01:31:47 -!- khadgar0 is now known as khadgar 01:57:39 <06mumra> I think I found a pathfinding bug now ... seems monsters are unable to pathfind through closed doors even if they are able to open them. (They can open doors if they can otherwise "go straight" to their destination, just as soon as the pathfinding logic kicks in they forget how to open them...) 01:58:58 <06mumra> And it's been this way since probably this commit: https://github.com/crawl/crawl/commit/323ad4e9fb1dbb588b479de76b9c62a829cc0646 01:59:43 <06mumra> (I think it's a few different incremental changes that added up to this behaviour so it's a bit hard to trace) 02:40:29 <04dracoomega> I was actually aware of this (though hadn't really dug into it since pathfinding code in Crawl is kind of ugly and hard to read). It became fairly obvious when coding/testing Blood for Blood and seeing a pileup of summoned allies on the other side of doors after opening them 02:41:27 <04dracoomega> Though I wasn't quite sure exactly how far the bug's effects were 02:42:13 <04dracoomega> (It also means that you can't get apostle challenges if the only spaces at the appropriate distance from the player have closed doors in the way, though I think this is hard to intentionally exploit 02:43:29 <04dracoomega> Due to a combination of the randomness in when said challenges can arrive, and how if it can't place one it just defers it a short distance into the future to try again later 02:46:15 <04dracoomega> (I admit, I had mostly forgotten about it again >.>) 03:32:44 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5208-geafff8c3b6 05:07:07 Unstable branch on crawl.akrasiac.org updated to: 0.32-a0-1188-gf657128 (34) 05:16:03 <06mumra> I think the fix is rather easy, these lines can just be removed: if (cell_is_runed(p)) return false; if (!mons->is_habitable(p)) return false; From here: https://github.com/crawl/crawl/blob/f6571284fb03e6208df6a58b60d291a8f15e699f/crawl-ref/source/mon-pathfind.cc#L435 Those two checks are already repeated in mons_can_traverse which is called just after, but mons_can_traverse first checks for door usage and 05:16:04 returns true before the is_habitable check (which returns false for closed doors). These checks should have been removed when clinging was removed, from the history it seems they were copied there supposedly to fix pathfinding bugs caused by clinging, but inadvertently broke doors in the process... 05:18:01 <06mumra> (I think is_habitable maybe at one time was returning true for closed doors because they didn't always have FFT_SOLID, which is maybe how the bug crept in through the side door... 05:25:59 <06mumra> Was digging into pathfinding code trying to get my wall monsters to actually stick to the walls a bit rather than always preferring floor, and the redundant checks jumped out as being a bit iffy 06:23:34 <09hellmonk> glad somebody found the source for that bug lol 06:25:55 03dolorous02 07* 0.32-a0-1189-g249f2362e2: Cantaloupe -> rock melon (jejorda2). 10(4 minutes ago, 2 files, 3+ 3-) 13https://github.com/crawl/crawl/commit/249f2362e2ce 06:30:32 <06mumra> The actual pathfinding code is not that terrible, the really eye burning stuff is handle_monster_move and everything connected there 06:34:28 03dolorous02 07* 0.32-a0-1190-gf786d4924a: Rock melon -> rockmelon (oops). 10(2 minutes ago, 2 files, 2+ 2-) 13https://github.com/crawl/crawl/commit/f786d4924a18 06:34:28 03dolorous02 07* 0.32-a0-1191-gf3498c87cc: Add another thing name to artefacts. 10(6 minutes ago, 1 file, 2+ 0-) 13https://github.com/crawl/crawl/commit/f3498c87cc09 06:39:48 03dolorous02 07* 0.32-a0-1192-gc771b14f0e: Add another Qazlal Donald line. 10(2 minutes ago, 1 file, 2+ 0-) 13https://github.com/crawl/crawl/commit/c771b14f0ef8 07:14:58 03dolorous02 07* 0.32-a0-1193-g0f580fe372: Add more colour names to Blorkula's dialogue. 10(25 minutes ago, 1 file, 60+ 0-) 13https://github.com/crawl/crawl/commit/0f580fe372f8 09:07:59 03dolorous02 07* 0.32-a0-1194-g39d9cdbf94: Add another fruit name to artefacts. 10(37 minutes ago, 1 file, 2+ 0-) 13https://github.com/crawl/crawl/commit/39d9cdbf9449 09:07:59 03dolorous02 07* 0.32-a0-1195-ga64f6baaea: Add still more colour names to Blorkula's dialogue. 10(5 minutes ago, 1 file, 96+ 0-) 13https://github.com/crawl/crawl/commit/a64f6baaeac6 09:07:59 03dolorous02 07* 0.32-a0-1196-g75f53f5641: Fix missorted colour entry. 10(5 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/75f53f56410e 09:07:59 03dolorous02 07* 0.32-a0-1197-g13c1908155: Add yet more colour names to Blorkula's dialogue. 10(3 minutes ago, 1 file, 4+ 0-) 13https://github.com/crawl/crawl/commit/13c1908155f4 09:22:38 <06mumra> FR: add hex codes for all these colours and we can make the screen flash the exact colour in tiles 09:47:26 <06mumra> fun fact: using X-ray vision in wizard mode can actually affect the path monsters take towards you 09:48:08 <06mumra> (as can ?revelation) 10:59:31 <06dolorous_84348> Good luck figuring out what exact color goblin scarlet is :gloweye: I found it in a massive dictionary some years ago, but there seems to be nothing online about it; Google only spits out either MCU references or RPG references. 11:00:43 <06dolorous_84348> Goblin blue, on the other hand, does show up. 11:10:30 <04dracoomega> That does not surprise me at all, really. (Does ?revelation really last long enough to do so, though?) 11:12:32 <02Monkooky> Didn't reply to this at the time because I got distracted by other projects- I'm not a fan of this but I'm having trouble with reasoning why 11:13:35 <02Monkooky> I would prefer an effect that isn't strictly early game, and it's a little too similar to Hermit's pendant 11:16:22 <02Monkooky> but neither of those are actual reasons against the idea 11:20:10 <04dracoomega> I don't think it's strictly earlygame if the bonus is large enough (and it felt like a cute and appropriate flavor tie-in to 'faking something') and I think being in such a different slot is decent differentiation from hermit's pendant, but if the vibes nonetheless feel off to you, I'm not so convinced as to keep pushing for it anyway 11:31:40 <06dolorous_84348> What about if it switched your Evocations and Invocations skills while equipped? If you have better aptitude at one than the other, and you need the one you have worse aptitude in, that could be useful. 11:34:20 <02Monkooky> that's very funny 11:35:36 <04dracoomega> That's cute, though keep in mind that whatever the effect is, it needs to be powerful enough to offset both the active Orb status downside and the opportunity cost of not putting something else in that slot 11:35:42 <02Monkooky> I don't think it's worth an orb slot much of the time, let alone the charlatan's orb downside yeah 11:36:05 <02Monkooky> in another world I'd also swap piety costs and wand/evoker charges but I think that's inviting a nightmare of edge cases 11:36:25 <04dracoomega> "Evocations becomes equal to the level of your highest other skill" 11:36:30 <04dracoomega> >.> 11:36:42 <06dolorous_84348> That could work too. 11:36:43 <02Monkooky> eh, that's just setting it to bignumber but less good 11:37:05 <02Monkooky> hm, no it does dodge some of the problems 11:54:32 <06dolorous_84348> Just imagine an oni. "Which do I want, giant spiked clubs and reliable teleports, or the ability to easily get the most out of evokables, as long as I keep my fighting skill the highest?" Assuming they've found a good enough set of evokables. Or is that a good enough tradeoff? 12:22:47 <02Monkooky> now that I think about it I'm almost inclined to do it the other way around 12:23:02 <02Monkooky> make every skill crosstrain from evocations while wielding charlatan's orb 12:26:18 <02Monkooky> but this is pretty workable. 12:27:19 <06pleasingfungus> Charlotte’s Orb and it summons orb spiders and hogs 12:39:15 <02Monkooky> so this makes training evocations strictly better than training 3+ other skills which is probably not good design 13:35:56 <06mumra> one turn is technically enough to make a difference: a monster could choose to move into a square that would not have been in your LOS otherwise. when revelation wears off they will now not be in view. from a practical point of view this will probably never make a difference. but it seems to be the observer principle in action 🙂 13:38:05 <06mumra> i can at least imagine what goblin scarlet might be. but i also saw that octarine is in the list, and that is a problem; by definition it is not going to lie in the RGB colour space ... 13:53:57 <06dolorous_84348> Indeed. It seemed appropriate for a character with wizard spells and, it can be presumed, the enhanced sight of a vampire if not a felid. 13:57:29 <06dolorous_84348> I was also briefly tempted to add the color described in https://www.gocomics.com/calvinandhobbes/1994/03/26, but not everyone would get the reference, and that would also be practically impossible to do even though it does lie in the RGB color space, since it's a deliberate mix of colors. 14:01:53 <06dolorous_84348> I was also tempted to put in the impossible colors described on the Wikipedia page of the same name, but I figured that was going a bit too far, and those colors are only accessible via visual tricks anyway. 14:08:53 Hello, I'm not sure if it's okay to ask questions here, but are there any requirements for hosting an official server that integrates with a scoreboard? 14:10:10 https://crawl.develz.org/wiki/doku.php?id=dcss:server:setting_up_dgamelaunch_and_webtiles might be helpful 14:11:00 for the scoreboard to work, you need to set things up so sequell can read your milestone log via https, and ask for it to be added to sequell 14:14:40 After setting up the server, do I just need to write the config/*.yml files in the Sequell GitHub repository and submit a pull request? 14:15:44 geekosaur Thank you for providing the documentation. 14:20:29 Actually, I'm a Korean DCSS player who used to play on the CWZ server. However, due to the prolonged instability of the CWZ server and the absence of its administrator, most Korean DCSS players are currently using the CPO or LLD servers. 14:20:30 We don't know when this situation will improve, but if it continues for much longer, there has been a suggestion to create a new server. That's why I wanted to ask about the procedures required to create an official server that integrates with the scoreboard system. 14:26:15 <02Monkooky> I think everyone would be happy to see a new server go up 14:38:38 Is the conversation in #crawl-dev connected to the Discord server? I am currently using a web IRC client, so I would like to check the previous conversation history when I am away (or participate in the chat via Discord). Where do I need to connect to access this channel on Discord? 14:39:29 you can't, it's a private server 14:41:18 <12ge0ff> ASCIIPhilia: the logs of this channel can be found here: http://crawl.akrasiac.org/logs/cheibriados/ 14:42:09 Thank you. I didn't see the notice right in front of me. haha 14:42:38 <12ge0ff> Are you on the Roguelikes Discord? You can find most of the devs there too 14:43:17 in #dcss channel? 14:43:29 <12ge0ff> yeah 14:54:05 <12ge0ff> ASCIIPhilia: you can usually find gammafunk or PleasingFungus on that Discord server, and they are the devs you might want to talk to 15:09:23 <06mumra> that's a hard reference to get, even though i read the strip possibly dozens of times when i was young 😂 probably an ETC_BARF if anything 15:23:22 <06dolorous_84348> Yes, as opposed to octarine which has its own Wikipedia entry at least. As for ETC_BARF, sounds right, although the limited number of console colors would make it a little tricky to pull off. 15:24:39 <06dolorous_84348> (Back when Makhleb's flavor got reworked, there were a few complaints about how no one knew what mortification of the flesh was, but I figured it had its own Wikipedia entry, so it wasn't all that obscure.) 15:38:02 Unstable branch on underhound.eu updated to: 0.32-a0-1197-g13c1908155 (34) 16:26:15 ASCIIPhilia: Hi! It's been awhile. It would be good to have a successor to CWZ if hong isn't available to administer it any more. 16:27:21 ASCIIPhilia: But one aspect that would be good to clarify is what hong's intentions are. 16:27:43 ASCIIPhilia: If he can't be reached, then I suppose it's best to move forward with your new server. 16:28:14 Technically there's no reason why we can't have two korean servers 16:51:43 <06mumra> Was able to coax a monster into pathfinding through a door with this change 16:51:49 New branch created: pull/3823 (1 commit) 13https://github.com/crawl/crawl/pull/3823 16:51:50 03mumra02 07https://github.com/crawl/crawl/pull/3823 * 0.32-a0-1198-g99eaaefb12: Remind monsters how to open doors 10(15 hours ago, 1 file, 0+ 5-) 13https://github.com/crawl/crawl/commit/99eaaefb1226 16:58:04 <06mumra> Let me know if any qualms but I'm pretty confident to merge this 17:11:14 <09hellmonk> notably, dracoomega seems to have fixed this exact bug in 2013 before it got broken again (in 38c2f1a) 17:13:21 <09hellmonk> so it was working for about 6 years and broken the rest of the time 17:21:16 zewo (L27 HuEE) ERROR in 'mon-cast.cc' at line 1852: Unknown monster spell 'Gell's Gravitas' cast by Canwutygu's ghost (Zig:7) 17:22:21 <04dracoomega> Hahaha, nice 17:24:04 <04dracoomega> I had altogether forgotten 17:24:45 <09hellmonk> I was doing a quick dig to see if how we handled door opaqueness had changed but no, it really does seem like this bug existed until 2013 and then again after 2019 17:25:01 <09hellmonk> which is pretty funny 17:25:15 <06mumra> amazing 17:25:48 <09hellmonk> it also means the bug is not in hellcrawl (vindicated) 18:22:50 <06mumra> ok this seems like a bug in melee_attack but i'm slightly not sure // Vhi's Electrolunge damage if (charge_pow > 0 && defender->alive() && defender->res_elec() <= 0) { int dmg = electrolunge_damage(charge_pow).roll(); int hurt = attacker->apply_ac(dmg, 0, ac_type::half); inflict_damage(hurt, BEAM_ELECTRICITY); } This is for a monster doing an electrolunge. It seems like it's applying its own AC 18:22:50 (attacker->apply_ac) rather than the defender's ? 18:33:38 that seems to happen a lot in melee_attack.cc? 18:34:56 haven't looked at circumstances yet, just did a grep 18:35:20 <06mumra> yeah normally it's because there's a retaliation effect being appliued 18:35:20 which may mean it's a copypasta 18:37:10 <06mumra> e.g. const int dmg = spines_damage(defender->type).roll(); int hurt = attacker->apply_ac(dmg); spines are hurting the attacker so it's correct to apply the attacker's AC 18:37:40 <06mumra> but for electrolunge there it's calculating attacker AC but then hurting the defender 18:46:19 <04dracoomega> ....oops >.> 18:46:50 <04dracoomega> That does just look like a bug, yes 18:51:49 <06pleasingfungus> nice 18:54:31 03mumra02 07* 0.32-a0-1198-g7932608573: Apply defender AC for electrolunge damage 10(44 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/793260857352 19:39:43 -!- The topic of #crawl-dev is: Crawl Development | https://github.com/crawl/crawl | Logs: http://s-z.org/crawl-dev/, temporarily http://crawl.akrasiac.org/logs/cheibriados/ | People with +v have commit access, devs on bridged discord as well | General Crawl-related chat to #crawl | Long stuff to a pastebin service, please 19:39:43 -!- The topic of #crawl is: Play Dungeon Crawl Stone Soup online now! Type ??online for instructions, ??lg / !lg for play stats | PM Sequell for long queries | http://crawl.develz.org | FooTV game replays: ??footv for instructions | #crawl-dev for dev discussion, #crawl-offtopic for offtopic 20:00:50 -!- ert7 is now known as ert 20:20:19 <06cao> Renew Success by crawl Success certs: crawl.akrasiac.org_ecc 20:26:35 <06pleasingfungus> that’s new. 20:44:31 -!- runrin_ is now known as runrin 20:53:45 -!- matta_ is now known as matta 21:00:40 <04dracoomega> I am not even certain what that means here 21:01:20 <06pleasingfungus> sounds like cao renewed its tls cert..? 22:35:20 Unstable branch on crawl.develz.org updated to: 0.32-a0-1198-g7932608573 (34) 22:57:38 Windows builds of master branch on crawl.develz.org updated to: 0.32-a0-1198-g7932608573 23:13:33 Unstable branch on cbro.berotato.org updated to: 0.32-a0-1198-g7932608573 (34) 23:55:07 Monster database of master branch on crawl.develz.org updated to: 0.32-a0-1198-g7932608573