00:31:29 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-4837-gdd944a72e4 01:09:33 -!- Amore is now known as earend1 03:24:17 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-4837-gdd944a72e4 03:37:05 <09gammafunk> makes sense 05:11:43 Unstable branch on crawl.akrasiac.org updated to: 0.30-a0-250-g41c1d76 (34) 06:07:37 03Roadster Tracker02 {GitHub} 07https://github.com/crawl/crawl/pull/2770 * 0.30-a0-224-g8550ff61da: Sneaking my name in to the credits:-) 10(17 seconds ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/8550ff61dad0 06:27:33 03Roadster Tracker02 {GitHub} 07https://github.com/crawl/crawl/pull/2770 * 0.30-a0-257-gcf37ddbffa: Merge branch 'crawl:master' into mons_from_base_type 10(65 seconds ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/cf37ddbffac5 07:25:36 -!- wchar_t is now known as cot 08:10:38 -!- allbery_b is now known as geekosaur 09:48:32 <09gammafunk> @advil this question is a bit out there, but I'm looking for a coordinate-free clua way of differentiating map locations, mostly because I can't use waypoints everywhere, particularly in the abyss where we couldn't even plausibly allow waypoints, due to things like abyss area shifts as the player gets close to the map edge. I mostly care about differentiating positions as opposed to needing relative coordonates. So a solution I had 09:48:32 in mind was simply making a string of concatenated feature names by iterating over los in a fixed order. Then I'd use this string as a Lua table key and do a simple test of existence to know if two los "hashes" are the same. The situations where two actually different positions had same string wouldn't really be relevant to this monster-related application. My only concern is using this ~5kb string as a table key and whether that's likely to cause 09:48:33 clua issues. Probably wouldn't have more than 10-20 of these sitting in memory 09:52:15 03PleasingFungus02 07* 0.30-a0-251-g98ac1a292c: Re-fix damage ratings 10(66 seconds ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/98ac1a292c36 10:04:54 Unstable branch on crawl.kelbi.org updated to: 0.30-a0-251-g98ac1a292c (34) 10:05:54 <06advil> I don't see any obvious clua problems that would cause, though I wonder if the waypoint system could be improved to work better for this use case 10:06:18 <06advil> there's probably no real technical reason to limit to 10 waypoints (I haven't checked how these are saved though) 10:06:41 <06advil> and abyss is easy, you just clear any waypoints when the map shifts or is deleted 10:07:00 <06advil> I assume if that's not implemented it's just because it's not very useful for a human 10:10:15 <06advil> I can't really remember how clua handles coordinates but I could also imagine something very straightforward to have non-leaky coords, like marking the initial point of level entry as 0,0 and providing a coordinate API that uses coordinates relative to that location 10:16:41 <06advil> well, travel cache serialization is a bit inflexible but it's in its own chunk and waypoints are at the end, so that probably would make it easy to modify 11:51:30 <10PleasingFungus> thought there were 10 waypoints for ui reasons? 12:14:55 03Roadster Tracker02 {GitHub} 07https://github.com/crawl/crawl/pull/2770 * 0.30-a0-259-g4ee4bd7304: Merge branch 'crawl:master' into mons_from_base_type 10(56 seconds ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/4ee4bd7304c3 12:40:00 <09gammafunk> Waypoint multiplicity isn't a concern for my use case. qw uses only 2 waypoints ever to maintain level cache for no more than two levels. It only needs one waypoint on a level to establish a coordinate system with 0,0 at the waypoint it set where it entered the level when that level's cache was initialized. Since caching many levels is prohibitive to some degree, I wouldn't need more waypoints than what's currently available. But a 12:40:00 non-leaky coordinate system that works on 100% of levels either as an API or by allowing waypoints everywhere would definitely be great and would obviate the need for my los hashing scheme 12:44:30 imamambug (L27 DjWr) Crash caused by signal #11: Segmentation fault (D (Sprint)) 12:44:32 <09gammafunk> And thanks for response, on mobile so my initial reply was a bit terse 13:19:31 <06advil> yeah it uses digits in the ui, and would be messy in various places to have more usable to a human player 14:07:23 I'm curious, what's the general thought on how one can get a very good idea on loot IDs via looking at vault files? And what about specifically in the context of portals, which are even more susceptible to that type of "metagaming"? 14:09:12 Although I guess in the cases in which this can really matter, the items you're trying it ID are likely type-ID'd.... 14:16:48 <05kate> i think most vaults where this was an issue have been changed to place either identified loot, or just random loot of the given consumable type, yeah 14:33:24 03advil02 07* 0.30-a0-252-gd42a0f4da5: feat: waypoints for all levels (gammafunk) 10(4 minutes ago, 5 files, 50+ 18-) 13https://github.com/crawl/crawl/commit/d42a0f4da5c4 14:37:36 03advil02 07* 0.30-a0-253-g6b895097cc: fix: remove some missed commented code 10(44 seconds ago, 1 file, 0+ 2-) 13https://github.com/crawl/crawl/commit/6b895097ccdd 14:54:30 <09gammafunk> @advil wow, thanks a lot! Looking at the size of that diff , it was slightly non-trivial although still on mobile and can’t read well at present. Will test it out with qw and let you know if there are problems 14:58:54 Unstable branch on crawl.kelbi.org updated to: 0.30-a0-253-g6b895097cc (34) 15:03:35 <06advil> it was pretty quick to write actually 15:03:51 <06advil> if I hadn't already done the abyss stuff for rune detection that would not have been true 15:03:54 <09gammafunk> Actually doesn’t look too bad, thankfully. Qw will need to check waypoint existence in the abyss each turn and reset cache but that’s easy 15:03:57 <09gammafunk> Aha 15:04:35 <06advil> I wonder if there's even a way to get that run position in lua 15:04:37 <06advil> *rune 15:04:47 <09gammafunk> I was also musing about a lua api for relative…haha yeah 15:05:41 <09gammafunk> Offset vector to sensed rune? Same approach as lua api for waypoints 15:06:50 <09gammafunk> I also did not realize until very recently that view.feature_at() works for out of los and uses your map cache 15:07:07 <06advil> for a general relative positioning api the stupid thing to do might just be to save level entry pos as a prop on player, and provide a single lua function that gets the value of that prop as player-relative coordinates? then you'd have to do some munging on the lua side as the player moves, but probably fine. It would also take some abyss-specific work 15:09:06 <09gammafunk> And then you and monster and item lua libs all have a pos function returning x,y relative to this 15:09:08 <09gammafunk> ? 15:10:01 <09gammafunk> Would simplify things and remove need for waypoints (for qw at least) 15:10:51 <09gammafunk> Also greatly simplify qw level cache I think 15:11:16 <06advil> In the simplest case I wouldn't bother to change their coordinate systems (which IIRC are player-relative?), but put it on the script to do coordinate adjustment 15:11:29 <06advil> probably better for backwards compatibility that way 15:12:14 <06advil> e.g. the script would have to subtract the arbitrary origin from any coordinates they get from the api 15:12:25 <06advil> not very elegant, but it'd be easy to implement 15:24:18 Unstable branch on underhound.eu updated to: 0.30-a0-253-g6b895097cc (34) 16:40:39 <09gammafunk> yeah, that would be fine 17:28:43 03Alex Jurkiewicz02 {advil} 07* 0.30-a0-254-g79d7498bdb: refactor!: Hardcode games_config_dir 10(8 weeks ago, 3 files, 5+ 19-) 13https://github.com/crawl/crawl/commit/79d7498bdb3e 17:28:43 03Alex Jurkiewicz02 {advil} 07* 0.30-a0-255-g9dacfef450: feat: Support webserver config.yml override file 10(8 weeks ago, 3 files, 27+ 1-) 13https://github.com/crawl/crawl/commit/9dacfef450f8 17:28:43 03Alex Jurkiewicz02 {advil} 07* 0.30-a0-256-g55547e2a59: refactor: Move all game definitions to games.d/ 10(8 weeks ago, 2 files, 27+ 34-) 13https://github.com/crawl/crawl/commit/55547e2a59a1 17:28:43 03Alex Jurkiewicz02 {advil} 07* 0.30-a0-257-g30c107533f: refactor: Keep use_game_yaml setting 10(8 weeks ago, 2 files, 6+ 2-) 13https://github.com/crawl/crawl/commit/30c107533f86 17:28:43 03advil02 07* 0.30-a0-258-g902f57cb4f: refactor: adjust how games are loaded 10(11 minutes ago, 3 files, 92+ 19-) 13https://github.com/crawl/crawl/commit/902f57cb4f28 17:32:35 <09gammafunk> refactor! 17:33:24 <06advil> haha yeah I'm not sure if that ! was meaningful, but I left it 17:33:34 <06advil> I think it maybe indicates something non-backwards-compatible? 17:33:58 <09gammafunk> oh, I see the upper-case note at the end, yeah 17:34:43 Unstable branch on crawl.kelbi.org updated to: 0.30-a0-258-g902f57cb4f (34) 17:34:53 <09gammafunk> this is again reminding me that I should pull out my commits to make various fixes and adjustments and probably the wizard user functionality in our dgamelaunch repo 17:35:28 <09gammafunk> the dgl scripts I made to install experimentals would probably also be useful 17:35:36 <09gammafunk> *install and update 17:36:04 03advil02 07* 0.30-a0-259-g4eb8f405f2: docs: update server admin comment for config.py 10(2 minutes ago, 1 file, 8+ 4-) 13https://github.com/crawl/crawl/commit/4eb8f405f293 17:37:03 <06advil> I wonder if Alex keeps adding these features (packaged in a way that I will commit) to try to get people to stop using dgamelaunch-config 😄 17:37:36 <09gammafunk> heh, well, what would really potentially stop people from using it would be if he could neatly automate all the things it does 17:38:11 <06advil> he's been chipping away at it (as have I for that matter, with moving a lot of dgamelaunch db management stuff to python) 17:38:22 <06advil> still a lot though 17:38:46 <06advil> and I don't really ever see dgamelaunch itself going away 17:38:53 <09gammafunk> yeah, until it can handle stuff like doing the build, installing the build, managing the trunk game transfer system.. 17:39:06 <09gammafunk> his setup is easier because, right, no console and also he just forces transfers for trunk 17:39:13 <06advil> save game transfers ought to be doable! 17:39:36 <06advil> ish 17:39:56 <09gammafunk> having python-side awareness of trunk installations etc? 17:39:57 <06advil> maybe I wouldn't really want to manage multiple trunk versions within the webtiles server though 17:40:50 <06advil> I've been thinking of adding something that could simplify a lot of the config templating though 17:40:57 <09gammafunk> @advil btw, I hope you appreciate what I took the time to implement here (you'll need to log in before this will do anything) https://crawl.dcss.io/#play-dcss-bot-git 17:41:15 <09gammafunk> made by observing the trunk game transfer code and making my own prompts 17:41:35 <06advil> rip (although maybe that's the point) 17:41:35 <06advil> https://cdn.discordapp.com/attachments/747522859361894521/1029916776131207188/Screen_Shot_2022-10-12_at_8.41.21_PM.png 17:41:39 <09gammafunk> based on the dgamelaunch bash code; the fact that it creates the menus from a bash script is hilarious 17:41:42 <09gammafunk> yep, that's the point 17:41:55 <09gammafunk> it's the same type of dialogue made by trunk transfer dgamelaunch stuff 17:42:06 <06advil> oh heh are you using floraline's lobby layout code 17:42:11 <09gammafunk> indeed! 17:42:23 <06advil> that's another thing I want to add in a more built-in way to game config 17:42:31 <09gammafunk> yes, would be great 17:42:46 <09gammafunk> like I was going to look into hiding these bot installs with a sort of crude html hack 17:42:55 <09gammafunk> but it's actually important that I be able to use it from webtiles as qw sometimes 17:42:57 <09gammafunk> to debug 17:43:04 <09gammafunk> so I just hid them under that drop down 17:43:10 <09gammafunk> and I can still get to them when I want 17:43:32 <09gammafunk> and yeah I guess we could just elect to force transfer users for trunk 17:43:35 <09gammafunk> and not give the option 17:43:39 <09gammafunk> it would certainly simplify many things 17:43:59 <09gammafunk> it's an awesome feature and I'm impressed that it's managed to work as well as it has 17:44:11 <09gammafunk> but seems to kind of strain our infrastructure 17:45:09 <06advil> I suppose that must be what CPO does? 17:45:14 <09gammafunk> yep 17:45:23 <09gammafunk> and other webtiles-only servers like cwz 17:45:25 <09gammafunk> presumably lld 17:46:06 <09gammafunk> alex is pretty hardcore, no console, no transfer prompts, and he even deletes morgues after like 6 months 17:46:17 <09gammafunk> in his words he wants zero maintenance 17:46:37 <09gammafunk> but I guess you'll just run out of disk space eventually if you allow morgues to pile up 17:47:05 <09gammafunk> !lg * cpo won end<2022-01-01 17:47:07 <04Cerebot> 3584. MrDizzy the Conqueror (L27 KoAM of Wu Jian), escaped with the Orb and 3 runes on 2021-12-31 14:36:58, with 1368982 points after 126613 turns and 9:12:38. 17:47:10 <09gammafunk> !lg * cpo won end<2022-01-01 -log 17:47:12 <04Cerebot> 3584. MrDizzy, XL27 KoAM, T:126613: https://crawl.project357.org/morgue/MrDizzy/morgue-MrDizzy-20211231-143658.txt 17:47:22 <09gammafunk> well that's still there 17:47:27 <09gammafunk> !lg * cpo won end<2021-01-01 17:47:29 <04Cerebot> 3176. Caerwyn the Conqueror (L27 DsBe of Trog), escaped with the Orb and 3 runes on 2020-12-30 15:55:59, with 1800758 points after 65208 turns and 1:49:50. 17:47:36 <09gammafunk> !lg * cpo won end<2021-01-01 -log 17:47:36 <04Cerebot> 3176. Caerwyn, XL27 DsBe, T:65208: https://crawl.project357.org/morgue/Caerwyn/morgue-Caerwyn-20201230-155559.txt 17:47:45 <09gammafunk> ok, maybe he didn't go through with that then 17:49:12 Unstable branch on crawl.kelbi.org updated to: 0.30-a0-259-g4eb8f405f2 (34) 18:51:38 <09hellmonk> if (you.duration[DUR_DEATHS_DOOR] > 25 * BASELINE_DELAY) you.duration[DUR_DEATHS_DOOR] = (23 + random2(5)) * BASELINE_DELAY; 18:51:46 <09hellmonk> is this the funniest crawl code goof of all time 19:07:37 <10PleasingFungus> uh 19:07:52 <10PleasingFungus> i don’t quite get it 19:08:02 <10PleasingFungus> what is it even trying to do? 19:09:29 <09hellmonk> it is trying to set a max on ddoor duration but not make it fixed, ig so you can't count the turns 19:10:03 I'm not sure what's funny about it? 19:10:08 <09hellmonk> this doesn't work for several reasons but the funny thing is that if you roll 26 duration naturally that's worse than rolling 25 19:10:52 in what sense? 19:11:58 it looks like the duration will have expected value 25 19:12:18 <09hellmonk> nm, I'm bad at math lol 19:12:20 also not sure why this doesn't work, unless this line is run every turn or something 19:52:07 <09hellmonk> ok I've actually got something this time: it appears that dithmenos shadow mimics a thrown item with HD dependent on your currently wielded weapon rather than throwing skill 19:55:36 03hellmonk02 07* 0.30-a0-260-g8b876d146b: randomly round disjunction duration 10(36 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/8b876d146b2b 19:55:36 03hellmonk02 07* 0.30-a0-261-ge375a5e35f: randomly round ddoor power checks 10(34 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/e375a5e35fdf 19:55:36 03hellmonk02 07* 0.30-a0-262-gc2e06944d9: adjust Shatter damage 10(27 minutes ago, 2 files, 5+ 4-) 13https://github.com/crawl/crawl/commit/c2e06944d944 19:55:36 03hellmonk02 07* 0.30-a0-263-ge22c4ad894: fix: Update battlesphere HD when recalling 10(2 minutes ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/e22c4ad89444 20:09:21 Unstable branch on crawl.kelbi.org updated to: 0.30-a0-263-ge22c4ad894 (34) 23:29:51 Unstable branch on cbro.berotato.org updated to: 0.30-a0-263-ge22c4ad894 (34)