracoomega> Since it already checks for no defined weight 12:05:09 <09gammafunk> I see vault at the map resolution level. Agree it should probably use 0 there, but don't know full context. Feels like an assert or at least error may be warranted for negative weight 12:05:45 <04dracoomega> It come up with something index was working on that involved scaling weights of certain monsters dynamically based on the depth a vault was placed at 12:06:04 <04dracoomega> A relevant snippet she sent me: {{ -- Three tiers scaling alongside depth: D:7-10, D:9-13, D:12-15. local d = math.max(21 - you.depth() * 2, 0) local e = 0 if you.depth() > 8 and you.depth() < 14 then local e = 4 end local f = math.max(-22 + you.depth() * 2, 0) mons("howler monkey w:" .. d .. " / jelly w:" .. d .. " / sky beast w:" .. d .. " / vampire mosquito w:" .. e .. " / ogre w:" .. e .. " / 12:06:04 wyvern w:" .. f .. " / basilisk w:" ..f) mons("wyvern w:" .. d .. " / two-headed ogre w:" .. e .. " / shadow w:" .. f .. "/ hornet w:" .. f) mons("water moccasin w:" .. d .. " / ugly thing w:" .. e .. " / large slime creature w:" .. f) }} 12:06:38 <04dracoomega> But if the math puts out a weight of 0, with the intention of 'do not spawn this', it counts as 10 instead 12:07:01 <04dracoomega> (It's a WIP) 12:07:41 <09gammafunk> Yeah seems no way this is desirable behavior 12:08:04 <04dracoomega> Hard to easily tell if any existing vaults are affected by this somehow, but it seems plausible some are 12:08:36 <09gammafunk> Callers can avoid complicated statement cleanup to remove 0 weight entries if it just respects 0 weight 12:09:09 <04dracoomega> Sure, I mean "I wonder if anyone else tried this and has had vaults spawning things in places they didn't expect already?" 12:09:47 <09gammafunk> unfortunately a lot of Lua mons code to inspect to determine that 12:09:54 <04dracoomega> Yeah, very non-trivial 12:10:52 <09gammafunk> Would say that error advisable for negative weight but maybe that's awkward somehow 12:11:39 <04dracoomega> It feels like it maybe should be an error, but reluctant to assert on it given that it would be invisible onto now that it might have happened. I wonder if negatives also counting as zero is fine enough? 12:22:12 <04dracoomega> Well, testing it out with a minor code change, telling weights of 0 to actually be 0 seems to work properly, without issue 12:22:25 <04dracoomega> (Index tells me that negative weights already show an error elsewhere entirely, apparently) 12:24:44 <04dracoomega> So, do you think it's fine to push a simple change to let weight 0 be 0 here? 12:27:11 Eli (L12 TeCj) ASSERT(!invalid_monster(&mons)) in 'mon-death.cc' at line 1637 failed. (D:14) 12:27:52 <04dracoomega> !crashlog Eli 12:27:54 <04Cerebot> 4. Eli, XL23 DsFE, T:85396 (milestone): https://crawl.kelbi.org/crawl/morgue/Eli/crash-Eli-20220902-021555.txt 12:28:16 <04dracoomega> Oh, geez, 0.29, haha 12:28:43 <04dracoomega> Wait, that's not even the right character 12:32:19 it takes a couple minutes for sequell to catch up with the crash milestone 12:33:22 !crashlog eli 12:33:23 5. Eli, XL12 TeCj, T:14827 (milestone): https://cbro.berotato.org/morgue/Eli/crash-Eli-20230907-192709.txt 12:39:19 <04dracoomega> Still not trunk, though, at least 14:28:44 -!- elliptic_ is now known as elliptic 15:26:03 Unstable branch on underhound.eu updated to: 0.31-a0-702-gf82c85558f (34) 16:51:28 New branch created: pull/3286 (1 commit) 13https://github.com/crawl/crawl/pull/3286 16:51:28 03Monkooky02 07https://github.com/crawl/crawl/pull/3286 * 0.31-a0-703-ga9002e1291: Rollpage heal 10(87 seconds ago, 4 files, 16+ 1-) 13https://github.com/crawl/crawl/commit/a9002e12915f 20:01:04 Palms (L7 KoHs) ERROR in 'mon-util.cc' at line 681: bogus mc (no monster data): invalid monster_type 1000 (1000) (D:5) 20:51:03 LexAckson (L14 GnSh) Crash caused by signal #11: Segmentation fault (D:12) 22:35:11 Unstable branch on crawl.develz.org updated to: 0.31-a0-702-gf82c85558f (34) 22:55:55 New branch created: pull/3288 (1 commit) 13https://github.com/crawl/crawl/pull/3288 22:55:55 03RypoFalem02 {GitHub} 07https://github.com/crawl/crawl/pull/3288 * 0.31-a0-703-g54c9e4a2f7: Correct a typo (credit to fbwer!) 10(4 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/54c9e4a2f704 22:57:30 Windows builds of master branch on crawl.develz.org updated to: 0.31-a0-702-gf82c85558f 23:26:32 Unstable branch on cbro.berotato.org updated to: 0.31-a0-702-gf82c85558f (34) 23:55:30 Monster database of master branch on crawl.develz.org updated to: 0.31-a0-702-gf82c85558f