01:02:08 <|amethyst> BTW, I'm sure some of you saw this when it was posted to reddit, but: https://i.redd.it/u5o37tjt6z071.png 01:03:11 <|amethyst> or, better, https://www.deviantart.com/kaonedong/art/Menkaure-Prince-of-Dust-880437011 01:04:03 <|amethyst> I'm willing to pay real money if kaonedong is willing to relicense that art (and maybe some of their other ones) such that we can use it as a splash screen 01:05:10 <|amethyst> https://www.deviantart.com/kaonedong lots of DCSS art, all of it quite good 01:07:40 <|amethyst> (also, I don't know if he'll be there this year, because the plague and also plague-related difficulties with CA-US travel; but I'm going to GenCon next week, and if I see Jeph Jacques, I will totally speak to him about commissioning a splash screen 01:07:44 <|amethyst> ) 01:09:53 <|amethyst> Also also, the new person we hired to teach the same class I taught last semester is again offering DCSS as a potential project. He isn't going to link the Discord, so yall don't have to worry about students showing up, but I am still hopeful we can get some useful patches out of the whole affair 01:24:08 <09gammafunk> nice 01:33:23 <|amethyst> was hoping to message JJ from @crawlcode, but apparently he has private messages blocked 01:34:21 <|amethyst> I guess I could publically @ him 01:34:39 Unstable branch on crawl.develz.org updated to: 0.28-a0-440-ge3cd3e9c68 (34) 01:42:47 <|amethyst> Also also also, if anyone else is going to Gen Con Live, hit me up 01:45:51 <10PleasingFungus> before offering money, i’d recommend asking nicely 🙂 01:48:09 <|amethyst> sure, but I feel like, in this post-capitalist society, artists should be rewarded for their art, so even if he wanted to do one for free, I'd still want to pay at least a little 01:52:38 <10PleasingFungus> oh 01:52:48 <10PleasingFungus> i meant kaonedong 01:53:03 <|amethyst> same there :) 01:53:33 <10PleasingFungus> i respect the desire to compensate work 01:53:42 <10PleasingFungus> but many people work on crawl 01:54:00 <10PleasingFungus> programmers, other artists, vaulters etc 01:54:22 <10PleasingFungus> seems strange to single out splash screen artists as the only people who should be paid 01:54:55 <02|amethyst> I mean, I'd pay the rest of you if I could afford it 01:54:56 <10PleasingFungus> unless we want everyone to be paid and for this to no longer be a volunteer project , which is a rather larger proposal 01:55:09 <10PleasingFungus> obviously do what you want 01:55:49 <10PleasingFungus> my feeling is that the ethics of compensation are different for a project which no one is profiting from 01:55:51 Windows builds of master branch on crawl.develz.org updated to: 0.28-a0-440-ge3cd3e9c68 01:56:01 <10PleasingFungus> or perhaps the morals 01:56:12 <10PleasingFungus> i always forget which is which 01:56:59 <02|amethyst> Yeah, but... I'd rather be paying artists and/or devs than CSZO's ISP :) 01:57:31 <02|amethyst> Or, worse, Amazon 01:58:15 <10PleasingFungus> is that a choice you get? 01:58:57 <02|amethyst> Well, hopefully the former won't double-bill me 01:59:46 <02|amethyst> Still a bit salty about that 02:19:52 <10PleasingFungus> reasonable 02:20:06 <10PleasingFungus> anyway, sorry for getting weirdly argumentative 02:20:20 <10PleasingFungus> i approve of art 🙂 02:30:32 <02|amethyst> No worries! I suppose it's in part because I'm capable of writing code, but not of doing art 02:53:56 Monster database of master branch on crawl.develz.org updated to: 0.28-a0-440-ge3cd3e9c68 03:43:22 <02|amethyst> BTW, since they are sort-of kind-of permanent, should Ignis's abilities be on capital letters? 03:43:52 <02|amethyst> I have found myself remapping them such, so that I don't accidentally waste my limited piety 03:45:05 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-4369-gf34790d913 04:01:53 Fork (bcadrencrawl) on crawl.kelbi.org updated to: 0.03-657-g71923310cf 04:38:49 <13Sastreii> As an artist i know the struggle of being recognised as a “regular worker” who needs to be payed for stuff, so i’m agree and i respect your comment. But since DCSS is 100% free to play, i’m offering willingly my support without asking any payment, and i know that there are other artist like me that want to help without making a profit:content: 04:39:39 <13Sastreii> Responding at this comment 04:49:13 <13Sastreii> Anyway, i saw somewere that there’s new wizlabs, were do i find better info of them? 06:22:55 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-4369-gf34790d913 08:53:23 <05kate> i don't know if they have learndb entries or anything yet but the commit has a decent summary: https://github.com/crawl/crawl/commit/bc928eaac75c4052eaff7402bcf43375c27ed236 09:37:43 <13Sastreii> awesome, thanks 10:59:15 <13Sastreii> spells: flame wave and scorch 10:59:16 <13Sastreii> https://cdn.discordapp.com/attachments/747522859361894521/886264634695376926/Schermata_2021-09-11_alle_16.58.14.png 10:59:16 <13Sastreii> https://cdn.discordapp.com/attachments/747522859361894521/886264636440182794/Flame_wave.png 10:59:17 <13Sastreii> https://cdn.discordapp.com/attachments/747522859361894521/886264638059184178/Scorch.png 11:02:31 <10PleasingFungus> ooh 11:15:57 03PleasingFungus02 07* 0.28-a0-441-gb603f484b2: Scorch and Flame Wave tiles (Sastreii) 10(12 seconds ago, 4 files, 4+ 2-) 13https://github.com/crawl/crawl/commit/b603f484b215 11:16:05 <10PleasingFungus> big fan of flame wave especially 11:16:14 <10PleasingFungus> got the good ripples 11:39:54 @PleasingFungus: thank you. 11:39:54 07PleasingFungus02 * 0.20-a0-372-gb9ad934c9e: Remove lava acquirement 10(4 years, 9 months ago, 1 file, 14+ 0-) 13https://github.com/crawl/crawl/commit/b9ad934c9e22 11:39:54 %git b9ad934c9e 11:43:42 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-441-gb603f484b2 (34) 11:44:41 <10PleasingFungus> haha 11:44:43 <10PleasingFungus> no problem 11:46:32 03Perry Fraser02 {GitHub} 07https://github.com/crawl/crawl/pull/2121 * 0.28-a0-422-gb337ad3f7f: Use have_passive(wall jump) for get_god_abilities silence logic 10(88 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/b337ad3f7fae 11:47:20 GitHub magic is neat. 11:47:48 Though the UI for writing a commit message/description for a suggested change is janky. Not used to it. 11:48:17 Oh I broke the build aaa 11:55:25 <09gammafunk> probably should have compiled my suggestion before suggesting it 11:56:42 03Perry Fraser02 07https://github.com/crawl/crawl/pull/2121 * 0.28-a0-423-ga603c6ee75: Add missing include for commit b337ad3f7 10(3 minutes ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/a603c6ee75c2 11:56:49 phew 11:58:24 Modern DevOps UIs honestly blow my mind. It feels like just two years ago that 99% of these features weren't present in /any/ web based repo managers. 12:14:59 does anyone know if the c_message hook gets much use? 12:15:31 imo it's super dangerous as implemented, I don't think it's safe to assume we can run arbitrary user lua code from every place mpr is called 13:03:39 03advil02 07* 0.28-a0-442-g8072b577e9: fix: step up lua throttle aggressiveness 10(79 minutes ago, 5 files, 25+ 3-) 13https://github.com/crawl/crawl/commit/8072b577e904 13:03:39 03advil02 07* 0.28-a0-443-g1af55ca459: fix: don't allow c_message reentry 10(54 minutes ago, 1 file, 11+ 1-) 13https://github.com/crawl/crawl/commit/1af55ca459e0 13:03:39 03advil02 07* 0.28-a0-444-g34660d7429: fix: improve SIGHUP handling in clua code on dgl builds 10(35 minutes ago, 1 file, 3+ 0-) 13https://github.com/crawl/crawl/commit/34660d742963 13:04:20 that last commit should fix the infamous ziggurnaut / AUTOQUIT IS NOT ENABLED 13:06:55 <12ebering> I think mgdelmonte contributed c_message to use it in lua boting 13:08:18 it seems like a useful idea in principle 13:09:29 I'm sort of considering removing it tbh, it's not very easy to implement in a safe way 13:09:59 that is also useful 13:11:22 advil: as long as you're looking at clua stuff 13:11:54 maybe take a look at CMA_LUA_CONSOLE and CLua::load_persist() 13:12:44 it's been my feeling for a while that there are probably some ways to sneak bytecode in there, but i haven't found the time to look deeply at it / write a poc 13:13:00 CMD_LUA_CONSOLE 13:14:16 like, the existing 0x1b checks don't actually cover all the cases where crawl calls luaL_loadbuffer, if i'm reading that right 13:15:26 i started looking at this the other day when y'all linked that weird wizmode game and i was like, hmm 13:15:34 hmm 13:15:36 indeed 13:15:48 this stuff is sort of beyond my lua interpreted knowledge 13:16:03 although tbf everything I just committed was beyond my lua interpreter knowledge as of yesterday 13:16:08 yes, mine too, i'm just half working off that youtube link that all this was based off of with the terrible audio 13:16:24 also there are a bunch of redis CVEs on mitre.org 13:17:54 that might present alternate paths for shenanigans 13:17:54 what youtube link? 13:17:54 and someone named Yongheng Chen also apparently came up with a bunch of lua vulns last summer on the lua-users.org mailing list 13:17:54 some of which apply to 5.1 13:17:54 Unstable branch on crawl.akrasiac.org updated to: 0.28-a0-441-gb603f48 (34) 13:17:54 https://www.youtube.com/watch?v=OSMOTDLrBCQ 13:18:18 http://lua-users.org/lists/lua-l/2020-09/msg00065.html 13:18:36 https://www.corsix.org/lua/bytecode_abuse_0_1.lua 13:18:58 <10PleasingFungus> i wonder if anyone has ever deployed lua 0days against crawl 13:19:08 <10PleasingFungus> would be quite a flex 13:19:12 <10PleasingFungus> as they say 13:20:43 well, presumably that's approximately what that game in question could have involved? 13:20:47 https://www.lua.org/bugs.html 13:20:54 though they didn't seem interested in really deploying it 13:21:13 the most recent security report (for which it appears I never hit publish on the security alert, ugh) came with a fully working poc, it was pretty impressive 13:21:37 youre talking about this one? 13:21:37 https://dpmendenhall.blogspot.com/2020/03/dungeon-crawl-stone-soup.html 13:21:42 or was there another 13:21:48 no, there was another 13:21:59 that one came with a sort of poc that none of us ever got to work 13:22:40 "that one" being the dpmendenhall one 13:24:17 but yeah, PF, my take on it is that crawl lua 0 days are probably live and have probably been live for ~10 years for anyone with expert domain knowledge there 13:26:38 having seen a working one in action, I do think that even if you know of the vulnerability it takes *serious* domain expertise beyond that to produce a lua exploit that actually lets you do anything 13:27:00 yes 13:27:33 in the case of the one that exploited a bug in the __gc metamethod the actual poc was way, way more impressive to me than just finding the bug in the first place 13:28:05 of course it just takes one person who is willing to share it rather than report it, I guess 13:28:54 "rawlins opened on Apr 26, 2020" I guess that timing has something to do with why releasing this got lost 13:29:07 New branch created: pull/2128 (2 commits) 13https://github.com/crawl/crawl/pull/2128 13:29:07 03Aliscans02 07https://github.com/crawl/crawl/pull/2128 * 0.28-a0-440-gdea9020f97: Delete 43 unused functions. Make 1 DEBUG-only. 10(22 hours ago, 54 files, 2+ 726-) 13https://github.com/crawl/crawl/commit/dea9020f972e 13:29:07 03Aliscans02 07https://github.com/crawl/crawl/pull/2128 * 0.28-a0-441-g059e2b9efe: Only compile 27 functions when USE_TILE (or similar) is set. 10(18 hours ago, 9 files, 38+ 10-) 13https://github.com/crawl/crawl/commit/059e2b9efea9 13:38:10 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-444-g34660d7429 (34) 13:44:28 <10PleasingFungus> // Dictionary says: // 1. (12) flesh -- (the soft tissue of the body of a vertebrate: // mainly muscle tissue and fat) // 3. pulp, flesh -- (a soft moist part of a fruit) // yet I exclude sense 3 anyway but include arthropods and molluscs. 13:51:16 04Build failed for 08master @ 34660d74 06https://github.com/crawl/crawl/actions/runs/1224745322 13:52:38 Re PR 2128: I've been tempted to do a few "clean-up" PRs that does a bit of modernization and simplification of older bits of the codebase, but I'm also worried that'd be a bit painful maintenance wise as I assume cherry picking that sort of thing into forks can take a lot more effort for that sort of change. I'm not sure if that fear is really legitimate though. 13:53:15 For an idea of what I'm thinking of, I'm talking about stuff like changing foo.size() == 0 into foo.empty() 13:55:18 <10PleasingFungus> we definitely encourage that sort of thing 13:55:27 <10PleasingFungus> we aren't really concerned about forks 13:55:41 <10PleasingFungus> maintaining one codebase is quite enough for us 14:04:13 <09gammafunk> would be nice to have a robust mod system, and there has been some work to that end (moving various data to yml files) and aidanh outlined an ambitious plan/todo for mod support, but he hasn't been around in a while 14:06:09 <09gammafunk> cpp switch (get_mon_shape(item.mon_type)) { case MON_SHAPE_HUMANOID: case MON_SHAPE_HUMANOID_TAILED: case MON_SHAPE_HUMANOID_WINGED: case MON_SHAPE_HUMANOID_WINGED_TAILED: return TILE_FOOD_BONE_HUMANOID + cs; default: return TILE_FOOD_BONE + cs; } good bone food 14:08:49 i'd be careful about moving more crawl setup stuff into yaml with the intention of "helping" fork authors without being very careful about how you do that... like, i'm pretty sure it's more work to add a species since they were moved to yaml than it was before that 14:09:05 if you're doing fork thing you still tend to end up grepping 14:09:25 but now you have to go through another abstraction layer and figure out how it ties in with the yaml 14:10:22 (maybe actual fork authors have a more informed opinion here) 14:10:26 <10PleasingFungus> ty 14:10:40 <10PleasingFungus> love yummin up those bones 14:13:04 eek, yaml 14:13:09 might be better to do toml? 14:14:54 Also I've seen what can happen if you end up going /heavy/ into programming-as-data methodology and I'm personally not a fan. Key example is Cataclysm DDA... it has 1,096,139 lines of JSON and 412,252 lines of C++. 14:15:41 Basic example: https://github.com/CleverRaven/Cataclysm-DDA/blob/master/data/json/monsters/fungus_zombie.json 14:17:11 <09gammafunk> having monsters etc a data would generally be good; it is way easier to parse various data format files than it is to directly parse C++ to extract that data (for things that aren't the crawl binary) 14:17:51 True—but where it gets bad is in what I'm trying to currently find which :P 14:18:05 <10PleasingFungus> don't think it matters that much 14:18:08 <10PleasingFungus> oh sorry irc 14:18:14 <10PleasingFungus> don't think it matters yaml vs toml in this case 14:18:17 <10PleasingFungus> scale is very small 14:18:36 Oh, small scale then I guess is meh. Toml is way nicer than yaml though :P 14:18:46 <09gammafunk> and yeah I wouldn't presume to know the ideal mod design, but the idea of moving species to data was just that it's generally better to have game data in a parseable format for sake of modification. What you're commenting about might be more "learning curve" for fork/mod people, and I'm sure there are pros and cons there 14:22:12 Ah, here's a moderately questionable use of JSON when I feel it really isn't ideal: https://github.com/CleverRaven/Cataclysm-DDA/blob/master/data/mods/Aftershock/npcs/sadie.json 14:22:22 Although admittedly that's not the worst use of it I've seen 14:22:22 <09gammafunk> a good article about some problems with toml https://hitchdev.com/strictyaml/why-not/toml/ . Just to illustrate that "which is better" is not entirely straightforward and that you can start running into problems as your data scales. I think a big win for yaml right now is just that it's far more popular. I think more people seem to prefer yaml, possibly for a variety of reasons, but it's pretty ubiquitous in certain sectors 14:23:04 There's definitely more anti-yaml than anti-toml. There's some people who are crazy vehemently against it—it's a bit crazy, to be honest. 14:23:11 just see https://hn.algolia.com/?dateRange=all&page=0&prefix=true&query=yaml&sort=byPopularity&type=story 14:23:40 <09gammafunk> well there's probably more of those because there's also not nearly as many pro-toml people either; toml just isn't very widely adopted 14:23:53 Fair point. 14:24:32 <09gammafunk> for certain types of data, txt may be more appropriate than either, for instance our speech and item/ability/spell/etc databases. Maybe those work best as simple text with our custom format 14:24:55 Yeah y'all's yacc work for .des stuff is awesome to be honest 14:25:00 It's a really, really great use of it 14:25:53 <09gammafunk> I'm sure there all sorts of tradeoffs when deciding formats and use cases, but it's hard to see how a C++ source code or header file is an ideal format for working with data. I guess if you have a robust "libcrawl" that other programs can link to, it might take out some of the pain of working with "code as data". But you still have to recompile whenever you chance those C++ files 14:26:29 C++ definitely isn't the best data serialization form out there, lol 14:27:08 I do remember when Gradle was fairly popular and Groovy as a config language took off for a short while. That was kinda neat and it actually worked (... somewhat) well. That's only tangentially related, though. 14:27:33 I know the main issue with YAML is that basically its syntax is as janky and complex and Perl 3 or 4 :P 14:30:10 gammafunk, that anti-toml analysis is pretty interesting, although the writer has a huge conflict of interest. Still seems pretty objective though. 14:31:22 Really, we should write all of our data and configuration in XML/XSLT /s 14:44:11 <06advil> yaml is already a dep and used in various places, it’s widely supported and already has had the work done of ensuring that all servers have it installed, I don’t think adding another sounds like a good idea to me 14:49:21 <09gammafunk> yeah full disclosure, I made all my bots use toml back when I was getting into asynchronous bot stuff since edlothiol used it for the now-defunct webtiles-changes branch as the new server config format. It was definitely perfectly fine and has worked for all three of the dcss relay bots I went on to make, but I'll be moving those over to yaml in the near term as well 14:50:38 <09gammafunk> I guess that'd be another area where it'd be nice to move config out of code; webtiles server itself, but of course that's a whole huge project in of itself that requires a bunch of coordination 14:51:50 advil, I wasn't aware of that context. I can't imagine any reason not to use it in that case :P 14:51:51 <09hellmonk> My personal experience wrt species additions is that the stuff exposed in yaml was the easier portion to figure out in c++ anyway and the hard part is always building whatever custom behavior i wanted. Hard to say what would be better for a new fork developer. 14:52:16 seconding that 14:52:23 <09hellmonk> Yaml is a lot nicer to read though 14:52:51 Maybe not always to write in complex cases though, but not prohibitively so, I suppose. /shrug 14:54:35 <06advil> yeah, the stuff you can do in pure yaml with species in a fork would make for some pretty boring species/forks 14:54:45 <06advil> probably MD is implementable 14:55:13 <06advil> I do think it's a lot easier to maintain though, so it's a net win 14:57:24 <06advil> not entirely sure how much dataifying other stuff is really worth it in roughly https://xkcd.com/1205/ terms, definitely not something I would bother with myself 14:58:54 <06advil> there's an old PR around to dataify backgrounds, spl-data.h and mon-data.h would be two other candidates, maybe the artefact data (which is in a custom txt format) 14:59:04 maybe the stuff in mon-data would make a good target for moving to yaml for fork developers? 14:59:18 there's a lot of stuff in there that's just flags that get reused a lot 14:59:26 but, it's effort for questionable benefit 14:59:56 <09hellmonk> Moving mon-data does sound reasonable, it also sucks to read that giant text block and try to remember what all the numbers are 15:00:09 <06advil> yes, I'm always scrolling to the top for that to look at the header comment 15:00:10 <09gammafunk> it's just not fork developers/mods, it's also just the issue of storing data in code 15:04:04 <06advil> yeah, probably more important to think about it that way 15:04:29 <06advil> if I were worried about making life easier for fork developers I'm not sure generalized quiver would exist 😄 15:05:14 <06advil> actually not sure if anyone's merged that yet 15:05:22 <09gammafunk> I do vaguely wonder if it will be hard at some point to maintain support for forks that are way out of sync with trunk; I guess we'de probably not run into many issues unless we reworked the webtiles client/server protocol in some way? 15:05:31 <06advil> (except monstercrawl ofc) 15:05:57 <09gammafunk> monstercrawl?! that's vapourware! (the joke is that vapour was an old crawl monster) 15:06:04 <06advil> right, that is a potential issue since they all do run with current trunk webtiles 15:06:40 <06advil> but the way things are set up, it's not so dissimilar to just making sure that older versions of crawl still can run from current webtiles 15:07:22 <09gammafunk> oh, that's true; unless we change something and decide to "cut off" support for sufficiently old crawl, or something 15:07:32 <09gammafunk> looking at you, 0.10 15:07:51 <06advil> haha 15:08:05 <06advil> that one would still work if it could build 15:08:24 <09gammafunk> even still is working on cue, as I found out some time ago 15:08:32 <06advil> oh interesting 15:08:37 <09gammafunk> I guess TZer0 got a request for it and had it installed a couple years back 15:09:07 <09gammafunk> !lg * cv=0.10 x=src 15:09:08 <04Cerebot> 247521. [src=cue] leNuup the Skirmisher (L4 HuFi), slain by an orc priest (a +0,+1 orcish hand axe) on D:3 on 2021-09-06 09:36:32, with 212 points after 1947 turns and 0:06:07. 15:09:30 <06advil> probably it's not completely impossible to get it to build, I just couldn't figure it out on cao 15:09:46 <06advil> (and then I accidentally deleted it :content: ) 15:42:09 03Seve Monahan02 07* 0.28-a0-445-g51f8dfeb47: fix: Clarity once again blocks *Rage berserk (Mantis 12640) 10(76 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/51f8dfeb475e 15:47:48 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-445-g51f8dfeb47 (34) 16:47:18 <13Sastreii> Quick question: there’s also a “regular” walking tome as well as the elemental ones? The Learndb is pretty cryptic 16:48:45 <13Sastreii> https://cdn.discordapp.com/attachments/747522859361894521/886352585014009886/Schermata_2021-09-11_alle_22.47.55.png 16:54:11 <10PleasingFungus> dummy monster 16:54:19 <10PleasingFungus> i think it might appear when xom does 'detect monsters'? maybe 16:54:26 <10PleasingFungus> it's extremely unimportant 16:59:42 <10PleasingFungus> it just exists as a 'genus', so the game can say 'these monsters are all related to each other' 17:00:08 <13Sastreii> Good to know 18:23:34 Unstable branch on underhound.eu updated to: 0.28-a0-445-g51f8dfeb47 (34) 23:07:57 03gammafunk02 07* 0.28-a0-446-gb986a403c0: fix: Don't give cloud rings to friendly pan lords 10(4 minutes ago, 5 files, 11+ 9-) 13https://github.com/crawl/crawl/commit/b986a403c025 23:36:35 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-446-gb986a403c0 (34)