01:44:37 Windows builds of master branch on crawl.develz.org updated to: 0.26-a0-771-g060e6a62af 02:15:59 Unstable branch on cbro.berotato.org updated to: 0.26-a0-771-g060e6a62af (34) 02:47:28 Monster database of master branch on crawl.develz.org updated to: 0.26-a0-771-g060e6a62af 03:31:01 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-3624-ge566ea1ced 08:47:20 -!- amalloy is now known as amalloy_ 11:02:19 03PleasingFungus02 07* 0.26-a0-772-g3149ed9: Increase Hydra Form duration (Lici) 10(4 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/3149ed920926 11:16:58 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-772-g3149ed9209 (34) 12:06:29 Unstable branch on crawl.akrasiac.org updated to: 0.26-a0-772-g3149ed9 (34) 13:30:11 Lightli: you're gonna have to be more specific, that save looks fine 14:11:22 huh, why does KMASK apply after shuffles but KPROP apply before? 14:12:02 well, not before necessarily, but to the current map 14:13:42 New branch created: pull/1605 (1 commit) 13https://github.com/crawl/crawl/pull/1605 14:13:42 03Nikolai Lavsky02 07https://github.com/crawl/crawl/pull/1605 * 0.26-a0-772-g1ec5cfe: Show monster spell damage on xv in WebTiles 10(10 minutes ago, 2 files, 4+ 15-) 13https://github.com/crawl/crawl/commit/1ec5cfebeeed 14:13:56 and followup question, is there a clean way to set opaque on a square that has a lot of SUBSTs besides just listing them all? 14:54:30 and what are you even supposed to do for explicitly masking e.g. only some floor tiles as opaque? 14:54:37 there really must be some builder syntax I'm missing here 14:58:53 !vault minmay_connected_triangles_in_corner 14:58:53 1/1. https://github.com/crawl/crawl/blob/master/crawl-ref/source/dat/des/variable/large_abstract.des#L703 14:58:53 this is a good example of a vault that should be done with explicit opaque masking 14:58:53 but I'm now very confused about how to do that, or even if that's currently possible 15:03:21 gammafunk any ideas? 15:08:26 hm the documentation for KMASK is also very confusing on this point, but I'm pretty sure from testing this in practice that the true part is "Like KFEAT, KMASK applies to cells that have the specified glyph after applying all shuffles and substitutions." 15:27:53 advil: oh, that may be the case re: the order, but couldn't you simply use a different glyph for the relevant floor tiles? 15:27:57 and then kmask those 15:28:29 and likewise add any item/monster glyphs to the kmask 15:29:13 ah, I see 15:29:54 yeah, pretty common to do that sort of thing in general to conditionally apply stuff to only certain floor tiles 15:29:57 I think I was just missing the basic thing of "`.` is not the only way to make a floor" 15:30:03 ah, yep 15:30:18 any glyph with no default mapping becomes floor 15:30:41 this syntax really needs some sort of flood fill for masks/props 15:30:52 so commonly used ones are: -'"` 15:31:41 by flood fill...how would that work? 15:32:47 for opaque what you mostly want to do is fill an entire zone as opaque, simplest thing to do would be to use the builder's connectivity algorithm 15:34:28 I'm just contemplating whether it would be more tedious to implement that or to manually do this vault I just linked 15:39:49 well, I came up with a way to simplify it in my editor, so that wins for now 15:40:06 I guess I'm wondering how this syntax would work; would it use a single glyph to "seed" the fill? 15:40:13 that's what I had in mind 15:48:16 it certainly could save some time; we manually do fills like this just all the time in vault making, and they do tend to be error prone in terms of missing a glyph 15:48:21 whoa, sublime text column selection is somehow autoselecting . regions, that's handy 15:48:44 but I guess a flood fill could lead to new types of mistakes 15:49:52 yes, though the most common sort of mistake would be equivalent to the current status quo (e.g. if you're filling with `opaque` and you mess up your walls, you'd get the equivalent of not tagging a vault transparent) 15:50:16 !vault kennysheep_water_balloons 15:50:16 1/1. https://github.com/crawl/crawl/blob/master/crawl-ref/source/dat/des/variable/large_abstract.des#L524 15:50:28 why would a vault like this not have an @ by the door in the lower left? 15:50:55 well, doors are defacto @, no? 15:51:12 I'm not sure the builder knows that 15:52:09 If a vault has no entry points defined, random traversable squares on its edge will be chosen and connected, as if they had the @ glyph. 15:52:14 according to syntax.txt 15:52:17 er 15:52:18 sorry 15:52:28 not what I wanted 15:52:29 The + and = glyphs, when placed on the edge of the map, are also considered 15:52:30 interesting 15:52:33 entry points. 15:52:40 I've seen cases where that doesn't appear to be happening 15:52:50 but those might have been vaults that had more than one such square 15:53:44 -!- aidanh_ is now known as aidanh 15:56:23 I've definitely been making vaults assuming it applied to all edge doors 15:56:34 although only one door is obviously the most common case 16:03:49 I think the weird cases I've seen did have more than one potential connection point, which if that doc sentence is correct, would lead to odd things 16:04:03 I wonder if it actually adds the @ glyph or not 16:05:24 for transparent vaults choosing randomly probably works fine (though that makes me worry, is it chosen at placement time or load time...) 16:11:03 03advil02 07* 0.26-a0-773-g80fa9cf: More transparency fixes 10(65 seconds ago, 6 files, 136+ 123-) 13https://github.com/crawl/crawl/commit/80fa9cfebab3 16:15:32 haha oops, I missed 6 .s in one of those minmay maps 16:15:39 luckily they don't really matter 16:17:14 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-773-g80fa9cfeba (34) 17:21:47 03advil02 07* 0.26-a0-774-g1415c89: Fix some missed .s 10(80 seconds ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/1415c89e12c5 17:22:16 Silurio (L12 GnWn) Crash caused by signal #6: Aborted (Lair:3) 17:32:09 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-774-g1415c89e12 (34) 17:55:59 http://crawl.kelbi.org/saves/Lightli-crawl-git-1415c89e12-201108-2254.tar.bz2 17:58:22 duvessa spawned but not dowan 18:25:24 Unstable branch on underhound.eu updated to: 0.26-a0-774-g1415c89e12 (34) 20:29:18 is there still a crawl forks channel or if anyone programs those, they discuss them here? 20:30:33 -!- amalloy_ is now known as amalloy 20:37:23 DarwinElf: We generally prefer discussion to stay on-topic to dcss. There is a forks channel, but it's probably unused, and there are various discords/forums for some specific forks 20:38:08 For another general discussion place, see #dcss-variants on the roguelikes discord 20:39:43 i'm definitely not going to discord just for that. If there were a Usenet forum, or traditional independent web-forum, or IRC channel for forks people here work on, I might go there... 20:42:16 you could certainly make one and advertise it! I think the issue is just that a large majority of people who might work on a fork prefer things other than irc these days 20:43:02 it'll be a long time before I might get around to that; I just like playing forks sometimes. Does it count if people here just add code to their own pulls on git then add it back in? 20:44:47 Not quite sure I understand. You're asking if making your own local changes counts as a fork? Sure, counts as a fork, just sounds like you're not advertising/organising it. 20:46:40 isn't that how a lot of projects work: official team members making their own branches/forks and adding stuff back into the main branch? 20:49:10 Well, sure, but that's describing how they go about testing/developing stuff that's intended to go into the game itself 20:53:58 and is any of that on GitHub/etc., like I could compile & test an beta/development/testing version? 20:54:04 'a' 20:54:30 (was going to say alpha but Dungeon Crawl Stone Soup has been way beyond that for decades) 20:55:44 ??repo 20:55:44 github[1/1]: The site that hosts the crawl git repository at https://github.com/crawl/crawl 20:55:44 that's the official repo, and INSTALL describes how you can compile the sources from a git checkout 20:56:01 many players actually play the current development version, which is called trunk (or sometimes 'git') 20:56:08 excellent; thanks! 20:56:11 np 21:07:50 DarwinElf: There's sometimes a bit of chat in ##crawl 22:50:17 gammafunk just to check, is there a reason why ghost vaults are mostly not transparent? (aside from often being relatively enclosed) 22:50:52 advil: no, I think that's just the default tagging we used for float vaults in general 22:51:00 that is, we tended to not mark them transparent 22:51:03 ok 22:51:11 yeah, that was true 22:51:46 from looking at a lot of vetoes, I kind of think that basically anything that has >1 outer passable square is better off transparent 22:52:18 yeah, I was going to look at your merged commits and try to figure out what the new 'rules' are for vaults 22:52:38 they're still in flux, what I'm doing is just going through many vetoes 22:52:54 your work is mostly trying to make vaults as transparent as possible to make better-connected levels? whereas before they defaulted to opaque and this caused issues? 22:52:59 contemplating seeing what would be involved in changing the defaults 22:52:59 yeah 22:55:33 for example, a relatively common source of veto-causing issues is when say a vault is opaque but has multiple floor tiles somewhere on the outside. If a small zone connects to only one of those spaces, it counts as disconnected, which will usually trigger a veto 22:57:11 !vault ebering_ghost_xom 22:57:12 1/1. https://github.com/crawl/crawl/blob/master/crawl-ref/source/dat/des/variable/ghost.des#L1243 22:58:15 like those four floor tiles there adjacent to the @ -- which look completely innocuous if you don't know that's how the builder works, in a level I was just looking at led to an unnecessary veto 22:58:32 and a lot of vaults have this 23:00:29 so in many cases a single vault with this kind of setup won't have trouble placing, but it definitely adds up 23:03:28 and I'll also add that so far, as long as I catch the cases that do need to be explicitly masked, I haven't really found any downsides to making vaults transparent in terms of placement 23:04:01 what would be nice is an automated test to determine if a vault is missing masking, maybe it can be done 23:04:56 though if we're contemplating a dec release I really need to get back to quivers :D 23:05:14 03Nikolai Lavsky02 {PleasingFungus} 07* 0.26-a0-775-g3b90a50: Show monster spell damage on xv in WebTiles 10(9 hours ago, 2 files, 4+ 15-) 13https://github.com/crawl/crawl/commit/3b90a50a07d3 23:16:07 04Build failed for 08master @ 3b90a50a 06https://github.com/crawl/crawl/actions/runs/353163804 23:17:06 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-775-g3b90a50a07 (34) 23:19:19 03PleasingFungus02 07* 0.26-a0-776-g8b94a72: Checkwhite 10(19 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/8b94a721938f 23:32:06 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-776-g8b94a72193 (34)