00:44:37 -!- dtsundere is now known as dtsund 00:53:53 !source layout_geoelf_castle.des 00:53:53 johnstein: You have 1 message. Use !messages to read it. 00:53:54 https://github.com/crawl/crawl/blob/master/crawl-ref/source/dat/des/builder/layout_geoelf_castle.des 00:53:57 -!- amalloy_ is now known as amalloy 00:58:09 hm. when I edit variables in the top part here, where it says the variable are global, they don't seem to change anything in the layout part below. https://github.com/jmbjr/crawl/blob/cryptic-layout/crawl-ref/source/dat/des/builder/layout_cryptic_divisions.des 00:58:45 specifically, if I change DEBUG_CASTLE_DESIGN = true, I don't get any of the print statements (which are now crawl.mpr statements in my current version) to display 00:59:19 but if I add DEBUG_CASTLE_DESIGN = true to the layout part of the code, down here, https://github.com/jmbjr/crawl/blob/cryptic-layout/crawl-ref/source/dat/des/builder/layout_cryptic_divisions.des#L106, then it works 00:59:43 I find it hard to believe that the code at the top isn't actually doing anything 01:08:01 -!- n1 is now known as n1k 01:08:18 -!- n1k is now known as n1 01:44:02 it is, *but* iirc it's only read once. possibly during des cache creation or some such. the globals are weird 01:45:12 somewhere in the documentation there's an overview of how it works and what the restrictions are on the global section 01:45:36 hm 01:45:57 you'd think after a recompile it would update. but maybe since it's a *.des I technically don't need to recompile 01:46:45 it does when recompiling under dgl because the caches are per version+commit hash. not necessarily for normal dev setups 01:47:56 (actually I think it'd be more correct to say building the cache collects all the globals together, and afterward changing the file without forcing a des cache rebuild won't update the collected globals, or some such) 01:48:30 because it's the crawl exe and not make that does the cache rebuild and it's not smart enough to check all of that, or something 01:49:12 that gives me a good lead. ty 01:57:02 Windows builds of master branch on crawl.develz.org updated to: 0.21-a0-113-g6841d05 02:34:57 where is the cache? 02:51:30 ah saves/des 02:53:14 Monster database of master branch on crawl.develz.org updated to: 0.21-a0-113-g6841d05 03:08:31 so, deleting saves/des, editing the layout files 'global prelude', then recompiling and starting a new game (generating new saves/des cache) still doesn't seem to register the change in the variables 03:11:36 Unstable branch on crawl.beRotato.org updated to: 0.21-a0-113-g6841d05 (34) 03:17:30 i.e. I still can't set DEBUG_CASTLE_DESIGN = true in the global prelude and have it stay true in the rest of the layout: https://github.com/jmbjr/crawl/commit/af4408fa735b1fc7321054ac738b69f38d8ae3c4 03:26:52 automatic targeting after use of LRD assumes the wall to be the target 13https://crawl.develz.org/mantis/view.php?id=11143 by Majang 04:18:49 Teleport in Swamp ends in forest clearing 13https://crawl.develz.org/mantis/view.php?id=11144 by Majang 04:18:52 imo remove lua 04:20:20 -!- amalloy is now known as amalloy_ 04:23:43 yes 04:29:12 automatic targeting after use of LRD assumes the wall to be the target 13https://crawl.develz.org/mantis/view.php?id=11145 by Majang 05:27:19 -!- amalloy_ is now known as amalloy 06:49:24 another targeting problem 13https://crawl.develz.org/mantis/view.php?id=11146 by Majang 08:24:56 johnstein: I tried to place layout_cryptic_divisions_a from that file and I get a failure afaict before any lua is run 08:25:23 I only copied that one des file from your repo, though 08:28:12 <|amethyst> johnstein: layout_geoelf_castle has the same variable 08:28:28 <|amethyst> johnstein: and since it's global... 08:28:59 ah good point 08:56:24 johnstein: can confirm that |amethyst's suggestion is what's going on, I'm guessing those preambles get run in alphabetical order 08:56:39 not entirely sure when 08:57:34 but I do get your debug messages when I set that var to true in layout_geoelf_castle.des, so I'd either do that or (better, if you're going to leave them in) change the variable name 09:14:22 also, this will be an issue with every variable in that preamble 09:16:30 maybe do something like what the rest of geoelf does with `geoelf = {}` in dlua/layout/geoelf.lua 09:16:41 not sure why castle isn't putting all those constants in that namespace 09:36:40 oooooh 09:36:54 for some reason I thought it was global wrt to all vaults in that single des file 09:37:06 not global global 09:37:27 you know, local global. 09:40:06 I'm still not sure if my approach here makes sense long term. seems like the logical thing is to separate out the "surround primary vault with a bunch of rooms" logic into a common routine. then modify geoelf to call that. but I've been having a tough time figuring out how to refactor it since layout stuff hurts my brain 09:40:18 thanks |amethyst and advil 10:17:07 -!- buki_ is now known as buki 10:58:34 ??rebuild[2 10:58:34 rebuild[2/2]: Trunk update start times in UTC (second number during summer time): CAO: 1700; CDO: 0600/0500; CUE: 2300/2200; CBRO: 0800/0700; CXC: 0500/0400; CPO: every 15 minutes; CJR: every hour 10:59:32 checkers: stupid question, does CPO pull stable versions on the same cycle (every 15 minutes) as it does trunk versions? 11:48:56 if so, I'm pretty sure that that's why we see this error on CPO so much 11:49:34 there appears to be a single shared des cache and there's all sorts of things that can trigger a cache rebuild, which can't happen without locking the files 11:50:40 it's fast enough per file that it's still rare, but every 15 minutes provides a lot of opportunities, at least when the branch is getting updated 14:01:27 03advil02 07* 0.21-a0-114-g8b85549: Pull a tiles dependency out of player.h 10(5 days ago, 10 files, 52+ 57-) 13https://github.com/crawl/crawl/commit/8b8554994a74 14:01:27 03advil02 07* 0.21-a0-115-g8779768: Factor tile-related #ifdefs out of AppHdr.h 10(5 days ago, 45 files, 102+ 81-) 13https://github.com/crawl/crawl/commit/8779768f0243 14:01:27 03advil02 07* 0.21-a0-116-gd2ad500: Remove some tiles-related ifdefs from state.h 10(4 days ago, 2 files, 3+ 6-) 13https://github.com/crawl/crawl/commit/d2ad50005d9b 14:01:27 03advil02 07* 0.21-a0-117-g9766c24: Remove an ifdef in message.h 10(4 days ago, 2 files, 5+ 4-) 13https://github.com/crawl/crawl/commit/9766c24e7203 14:01:27 03advil02 07* 0.21-a0-118-g66b5d33: Relocate an ifdef'd function from view.h to tiles-build-specific.h 10(4 days ago, 8 files, 14+ 6-) 13https://github.com/crawl/crawl/commit/66b5d3384292 14:01:27 03advil02 07* 0.21-a0-119-g868a11d: Remove an old function definition 10(4 days ago, 1 file, 0+ 4-) 13https://github.com/crawl/crawl/commit/868a11d9fde6 14:01:27 03advil02 07* 0.21-a0-120-g7f6a83f: Another batch of removing/relocating ifdefs from .h files 10(4 days ago, 14 files, 41+ 31-) 13https://github.com/crawl/crawl/commit/7f6a83f035ce 14:01:27 03advil02 07* 0.21-a0-121-g4dad355: Use a different strategy for `flash_monster_colour` 10(4 days ago, 6 files, 9+ 13-) 13https://github.com/crawl/crawl/commit/4dad3554f68c 14:01:27 03advil02 07* 0.21-a0-122-g43b3ea6: An attempt to update the MSVC project 10(4 days ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/43b3ea631c2e 14:13:40 wonder if that will screw up the android build somehow 14:23:39 not sure the android build was working even before this 14:23:52 there've been a number of reports of it blowing up 14:24:20 I'm not sure about working, but it apparently builds 14:24:34 every time I do something to break it the people doing it pop in here 14:24:52 nightly builds at http://crawl.newtzie.com/ 14:25:26 it's the same developer(s) who do the webtiles app 15:09:18 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-122-g43b3ea6 (34) 15:47:42 -!- crate_ is now known as crate 16:24:34 -!- Amnes1ac is now known as Amnesiac 18:15:14 Unstable branch on underhound.eu updated to: 0.21-a0-122-g43b3ea6 (34) 18:35:26 -!- wHATEver is now known as Guest3721 21:48:36 |amethyst, how do crawl servers know which des cache folder to use? there's multiple in a given crawl-version/saves/ folder 21:59:28 -!- steven_ is now known as stebl 22:29:25 they should be keyed by commit hash 22:29:30 and/or version 22:50:56 johnstein: i'm not sure if you care about this or want to think about it at all, since it's kind of an annoying problem to deal with, but: 22:51:15 !lg nofunallowed month s=src,char 22:51:15 537 games for nofunallowed (month): 537x cbro (536x CeCK, CeFE) 22:51:29 they're running a copy of qw to try to make the worst ghosts possible it seems 22:51:36 !hs nofunallowed month 22:51:37 537. NOFUNALLOWED the Bludgeoner (L15 CeCK of Xom), succumbed to a sea snake's poison on Shoals:2 on 2017-07-12 20:18:09, with 105081 points after 29506 turns and 0:26:17. 22:52:11 !lg nofunallowed month s=xl 22:52:12 537 games for nofunallowed (month): 100x 4, 91x 5, 77x 3, 59x 6, 44x 7, 42x 2, 35x 8, 32x 9, 25x 10, 14x 1, 7x 11, 4x 12, 3x 13, 2x 14, 2x 15 22:52:21 from what I can tell it's solely for griefing purposes (fast moving species, chaos melee) 22:52:35 That's pretty awesome 22:53:20 obviously it's a pain to try to police anything like this, and they could just create a new account unless they got some kind of ip block I guess 22:53:29 !gkills cbro 22:53:30 No games for * (ikiller~~^(cbro)'s?_ghost). 22:53:43 you know what the solution is, gfunk 22:53:58 remove accounts of all players who have posted on 4chan, I know 22:54:02 sorry nikheizen, we'll miss you 22:54:31 !lg * cbro cikiller=player_ghost s=ikiller 22:54:34 30973 games for * (cbro cikiller=player_ghost): 620x gw's ghost, 172x Phlounder's ghost, 171x sugabear's ghost, 156x antares's ghost, 151x gressup's ghost, 143x ProzacElf's ghost, 142x MetaCosm's ghost, 132x grandjackal's ghost, 131x Iamsock's ghost, 127x LogicNinja's ghost, 124x daryashkoh's ghost, 124x tedric's ghost, 116x apoch's ghost, 114x Buffagor's ghost, 105x w00t's ghost, 102x qqqbard's g... 22:54:41 advil: it doesn't ever, I update by hand when something is tagged 22:54:45 advil: only trunk auto-updates 22:54:45 !lg * cbro cikiller=player_ghost year s=ikiller 22:54:48 18106 games for * (cbro cikiller=player_ghost year): 137x antares's ghost, 110x grandjackal's ghost, 105x w00t's ghost, 100x CanOfBees's ghost, 93x Elysium's ghost, 92x elan's ghost, 92x bande's ghost, 84x Buffagor's ghost, 79x HeyBub's ghost, 75x cheese's ghost, 70x Psyclone's ghost, 68x rogsnak's ghost, 68x JeffiDaFish's ghost, 68x NOFUNALLOWED's ghost, 67x aTarkinC's ghost, 66x daryashkoh's gho... 22:54:56 !lg * cbro cikiller=player_ghost month s=ikiller 22:54:57 1185 games for * (cbro cikiller=player_ghost month): 68x NOFUNALLOWED's ghost, 30x sscrono2's ghost, 25x azarias's ghost, 24x mythresam's ghost, 19x grandjackal's ghost, 19x Vizrt's ghost, 17x VonBacon's ghost, 16x bande's ghost, 15x JohnnyMnemonic's ghost, 14x Parsimony's ghost, 13x octopodefightclub's ghost, 12x Potronius's ghost, 12x charan's ghost, 11x sambo's ghost, 11x Dharilos's ghost, 10x ... 22:55:17 I bring it up because dynast was complaining about fighting tons of these 22:55:18 seems to me like the solution to "some ghosts are complete bs to deal with" is to rework/remove ghosts 22:55:32 no I'm sure that my 4chan solution would work 22:55:34 advil: the des cache is different for each build of trunk, the version hash it baked into the pathname 22:56:02 gammafunk's 4chan-posting detector-linker is _very_ robust. 22:56:11 !lg * cpo cikiller=player_ghost month s=ikiller 22:56:12 211 games for * (cpo cikiller=player_ghost month): 18x zilverxoul's ghost, 15x enarc13's ghost, 13x octopodefightclub's ghost, 12x ultrashaft's ghost, 10x RebelPilot31's ghost, 9x MatthieuS's ghost, 9x NoxCultor's ghost, 8x Caerwyn's ghost, 7x aesoa's ghost, 7x Llamajah's ghost, 7x Spazmo's ghost, 6x Heartnet's ghost, 5x Kaishin's ghost, 5x Raine357's ghost, 5x loopi's ghost, 5x Duelies's ghost, 4... 22:56:26 octopodefightclub 22:56:30 that is 22:56:32 !lg octopodefightclub s=char 22:56:32 what if I posted on 4chan once, about 9 years ago 22:56:33 363 games for octopodefightclub: 192x OpMo, 168x OpGl, 3x OpHu 22:56:33 a really good name 22:56:41 do I need to find a new server 22:56:44 hellmonk: done for 22:56:47 rip 22:56:52 sorry hellmonk, no exceptions 22:56:56 and rip spicerack as well I guess 22:56:59 notlikethis feelsbadman 23:04:45 ghost rework idea: decide which archetype the dead character fit into (BATTLEMAGE/mibe/etc.) and generate a monster version of that 23:04:59 kind of like the hep friend 23:06:30 yeah that's what I think i would do if I weren't outright removing bones 23:09:44 what kind of archetype of octopodefightclub fit into? 23:09:49 s/of/would/ 23:10:02 free exp 23:10:17 some ghosts just generate a potion of experience 23:10:28 ooh 23:10:31 even better 23:10:38 a toadstool of experience 23:10:43 you have to destroy it to get the xp 23:11:44 yes!!! 23:12:01 ghost categories: turbo murdermachine, unkillable wall, tediumbot, toadstool of experience 23:12:46 only catch is if you worship weedl0rd the dank you can't kill the toadstool without getting penance 23:13:46 ty for referencing my cyc thankbait post, but weedl0rd doesnt give a fuck about killing plants 23:14:29 yeah i mean think of all those pipelines he authorized 23:14:43 well if we ruined your idea by adding it to crawl we'd make sure to add new annoying conducts like that 23:14:49 <|amethyst> erm 23:14:49 reee 23:14:50 also real talk can we pretty please get a fedhas rework into 0.21 23:15:01 was weedlord a serious rework proposal? 23:15:09 I told someone that it was 23:15:15 [22:13] ty for referencing my cyc thankbait post 23:15:19 sans the meme parts 23:15:32 well I just mean I don't think anyone has proposed any rework of fedhas recently then 23:15:33 <|amethyst> gammafunk: how are you supposed to smoke a plant without it dying? 23:15:48 hrm, what if you didn't smoke it all? 23:15:55 and left the roots intact 23:15:59 it was a semi-legit attempt to make use of that one guy's idea to lig you when you use fedhas abilities 23:16:21 hence why its basically just fedhas except weaker and you lig yourself 23:18:33 mostly it was a joke post though 23:24:23 "I'm only serious if you unironically liked it, otherwise it definitely wasn't serious" 23:26:52 it works on multiple levels, sry if it was too deep in the meta-irony hole 23:28:58 please can I get all yall to smash that mother**** like button, come on ma homies....like and subscribe 23:29:44 agree 23:32:55 just implement all the griggy abilities instead of what fedhas has 23:33:47 hey gammafunk i have good new and bad new, would you like the good new or the bad news first 23:33:59 +sss 23:34:25 that is impressive typo consistency 23:34:47 can I get both at once 23:35:10 was griggy the thing that became wudzu (rip) 23:35:15 yes, because it's the same news! I have more vaults for you to review, and some of them are p.good and some of them are p.bad 23:35:31 wudzu was bad but I would unironically support replacing fedhas with wudzu 23:35:40 oh, that's fine; I am behind on reviewing vaults, just like I'm behind on doing some other things I said I'd do 23:35:44 but more are always welcome 23:35:54 this travel code, man 23:36:28 every time I think I've understood enough of it to do what I want, something else pops up 23:36:42 super in need of some major refactoring 23:37:55 New branch created: pull/569 (1 commit) 13https://github.com/crawl/crawl/pull/569 23:37:55 03Alex Jurkiewicz02 07https://github.com/crawl/crawl/pull/569 * 0.21-a0-95-gc174bda: Add new vaults and a serial 10(8 days ago, 3 files, 333+ 0-) 13https://github.com/crawl/crawl/commit/c174bdaed3bd 23:46:01 hellmonk, griggy was my god suggestion and then morganleah coded it up and changed the name to wudzu 23:46:14 ok 23:46:15 I didn't think the god was good enough in the OP to be a god 23:46:40 I agree but it would still be better than fedhas 23:46:49 !lg . wudzu 23:46:50 1. Hellmonk the Apex Predator (L27 DsCj of Wudzu), escaped with the Orb and 15 runes on 2016-10-06 04:23:52, with 13513248 points after 118108 turns and 11:55:41. 23:47:45 ok i think i'll turn nofunallowed off now that it's been noticed 23:48:21 apologies to any promising runs it may have ruined 23:49:21 remove ghosts etc. etc.