00:00:23 <floraline> alexjurkiewicz: what happens if you remove a game from the config while someone is playing it
00:31:16 <Eksell> Unstable branch on CRAWL.XTAHUA.COM updated to: 0.24-a0-10-gb84795d (34)
00:37:55 <Cheibriados> 03ebering02 07* 0.24-a0-11-g46ee55f: Don't use player spell colours for monster spellbooks 10(70 seconds ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/46ee55fc9466
00:47:07 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-11-g46ee55fc94 (34)
01:06:57 <Eksell> Stable (0.23) branch on CRAWL.XTAHUA.COM updated to: 0.23-b1-3-gdda8187
01:16:58 <alexjurkiewicz> floraline: i never do that
01:17:25 <alexjurkiewicz> floraline: the games I remove are experimentals and forks, and generally they only get removed when nobody has played them in weeks
01:18:52 <floraline> what if you didn't know they started playing it
01:18:55 <alexjurkiewicz> feature request: crawl binaries can output a list of places and their correct grammatical details
01:19:03 <alexjurkiewicz> floraline: I check the lobby
01:19:16 <floraline> it's possible though, right?
01:19:32 <alexjurkiewicz> oh, in terms of theoretical concern yeah who know what would happen, definitely something bad
01:20:35 <alexjurkiewicz> you could get around this by adding an "enabled = true|false" option to each game I guess, which defines if players can start the game but otherwise leaves the game defined. So removing a game would become a two step process
01:20:35 <alexjurkiewicz> game game game
01:20:43 <floraline> that's a good idea
01:20:53 <floraline> what if someone loads the lobby but doesn't click anything yet
01:20:57 <floraline> then you remove the game and then they try to launch it
01:21:00 <ebering> alexjurkiewicz: -place-json patch welcome :)
01:23:38 <alexjurkiewicz> it's so tempting. I've written grammar logic about three times in each language
01:24:06 -!- amalloy_ is now known as amalloy
02:40:14 <floraline> alexjurkiewicz: have you had any problems as a result of toml being unordered?
02:48:38 -!- amalloy is now known as amalloy_
03:51:23 <Kelbell> Stable (0.23) branch on crawl.kelbi.org updated to: 0.23-b1-3-gdda8187c6f
05:44:07 <alexjurkiewicz> floraline: nope, but sidenote I really dislike toml aesthetically
08:33:48 <Cheibriados> 03advil02 07* 0.24-a0-12-ga9077f9: More improvements to randomization 10(19 hours ago, 5 files, 24+ 42-) 13https://github.com/crawl/crawl/commit/a9077f9e91af
08:33:48 <Cheibriados> 03advil02 07* 0.24-a0-13-g18b6515: Remove crawl.random_element and replace with something more stable 10(20 minutes ago, 15 files, 183+ 204-) 13https://github.com/crawl/crawl/commit/18b65154750a
08:33:48 <Cheibriados> 03advil02 07* 0.24-a0-14-g542fcb7: Don't use ui_random for bones file choice 10(11 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/542fcb7d11b2
08:33:48 <Cheibriados> 03advil02 07* 0.24-a0-15-ge63ed06: More RNG initialization improvements 10(8 minutes ago, 3 files, 13+ 4-) 13https://github.com/crawl/crawl/commit/e63ed0601461
08:37:03 <advil> I ... may have underestimated how many subtle seed issues there would be
08:37:34 <advil> I think there's still something lurking on the lua side
08:37:43 <advil> possibly just those other util random functions
08:38:18 <advil> it's at least getting harder to replicate seed divergence though
08:43:14 <advil> although half of this process is just figuring out what global state I need to reset for tests to work properly
08:43:46 <advil> like, it turns out that the number of octopus kind rings spawned is global state on `you`, and has a huge impact on generation, sooner or later, if you don't reset it
08:44:00 <advil> *king
08:47:43 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-15-ge63ed06014 (34)
11:23:12 <Lantell> Stable (0.23) branch on underhound.eu updated to: 0.23-b1-3-gdda8187c6f
12:26:49 <Henzell> Unstable branch on crawl.akrasiac.org updated to: 0.24-a0-15-ge63ed06 (34)
12:50:14 <floraline> i don't think yaml or toml will work for config files because they are unordered
12:50:38 <floraline> in trunk, the order in which the games are loaded is important
12:53:03 <floraline> it might be best to just move the game definitions and milestone_files into a separate module and just control access to it through config.py
12:55:29 <floraline> this also means it doesn't make a lot of sense to spread out your game defs across multiple config files, it would be tedious to make sure they are loaded in the right order
12:59:53 <gammafunk> floraline: well, webtiles-changes uses toml
13:00:04 <gammafunk> which I've used for other projects and it does work well
13:00:31 <gammafunk> floraline: oh, I see what you're saying
13:00:38 <gammafunk> I think toml supports arrays ok
13:01:45 <gammafunk> I'm not sure what you mean by order of game loading being important for trunk though
13:01:56 <gammafunk> if you're talking about menu-wise, well that's simply data that could be in a field
13:02:04 <floraline> i'm just talking about the menu
13:02:14 <floraline> i want to avoid having to handle ordering with another setting though
13:02:20 <floraline> that would get really tedious
13:03:43 <gammafunk> floraline: https://github.com/toml-lang/toml#array-of-tables
13:04:16 <gammafunk> I thik webtiles-changes might already be using this for its config, or maybe I'm confising that with my beem/cerebot project
13:04:20 <floraline> oh
13:04:25 <gammafunk> but I don't think it'd be hard to get an ordering, yeah
13:04:44 <floraline> i didn't see that part, thanks
13:04:44 <floraline> yes webtiles-changes has that exact syntax
13:05:08 <gammafunk> again, toml is just what edlothiol chose at the time, quite some years ago, when web 2.0 stuff was a lot newer
13:05:18 <gammafunk> I don't know if there's a better format choice these days
13:06:26 <floraline> i was actually looking at yaml over toml just because we now already have pyyaml, and it's a lot more popular
13:06:32 <gammafunk> it would certainly serve our purposes well enough, but if there's a better format with reasonably python support
13:06:38 <gammafunk> yeah
13:06:48 <gammafunk> don't know if yaml has that issue you describe
13:06:53 <floraline> it seems you can do the same array of tables in yaml
13:07:06 <gammafunk> probably a more reasonable choice then
13:07:45 <floraline> ok, well that solves the problem of file format. i already wrote it all in yaml so i don't have to rewrite that now
13:07:59 <Kelbell> TimeString (L5 HOFi) ASSERT(load_mode != LOAD_VISITOR) in 'files.cc' at line 1439 failed. (D:255)
13:08:17 <floraline> i'm still not sure i like pulling these definitions from multiple files
13:08:30 <floraline> you would have to make sure your files are being ordered the way you want them, which again can get tedious
13:10:02 <gammafunk> floraline: well, that tends to be solved on unix systems with a simple numeric filename scheme
13:10:18 <gammafunk> perhaps that style has fallen relatively out of favor but I think it's still pretty widely used
13:10:52 <floraline> yeah i like that style but i was hoping to avoid something like that
13:11:03 <floraline> any time you wanted to reorder things you might have to rename a ton of directories
13:11:21 <gammafunk> directories? wouldn't it simply be files
13:11:30 <floraline> oh right yes
13:11:31 <floraline> just files
13:11:49 <floraline> i guess if i just wanted to have one big file i can do that. and others can have multiple files if it doesn't bother them
13:12:01 <gammafunk> yeah
13:12:13 -!- amalloy_ is now known as amalloy
13:12:44 <floraline> well, that's very cool. thanks for finding that section in the documentation
13:13:01 <floraline> i was only looking at maps which are all unordered
13:14:10 <floraline> the only problem now is to figure out what needs to happen when we pull the new values in
13:14:26 <floraline> i know at the very least, the milestone file watchers will have to be restarted
13:15:00 <floraline> i want to make this work while changing as little as possible
13:16:23 <gammafunk> well, webtiles-changes can give you an idea about that, much of that code will be the same
13:21:15 <gammafunk> well, I guess the relevant commits aren't quite as informative but
13:21:16 -!- MarvinPA_ is now known as MarvinPA
13:21:28 <gammafunk> %git b2665bdd2e
13:21:28 <Cheibriados> 07edlothiol02 * 0.15-a0-701-gb2665bd: Load Webtiles server config from a JSON file. 10(4 years, 9 months ago, 13 files, 324+ 249-) 13https://github.com/crawl/crawl/commit/b2665bdd2e27
13:21:37 <gammafunk> this is in general a good commit to check out
13:21:51 <gammafunk> especially since it's relative to webtiles trunk (at the time)
13:22:04 <gammafunk> it's the implementation of loading config from json instead of from python config
13:22:22 <gammafunk> and as part of this implementation, basic reloading of config on the fly is supported
13:22:52 <gammafunk> his commit to actually implement that via a signal is a bit after and is very simple, just setting up the handler to call that load() method he implements for Conf
13:23:19 <gammafunk> some of the details of that approach we'd not want to do, particularly how he instantiates Conf
13:24:12 <gammafunk> basically trying to set it up when the module is imported
13:24:23 <gammafunk> I actually used this approach initially for beem, but I found it's just problematic
13:25:18 <gammafunk> so I went with having a class for my server that just loads the config as part of its __init__ method in a try/exception
13:26:16 <gammafunk> tryig to create a global config object when the config module is loaded is just really a pain to deal with
13:26:33 <gammafunk> because managing order of execution becomes a pain
13:26:46 <floraline> that makes sense
13:26:48 <gammafunk> better to just have all that happened at a specific time so you can control the sequence more sanely
13:27:03 <floraline> i'm not seeing where the milestone tailers are being reloaded in webtiles-changes
13:27:47 <gammafunk> if you want to see my approach in terms of inspiration, check out https://github.com/gammafunk/beem/blob/master/beem/server.py and https://github.com/gammafunk/beem/blob/master/beem/config.py and my config toml file for reference is here: https://github.com/gammafunk/beem/blob/master/beem_config.toml.sample
13:28:05 <floraline> thanks, i will take a look at that
13:28:14 <gammafunk> hrm, milestone tailers
13:28:29 <gammafunk> not sure where those are in trunk webtiles
13:29:52 <floraline> it happens once here in server.py on startup: https://github.com/crawl/crawl/blob/webtiles-changes/crawl-ref/source/webserver/server.py#L281
13:30:06 <floraline> so far it looks the same as trunk
13:30:30 <floraline> but when the config is reloaded, i dont
13:30:42 <floraline> see any code going back to check these tailers
13:34:47 <floraline> that's the kind of thing that worries me most about making this change, reloading stuff from disk is trivial
13:35:08 <floraline> it's making sure that the code expects these values to change at any time
13:35:37 <floraline> things could change in the middle of a request
13:39:31 <gammafunk> yes, this is the concern I had as well
13:40:06 <gammafunk> as far as I could tell, nothing much was done to ensure integrity of the server state and connections' states upon reload
13:41:54 <gammafunk> it's certainly possible to do that in terms of having methods for each connection where that method will have the connection reconnect, essentially
13:42:01 <gammafunk> but exactly what to do requires some careful looking I think
15:31:07 <alexjurkiewicz> floraline: about game file ordering I just order my config files, eg 10-trunk.toml, 20-0.22.toml
15:32:03 <floraline> that's what i was picturing
15:32:13 <alexjurkiewicz> floraline: the advantage of different files is that if you need to write a config automatically (say when you update trunk) you can do this without needing to touch unrelated config in the same file
15:32:42 <floraline> why do you have to change the config when you update trunk?
15:32:49 <alexjurkiewicz> to point to the latest version
15:33:24 <floraline> oh you're not using dgl
15:33:24 <alexjurkiewicz> crawl_binary = "/var/dcss/builds/dcss-trunk/0.24-a0-15-ge63ed06/bin/dcss"
15:33:24 <alexjurkiewicz> right, I just start the binary directly
15:33:53 <gammafunk> floraline: one of the things I looked into working on, not sure if I ever pushed anything here
15:34:00 <gammafunk> was automating creation of dgl menu files
15:34:16 <gammafunk> this would be necessary if you wanted to make adding branches more painless
15:34:29 <gammafunk> extending the dgl scripts to better support webtiles-changes and automation in that regard
15:34:36 <gammafunk> actually I think I did put a branch to the dgl script repo
15:35:23 <gammafunk> yes, guess neil gave me access to this repo
15:35:24 <gammafunk> https://github.com/neilmoore/dgamelaunch-config/tree/webtiles-changes
15:35:47 <gammafunk> hrm
15:35:53 <gammafunk> did neil actually make these commits, or did I
15:36:30 <gammafunk> he must have made those and maybe I made some more on top
15:36:55 <gammafunk> ah, yes
15:36:56 <gammafunk> https://github.com/gammafunk/dgamelaunch-config/commits/webtiles-changes
15:37:05 <gammafunk> so I made a fork with some additional commits
15:37:52 <gammafunk> I made some dgl scripts to help manage webtiles
15:38:49 <gammafunk> this was a long time ago, and I've forgotten most of these details, but yeah on the dgl side I guess there's some script changes we'd want to make to better support webtiles config management
15:39:36 <gammafunk> this in part depends on a commit I made in webtiles-changes to split out the client and the server in the makefile
15:39:45 <gammafunk> although I think some of that has changed in our current makefile
15:40:12 <gammafunk> also, that one person who was working on porting our build to cmake...I guess they stopped working on that
15:41:27 <gammafunk> there are a lot of pieces to put together when trying to change this stuff
15:51:38 <switch__> did they stop working on it? last i checked they had it working for their dev env and hadn't gotten around to making sure it worked for the other environments
15:52:11 <switch__> last commit was 23 days ago, heh
16:00:30 <floraline> i feel like we should be able to at least take small steps forward
16:00:53 <floraline> allowing certain config items to be reloaded on the fly should be doable now without too many changes
16:04:01 <floraline> my goal right now is to find the smallest changes that solve the problems i'm looking at right now and also don't require any action from any server operators
16:10:32 <gammafunk> floraline: yes, although it's actually ok if this does require setup from server operators, as you and alexjurkiewicz are the two admins who most want this functionality
16:10:32 <gammafunk> and you both are find with doing a configuration change or two
16:11:43 <gammafunk> if you mean in terms of interrupting operation for normal admins, you'd probably have to make some kind of unusual change to the client for that to even be possible
16:12:19 <gammafunk> hrm, maybe I'm misremembering how changes to webtiles server are rolled out
16:12:40 <gammafunk> I guess normal updates do change webtiles server, don't they
16:12:52 <gammafunk> we just exclude e.g. config.py, or something like that?
16:14:54 <floraline> some files are copied from webserver/ when install-trunk runs
16:15:30 <floraline> so a lot of servers will fail to restart if the new code does not run using old config
16:16:16 <gammafunk> ah, yeah, I see
16:16:36 <floraline> so i'm fine with using yaml because everyone should have pyyaml now
16:16:42 <floraline> but i don't want to use toml
16:17:01 <gammafunk> well, that's fine re using yaml, but again how would it be possible to not interrupt server operation
16:17:12 <gammafunk> if config moves to yaml, they must convert their config to yaml
16:17:27 <floraline> it can support both
16:17:47 <floraline> at the risk of complicating the codebase
16:18:01 <gammafunk> yeah, it may be a lot wiser to simply coordinate with admins to roll out these changes
16:18:42 <gammafunk> I agree with your idea of minimalism for initial changes, but I think trying to guarantee a 0 effort upgrade is possibly not feasible
16:19:15 <gammafunk> for dgl based servers where an update would break them, we do have cue cxc cao and cbro to consider
16:19:50 <gammafunk> probably those last two present a bit of challenge since the admins have lest time available
16:20:26 <gammafunk> might want to get opinions from |amethyst and johnstein themselves, especially the former as to what is the best approach
16:21:09 <gammafunk> I think Medar and TZer0 would probably be ok with going through some steps to convert their config over if given notification
16:25:14 <gammafunk> we could also provide a script that converts config.py data to yaml
16:25:37 <gammafunk> since config.py has python data, that's easy to just load, and I imagine it wouldn't be too hard to convert that to yaml using pyyaml
16:58:19 <floraline> my original idea was to move options that were made to be reloadable into a config.yml and everything else stayed in config.py
16:58:38 <floraline> that way config.py could be changed to allow the code to keep asking for config.games or whatever so it would neatly fallback in the event that you had an old config.py
17:02:02 <floraline> i guess we could move everything into a config.yml all at once, and just have reload pick up some changes and not others
17:03:37 <floraline> in that case, migration might be easily automated. just yaml.dump(config.py) into config.yml and you're good to go. your new config won't look pretty that way, but it would work
17:09:25 <Cheibriados> 03gammafunk02 07* 0.24-a0-16-g0a484b6: Tweak the no-teleport message for Gauntlets 10(6 hours ago, 1 file, 4+ 1-) 13https://github.com/crawl/crawl/commit/0a484b665a29
17:09:25 <Cheibriados> 03gammafunk02 07* 0.24-a0-17-g2c779b6: Prevent use of Passage of Golubria in Gauntlets (Yermak) 10(6 hours ago, 2 files, 12+ 0-) 13https://github.com/crawl/crawl/commit/2c779b6b6d14
17:09:25 <Cheibriados> 03gammafunk02 07* 0.24-a0-18-gcd3d6b9: Update documentation of a function 10(3 hours ago, 1 file, 4+ 1-) 13https://github.com/crawl/crawl/commit/cd3d6b93d089
17:12:15 <Cheibriados> 03gammafunk02 07[stone_soup-0.23] * 0.23-b1-4-g53b5eb8: Tweak the no-teleport message for Gauntlets 10(6 hours ago, 1 file, 4+ 1-) 13https://github.com/crawl/crawl/commit/53b5eb8da2c0
17:12:15 <Cheibriados> 03gammafunk02 07[stone_soup-0.23] * 0.23-b1-5-gf1159c6: Prevent use of Passage of Golubria in Gauntlets (Yermak) 10(6 hours ago, 2 files, 12+ 0-) 13https://github.com/crawl/crawl/commit/f1159c66918a
17:12:15 <Cheibriados> 03gammafunk02 07[stone_soup-0.23] * 0.23-b1-6-ga921848: Update documentation of a function 10(3 hours ago, 1 file, 4+ 1-) 13https://github.com/crawl/crawl/commit/a921848b4870
17:12:31 <alexjurkiewicz> fwiw with webtiles changes you can send a particular signal (USR1) and *only* games are reloaded
17:15:40 <gammafunk> !learn edit gauntlet[1] s/teleport/teleport or use {passage of golubria}/
17:15:40 <Sequell> gauntlet[1/3]: A new portal in 0.23 that replaces the {labyrinth}. Players choose a path to fight through combat arenas with monsters and loot before fighting the {minotaur} for its hoard. Access to each arena is one-way through transporters until the minotaur and exit are reached. Players can't teleport or use {passage of golubria}, but other translocations work normally.
17:17:35 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-18-gcd3d6b93d0 (34)
17:31:35 <floraline> if nobody has any objection i think i'm going to turn off this trunk build announcement
17:31:41 <floraline> as long as it's on a 15 minute schedule it's more annoying than anything
17:31:49 <advil> it's actually sort of useful when making changes
17:32:05 <advil> I mean, if it's annoying you, it's your bot though :)
17:32:26 <advil> but it probably won't announce unless there have been commits, right?
17:32:53 <advil> anyways, I do like to know when my things go live
17:47:11 <Cheibriados> 03advil02 07* 0.24-a0-19-g0b81bb8: Reset temple vault tags when doing temple setup 10(19 minutes ago, 2 files, 18+ 1-) 13https://github.com/crawl/crawl/commit/0b81bb8f681f
17:47:45 <advil> oops, that somehow ended up as one commit
17:47:47 <gammafunk> floraline: it you do decide to keep it up, you can also have your irc client specifically block those messages from Kelbell
17:47:56 <gammafunk> without blocking other ones
17:48:07 <gammafunk> well, you probably can; most clients support that kind of filtering
17:55:58 <floraline> oh i can keep it then, no problem. i had no idea whether they were useful to anyone
17:56:08 <floraline> i'm using irssi, maybe i'll see if i can figure that out
18:02:30 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-19-g0b81bb8f68 (34)
18:11:13 <travis-ci> The build has errored. (master - cd3d6b9 #11013 : gammafunk): https://travis-ci.org/crawl/crawl/builds/487164782
18:13:22 <gammafunk> false alarm, but TIL we have a crosscompile build in travis
18:13:28 <gammafunk> I probably knew that at one point, but forgot
18:21:49 <Lantell> Unstable branch on underhound.eu updated to: 0.24-a0-19-g0b81bb8f68 (34)
19:07:28 <travis-ci> The build has errored. (stone_soup-0.23 - a921848 #11014 : gammafunk): https://travis-ci.org/crawl/crawl/builds/487165728
19:50:29 -!- amalloy is now known as amalloy_
19:53:29 <geekosaur> well. this is fascinating.
19:53:50 <geekosaur> ^O shows Vaults entrances on D:12 and D:13. The one on D:12 is not obvious.
19:54:06 <geekosaur> go to D:12, ^GV0, "You're already here!"
19:54:13 <geekosaur> this sounds confused.
19:55:33 <geekosaur> (anywhere on the level, apparently)
20:06:23 <advil> !source _place_specific_trap
20:06:24 <Sequell> 1/1. https://github.com/crawl/crawl/blob/master/crawl-ref/source/dungeon.cc#L93
20:06:39 <advil> ebering: do you happen to know why that is an error mprf?
20:07:00 <advil> is it not supposed to happen or something?
20:07:37 <advil> geekosaur do you have a save?
20:08:04 <geekosaur> I can make one, although I'm in  Snake right now
20:09:33 <advil> that's probably fine
20:13:43 <geekosaur> hm, hadn't updated since the branch either
20:15:41 <Cheibriados> crawl thinks all of D:12 is a Vaults entrance? 13https://crawl.develz.org/mantis/view.php?id=11850 by geekosaur
20:19:19 -!- amalloy_ is now known as amalloy
21:06:40 <alexjurkiewicz> !source _make_level
21:06:40 <Sequell> 1/1. https://github.com/crawl/crawl/blob/master/crawl-ref/source/files.cc#L1244
21:06:52 <alexjurkiewicz> is it possible for abyss to have portals to pan?
21:07:01 <alexjurkiewicz> hm i guess not. That was something a fork had
21:19:07 <alexjurkiewicz> advil: just noticed this "rng_generator" = random number generator generator
21:23:44 <alexjurkiewicz> is there a better way to write this? level_effect_type level_effect = static_cast<level_effect_type>(random2(level_effect_type::COUNT))
21:51:42 <Cheibriados> 03advil02 07* 0.24-a0-20-g05a21ba: Various fixes for more stable lua iteration order 10(2 minutes ago, 7 files, 54+ 28-) 13https://github.com/crawl/crawl/commit/05a21ba7acd5
21:51:56 <advil> alexjurkiewicz it's really a sub-generator
21:53:35 <advil> apparently nethack 4 has 209 sub-generators
21:58:51 <gammafunk> alexjurkiewicz: before you go too crazy with implementing that, you probably want to put something somewhere for devs to see what you're actually planning to implement effect-wise
21:59:09 <gammafunk> because I'm guessing there will be some that we'd not merge in the first place
21:59:24 <advil> %git a39c87378c
21:59:24 <Cheibriados> 07ebering02 * 0.23-b1: Changelog through 0.23-a0-899-gbbae0a335a 10(3 days ago, 1 file, 29+ 3-) 13https://github.com/crawl/crawl/commit/a39c87378c14
22:01:09 <advil> geekosaur: this wasn't a pregenerated dungeon was it? (Your version # is after that was merged)
22:02:11 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-20-g05a21ba7ac (34)
22:05:32 <floraline> doesn't the save file indicate that?
22:05:47 <advil> I didn't add any ui element to say
22:05:52 <advil> it's probably a good idea though
22:06:39 <advil> though going to a new level builds it, so I guess not
22:06:46 <geekosaur> no, unless it somehow did that by default. I didnt secify  seed or anything
22:08:56 <advil> something is off with your travel cache, but it will clear up if/when you eventually enter the vaults
22:09:59 <advil> it doesn't seem to be about the level or anything like that
22:30:00 <advil> !source dungeon.cc:5909
22:30:00 <Sequell> https://github.com/crawl/crawl/blob/master/crawl-ref/source/dungeon.cc#L5909
22:30:16 <advil> ebering my question earlier was supposed to be about that ^
22:31:42 <advil> I might ditch that mprf, it doesn't seem to actually be an error really, in that there are plenty of vaults that can place a shaft and can be at a branch end
22:32:39 <advil> maybe it's an mprf error because 4 years ago wheals downgraded it from an ASSERT
22:33:10 <advil> %git c068da2d1
22:33:10 <Cheibriados> 07wheals02 * 0.16-a0-794-gc068da2: Be a little less violent about invalid traps. 10(4 years, 4 months ago, 1 file, 3+ 2-) 13https://github.com/crawl/crawl/commit/c068da2d193e
22:36:05 <advil> %git 7c65259dd8
22:36:05 <Cheibriados> 07wheals02 * 0.16-a0-730-g7c65259: Handle vaults placing shafts in branch ends more gracefully. 10(4 years, 4 months ago, 1 file, 3+ 0-) 13https://github.com/crawl/crawl/commit/7c65259dd833
22:36:16 <advil> ok, at that point I guess it was considered a bug
22:48:01 <gammafunk> ??tournament
22:48:01 <Sequell> tournament[1/2]: See {past tournaments} for links to previous tournaments.
22:48:07 <gammafunk> ??tournament[2
22:48:07 <Sequell> tournament[2/2]: Repository of tournament code: https://github.com/crawl/dcss_tourney
22:50:20 <gammafunk> !learn add tournament[1] The 0.23 tournament will run from 20:00 UTC Friday 8 February 2019 to 20:00 UTC Sunday 24 February 2019. Tournament home page: https://crawl.develz.org/tournament/0.23/
22:50:20 <Sequell> tournament[1/3]: The 0.23 tournament will run from 20:00 UTC Friday 8 February 2019 to 20:00 UTC Sunday 24 February 2019. Tournament home page: https://crawl.develz.org/tournament/0.23/
22:51:43 <advil> oh, heh, for some reason shafts were disallowed in tests (??) and that's why I'm seeing that mprf all over the place
22:54:32 <gammafunk> !time
22:54:33 <Sequell> Time: Feb 01, 2019, 03:54:32 AM, UTC. The 2019 0.23 tournament starts in 7 days, 16 hours, 5 minutes and 27 seconds.
22:54:52 <advil> are we doing any rules changes?
22:55:02 <advil> there seemed to be a lot of good ideas around after the last one
22:56:09 <alexjurkiewicz> gammafunk: yeah. I'm planning to just build a skeleton and implement a couple of prototype effects
22:56:45 <gammafunk> 💯
22:57:18 <alexjurkiewicz> emoji from the funk? what
23:00:14 <gammafunk> emoji are just uncool emotes don't @ me
23:02:01 <Cheibriados> 03advil02 07* 0.24-a0-21-ge4bce67: Allow shafts in tests 10(58 seconds ago, 1 file, 1+ 4-) 13https://github.com/crawl/crawl/commit/e4bce670a794
23:03:59 <alexjurkiewicz> what would be the right place to store metadata about a level? eg the level effect
23:04:11 <alexjurkiewicz> my hacky implementation is storing it in you, which is definitely wrong
23:07:55 <gammafunk> I believe each level has a props hash
23:08:05 <gammafunk> hrm, that may be grid-based
23:08:31 <gammafunk> lucy's level corruption would be the model we want, probably
23:12:09 <gammafunk> I guess env.map_markers
23:12:26 <gammafunk> that is position-based though
23:13:17 <gammafunk> ah, yeah but that's probably what you want
23:14:09 <gammafunk> or no
23:14:13 <gammafunk> env.properties
23:14:28 <gammafunk> alexjurkiewicz: env.properties is a per-level CrawlHashTable
23:15:06 <gammafunk> if you wanted something position-based you can use env.map_markers
23:17:15 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-21-ge4bce670a7 (34)
23:17:18 <alexjurkiewicz> env.properties sounds perfect then
23:17:24 <alexjurkiewicz> thanks for the digging
23:29:14 <alexjurkiewicz> !source grd
23:29:15 <Sequell> Can't find grd.
23:29:49 <alexjurkiewicz> !
23:31:13 <alexjurkiewicz> how does this witchcraft work
23:31:46 <gammafunk> it's a define; try it with ctags or something equiv in your code editor
23:31:53 <gammafunk> it's a define for env.grid
23:33:44 <alexjurkiewicz> aha, searching just in *.h found it
23:36:58 <advil> omg don't add any more dungeon builder state to you
23:37:30 <advil> wonder what is with all the "Multiple branch exits on Tomb:1" errors
23:40:54 <Cheibriados> 03advil02 07* 0.24-a0-22-g2a401c0: Clear more global state for tests 10(32 seconds ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/2a401c07d103
23:47:27 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-22-g2a401c07d1 (34)
23:59:17 <alexjurkiewicz> yes. I can turn all doors into blood fountains
23:59:43 <alexjurkiewicz> probably this is good enough to be a real effect. It adds more blood fountains to the game, which is a net positive any way you look at it