00:26:53 <alexjurkiewicz> advil: a problem with pregen_dungeon is that morgues contain vaults for every floor, eg https://crawl.project357.org/morgue/Ihryk/morgue-Ihryk-20190205-050951.txt
00:27:11 <alexjurkiewicz> advil: I guess this is technically information leakage, but mostly it's just morgue inflation
00:32:08 <Eksell> Unstable branch on CRAWL.XTAHUA.COM updated to: 0.24-a0-67-ga78bd38 (34)
00:34:05 <gammafunk> wow
00:34:11 <gammafunk> that morgue offends me
00:34:16 <gammafunk> zero gammafunk vaults!!!
00:36:00 <Eksell> Stable (0.23) branch on CRAWL.XTAHUA.COM updated to: 0.23-b1-67-g8bd44de
00:55:15 <alexjurkiewicz> and two chequers ones. That's the good part of adding a serial. A million new tiny minivaults
03:15:00 <Cheibriados> wrong description on x-v monster's fire storm range information 13https://crawl.develz.org/mantis/view.php?id=11853 by drifter255
03:32:12 <Kelbell> Stable (0.23) branch on crawl.kelbi.org updated to: 0.23-b1-67-g8bd44de8e4
06:32:32 <Lantell> feta (L7 MfGl) ERROR in 'tileweb.cc' at line 219: Socket write error: Resource temporarily unavailable (D:3)
07:40:51 <Lantell> Webtiles server stopped.
07:41:17 <Lantell> Webtiles server stopped.
07:41:18 <Lantell> Webtiles server started.
08:19:14 <Lantell> TheWayOfAllFlesh (L22 GnTm) Crash caused by signal #15: Terminated (Depths:3)
09:19:37 <advil> oops
09:19:43 <advil> definitely an info leak if seeds are shown
09:19:59 <advil> also, wish I'd known that when testing things
09:20:11 <advil> !crashlog TheWayOfAllFlesh
09:20:15 <Sequell> 1. TheWayOfAllFlesh, XL22 GnTm, T:70568 (milestone): https://underhound.eu/crawl/morgue/TheWayOfAllFlesh/crash-TheWayOfAllFlesh-20190205-131913.txt
09:20:47 <advil> hm something on the new layers view on webtiles maybe?
10:06:43 <advil> this is an annoyingly complicated bug to apply a quick fix to
10:07:04 <advil> the info I need for it is currently saved with the level
11:11:04 <gammafunk> ebering: I was going to copy your Fair Crawling section from csdc to become a section in the tournament rules
11:11:28 <gammafunk> do you have any objection to that or any changes/edits you want to make to it?
11:23:19 <Lantell> Stable (0.23) branch on underhound.eu updated to: 0.23-b1-67-g8bd44de8e4
11:30:42 <advil> maybe I just need to give in and add a "level_visited" prop on the player, was trying to avoid that
11:33:34 <gammafunk> to handle out of order level visitation for pregenerated dungeons?
11:33:58 <advil> it's for the morgue bug that alexjurkiewicz pointed out, but might help other things too
11:34:31 <advil> right now, the information is stored in that they have 0 turns on levels that they haven't visited
11:34:41 <advil> but that is saved in an awkward way and only available as env.turns_on_level when a level is currently loaded
11:35:05 <gammafunk> ah, yeah
11:35:14 <advil> so to better control vault printing for the morgue, I'd need to reload every level, which is a pain
11:35:35 <gammafunk> I noticed that certain levels were excluded from the listing; I guess that's due to what specific levels get pregenerated?
11:36:13 <gammafunk> zot:5 was excluded, shoals:1-3 also, but shoals:4 included
11:36:26 <advil> huh, weird
11:36:30 <gammafunk> wasn't sure why that particular pattern was present
11:36:34 <advil> that might be an independent bug in the crashdump code
11:36:48 <advil> pregen should be generating those
11:36:52 <gammafunk> oh, maybe no vaults selected? hrm, three shoals levels with no vaults seems odd
11:37:08 <gammafunk> I know hall of zot on zot:5 has no_dump
11:37:20 <advil> oh I bet I know
11:37:20 <gammafunk> so I could see that one simply not having any vaults
11:37:23 <advil> yeah
11:37:30 <advil> this listing doesn't show layout, only vaults
11:37:47 <gammafunk> so they must have gotten three shoals levels that just didn't have vaults
11:37:49 <advil> so given some layouts it's entirely possible for a level to lack vaults
11:38:02 <gammafunk> sort of makes sense as there are probably not so many shoals vaults, but that seems a bit surprising
11:38:12 <gammafunk> maybe that layout has a hard time placing them in the first place
11:38:47 <advil> so for example seed 1 has only layouts for crypts 1-2
11:39:05 <advil> I also wouldn't want to go to bat for the completeness of what gets stored in you.vault_list
11:40:57 <advil> e.g. I suspect that hall_of_Zot_1 etc aren't there
11:41:03 <gammafunk> yeah
11:41:16 <gammafunk> there are also some strange instances of layouts actually showing up in map/kmap fields
11:41:24 <gammafunk> !lg * map~~layout x=map
11:41:33 <advil> I guess I can test this by doing a morgue for seed 1, my travis test prints this stuff exhaustively
11:41:50 <Sequell> 328. [map=layout_vaults] FantomFang the Phalangite (L18 MfDK of Yredelemnul), quit the game on Vaults:1 (layout_vaults) on 2013-08-26 05:30:22, with 227012 points after 52279 turns and 5:01:23.
11:41:58 <gammafunk> ah, interesting
11:42:05 <gammafunk> !lg * kmap~~layout x=kmap
11:42:06 <Sequell> 2480. [killermap=layout_roguey; special_room_mythical_zoo] Zabinka the Archer (L13 CeHu of Makhleb), slain by a manticore (kmap: layout_roguey; special_room_mythical_zoo) on Lair:5 on 2019-02-04 17:05:04, with 37599 points after 23075 turns and 1:29:27.
11:42:11 <gammafunk> there we go
11:42:16 <advil> ah
11:42:34 <advil> for vaults in particular the layout/vault distinction is even more confusing
11:43:07 <gammafunk> right, I can't actually use transporters in Vaults due to how the layout places those rooms
11:43:23 <gammafunk> I need to look into that, but it causes the level to fail connectivity tests
11:44:16 <advil> do you know if there's any principled reason not to show layouts in morgue files?
11:44:26 <advil> it might just have been the easiest way to implement this
11:45:24 <gammafunk> hrm
11:45:32 <gammafunk> I suppose it wouldn't add all that much to the listing
11:45:41 <gammafunk> and might actually help identify problems
11:45:47 <advil> I wonder if I should instead redo dump_vault_maps to do what vault_catalog.lua does
11:45:57 <gammafunk> !source vault_catalog.lua
11:45:57 <Sequell> https://github.com/crawl/crawl/blob/master/crawl-ref/source/test/vault_catalog.lua
11:46:41 <gammafunk> heh, Elf generated after Hell in the generation order
11:46:56 <gammafunk> seems a little odd since it's a subbranch of Orc
11:46:59 <advil> yeah, after the vestibule at least
11:47:04 <amalloy> truly the most dangerous branch
11:47:48 <advil> the generation order I went with is a stripped-down version of what bel originally proposed
11:48:34 <advil> it has some similarities to how a player might traverse the dungeon, but those aren't really crucial to get right in a very detailed way
11:50:27 <advil> I think he probably went with that order because vestibule is a depths sub-branch, which seemed ok to me; elf is later than orc based roughly on when a typical game might enter it
11:50:28 <gammafunk> yeah, Elf would more typically be pre-Vaults or post-Vaults
11:50:33 <gammafunk> but certainly pre-Depths
11:50:42 <gammafunk> not that it matters much, like you say
11:51:27 <advil> yeah, definitely not a statement on what the player should be doing
11:52:29 <gammafunk> what would be the difference between the two approaches? I see that catalog_dungeon_vaults() is just printing the set for each level in that given level order
11:53:00 <gammafunk> which is not the same order as the branch iterator we have that I think goes by absdepth
11:53:58 <advil> dump_vault_maps uses you.vault_list, which is global, doesn't require loading the levels, and is a subset of what debug.vault_names() (which does require loading a level) gets
11:54:53 <advil> !source level_vault_names
11:54:54 <Sequell> 1/1. https://github.com/crawl/crawl/blob/master/crawl-ref/source/dbg-util.cc#L40
11:55:50 <advil> that's what debug.vault_names uses, it's the full list populated by levelgen and stored in each level chunk
11:57:53 <advil> .huh, maybe you.vault_list is only used for morgue generation
11:59:37 <gammafunk> yeah
11:59:45 <gammafunk> I was just trying to track down what you.vault_list is
11:59:54 <gammafunk> and I only see it being used for morgue dumps
12:02:10 <gammafunk> so I guess you could say that vault_names is refering to the list actually used for level gen purposes
12:02:25 <gammafunk> and you.vault_list is what we want to report to the user
12:02:51 <gammafunk> but I guess that this is a bit murky since the only users are people doing things with dev and testing
12:04:23 <advil> what I might do given this is just change how you.vault_list is populated, it doesn't need to be populated by the level builder
12:04:37 <advil> I was assuming it was used somehow in levelgen, like the other misc vault things on you
12:06:30 <Henzell> Unstable branch on crawl.akrasiac.org updated to: 0.24-a0-67-ga78bd38 (34)
12:09:42 <advil> except this is used in a way I don't understand for marshalling/unmarshalling map data
12:13:21 <advil> oh, this is just saving that table
12:29:26 <advil> why would a map be tagged with no_dump?
12:30:07 <advil> these are dummy maps of some sort?
12:32:52 <Cheibriados> New branch created: pull/977 (1 commit) 13https://github.com/crawl/crawl/pull/977
12:32:52 <Cheibriados> 03NormalPerson702 07https://github.com/crawl/crawl/pull/977 * 0.24-a0-68-g74d4781: Fix dancing weapon description when multiple monsters enter LOS (11843) 10(6 minutes ago, 1 file, 3+ 1-) 13https://github.com/crawl/crawl/commit/74d478157905
12:45:46 <advil> oof just found a last-minute seed stability bug that I'll need to fix tonight
12:48:50 <gammafunk> advil: no_dump for things that always place, like hall of zot on zot:5 or whatnot
12:48:50 <advil> ah ok
12:50:48 <advil> for now I'm going to show layouts in this listing
12:55:45 <advil> the other related bug I need to address us X] with pregen
12:57:26 <Cheibriados> 03advil02 07[stone_soup-0.23] * 0.23-b1-68-g643a44d: For pregen morgues, only show visited levels (alexjurkiewicz) 10(3 minutes ago, 5 files, 13+ 27-) 13https://github.com/crawl/crawl/commit/643a44d428fe
12:57:47 <advil> heh wrong branch
12:57:51 <advil> I always do that
12:58:27 <Cheibriados> 03advil02 07* 0.24-a0-68-gd6fbd69: For pregen morgues, only show visited levels (alexjurkiewicz) 10(4 minutes ago, 5 files, 13+ 27-) 13https://github.com/crawl/crawl/commit/d6fbd69efe4d
12:58:40 <gammafunk> yesterday I made the mistake of typing git rebase -i origin/stone_soup-0.13 instead of git rebase -i orgin/stone_soup-0.23
12:58:40 <gammafunk> big mistake
12:58:40 <gammafunk> git was busy for a while
12:58:40 <advil> lol
12:58:42 <gammafunk> *origin
13:04:50 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-68-gd6fbd69efe (34)
13:20:02 <switch___> re NP7's trove PR
13:20:17 <switch___> even if you're doing an identity rebalance in the near future, why not merge it as-is anyway?
13:20:38 <switch___> unless you disagree with the effects proposed
14:18:45 -!- aidanh_ is now known as aidanh
14:49:28 <amalloy> |amethyst: your lovely retelling of the fable of stone soup was so popular that someone reported it to the mods, with reason "why is this stupid prattle voted up so high"
14:50:25 <amalloy> to be clear i do not endorse or agree with that assessment, it's just so absurd i thought you might join me in finding it a funny complaint
14:52:26 <travis-ci> The build has errored. (master - d6fbd69 #11080 : advil): https://travis-ci.org/crawl/crawl/builds/489165886
15:15:25 <gammafunk> was the reporter maybe a longtime reddit troll
15:15:27 <gammafunk> that'd be my guess
15:22:05 <n1> no it wasn't me
15:22:27 <gammafunk> Yeah I realize
15:22:36 <gammafunk> I have a special list of things to blame you for
15:22:47 <gammafunk> thanks again for global warming btw
15:23:04 <n1> yeah that one was my bad
15:37:08 <alexjurkiewicz> 
15:38:32 <NP7> what about that space alexjurkiewicz?
15:39:28 <alexjurkiewicz> oops
15:48:54 <floraline> i still don't know exactly what i should be doing when checking out different branches in my nightly build script
15:49:10 <floraline> every once in a while i find out there is some thing git complains about that blocks a certain build for days
15:49:54 <floraline> i'm just calling update-stable one after the other but apparently that's not good enough for every situation
15:50:16 <Kelbell> Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-1672-gc654023060
15:51:03 <gammafunk> floraline: can you be more specific re: git complaints
15:54:43 <Kelbell> Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-1672-gc654023060
16:02:57 <floraline> here's the context: https://pastebin.com/rjD4M1dZ
16:03:04 <floraline> the build log on the bottom indicates that i'm in the middle of the script at the top
16:03:31 <floraline> i think it's failing when it tries to git checkout the branch it will build, because there are some unexpected local changes
16:04:28 <floraline> i think i would want to change git checkout to use -f but this is a standard script that is part of dgl-config repo
16:04:48 <floraline> so i'm trying to figure out why i'm seeing this and what a reasonable fix would be
16:05:37 <floraline> the overnight script has just checked out a couple different branches prior to failing on bcrawl
16:05:41 <gammafunk> hrm
16:05:55 <floraline> it went from 0.23 to trunk to hellcrawl to bcrawl
16:06:06 <floraline> somewhere that weird untracked thing popped up
16:06:13 <gammafunk> yeah, it looks like
16:06:24 <gammafunk> it's trying to check out in the same repo as trunk?
16:06:34 <floraline> different remotes
16:07:03 <gammafunk> well, I think that's possible the issue then..it's all the same repo?
16:07:04 <floraline> master hellcrawl and bcrawl are all tracking different remotes in my local repo
16:07:19 <gammafunk> ideally you would have distinct repos for each fork I guess
16:07:21 <gammafunk> hrm
16:07:42 <gammafunk> ah
16:07:42 <gammafunk> yeah
16:07:53 <gammafunk> floraline: so for us, species-type.h is autogenerated, not in the repo
16:08:04 <gammafunk> for those forks don't have e.g. datafiles branch stuff merged
16:08:10 <gammafunk> those headers are still in their repo
16:08:11 <floraline> oh i see
16:08:15 <floraline> untracked here, tracked there
16:08:19 <gammafunk> right
16:08:23 <floraline> and it doesn't want me to lose my changes
16:08:34 <gammafunk> either script needs to be prepared to force overrite local "changes"
16:08:41 <gammafunk> or you need to move those to distinct repos
16:08:44 <floraline> ok, that's what i want to do
16:08:54 <floraline> i want every branch to totally disregard any local changes on every build
16:08:58 <gammafunk> right
16:09:05 <floraline> can i just change the git checkout to git checkout -f and that will let it do that?
16:09:22 <floraline> or should i be using some kind of git reset or other command somewhere here
16:09:46 <gammafunk> yeah, that might be the best way, not sure if it has other implications
16:09:53 <gammafunk> I often do e.g. git reset --hard <target>
16:10:01 <floraline> i would be happy to just rm -rf the whole thing and re-clone and build
16:10:06 <floraline> so whatever git commands are the equivalent of that, i'm fine with
16:10:07 <floraline> i think
16:10:16 <gammafunk> yeah you shouldn't do that command manually I think
16:10:21 <gammafunk> maybe try git checkout -f
16:10:38 <gammafunk> although I wonder
16:11:15 <gammafunk> so what happens if a fork has file that influence your build of say trunk
16:11:18 <gammafunk> or vice versa
16:11:32 <gammafunk> but that aren't tracked in one branch
16:12:08 <floraline> i thought checkout would handle such a case
16:12:29 <amalloy> reset --hard still won't touch untracked files, i think
16:12:44 <amalloy> if you want to nuke it from orbit that's what git clean -df does, but it is a pretty scary command
16:13:06 <floraline> i think i'm comfortable with the effects of nuking from orbit, but i would prefer to use a more graceful git equivalent
16:16:05 <amalloy> floraline: perhaps https://stackoverflow.com/a/7588333/625403 ?
16:16:32 <geekosaur> if you haven't locally committed, there's git stash
16:16:49 <amalloy> since i think the only problem is when a data-files fork generates a file but doesn't commit it (as it shouldn't), and then you try to check out a fork that has a version committed
16:17:01 <amalloy> geekosaur: i don't think that's such a great plan for auto-generated changes that will be re-generated with every build
16:18:27 <gammafunk> yeah, it seems that if you checkout a branch with a file that another branch doesn't have, when you switch back said file is gone, which I guess makes the most sense
16:18:47 <amalloy> gammafunk: right, and that's even the behaviour we want in this case
16:18:54 <gammafunk> yes
16:19:08 <gammafunk> but I suppose the problem might be if a fork generated a file that is not tracked that we likewise don't track
16:19:13 <gammafunk> and this file influences crawl
16:19:18 <floraline> right
16:19:21 <floraline> that sounds like a concern
16:19:41 <gammafunk> this is sort of why maybe they should have their own repos I guess
16:19:49 <amalloy> nah, i think they should share a repo
16:19:58 <gammafunk> and just git clean?
16:20:00 <amalloy> this is what git-clean is for
16:20:24 <gammafunk> hrm, not sure since this is an expensive operation, right?
16:20:38 <amalloy> i mean i guess if generation is expensive enough that you're willing to pay for N copies of the repo on your disk
16:20:55 <gammafunk> well there are few forks
16:21:01 <gammafunk> and few cpus
16:21:13 <gammafunk> but yeah I'd defer to someone with more knowledge about how to best do this
16:21:21 <gammafunk> these repos really are not related to each other
16:21:34 <gammafunk> is there actually a benefit to having them as remotes in one repo?
16:21:41 <floraline> that was the first way that i got to work
16:21:42 <gammafunk> aside from possibly disk savings (which I assume is no big issue)
16:21:52 <floraline> that's the only reason i'm doing it this way
16:22:12 <gammafunk> sounds like there might be some tradeoffs either way; don't know if your scripts can work ok with multiple repos
16:22:25 <gammafunk> but sounds like you have a plan for either approach you want to take, at least
16:22:26 <amalloy> actually i bet you could get disk savings even with N copies, by cloning them from each other instead of directly cloning each from upstream
16:23:00 <amalloy> the default behaviour there is to create a hardlink instead of copying the file
16:24:21 <gammafunk> maybe git clean is perfectly fine
16:24:36 <gammafunk> it's not like it's deleting all files and recreating, just stuff that's not tracked
16:27:27 <floraline> i think git clean sounds most attractive right now
16:30:07 <amalloy> i think git clean -dx is the best fit for what you want to do. -dfx would probably be fine but seems a bit overkill; i wouldn't turn that on unless i saw a case where -dx wasn't enough
16:31:37 <amalloy> well, maybe i'm wrong. maybe -dx makes no sense with -f. i dunno. the man page seems to say that without -f not much will happen
16:37:08 <gammafunk> yeah I think you may indeed need git clean -dfx
16:38:09 <gammafunk> hrm
16:38:14 <gammafunk> this almost needs like you need -f twice
16:38:48 <gammafunk> If an untracked directory is managed by a different Git
16:38:49 <gammafunk> repository, it is not removed by default. Use -f option twice if you really want to remove such a directory.
16:39:14 <gammafunk> oh, what does "different git repository" mean
16:39:34 <gammafunk> that's referring to when it finds a .git dir at some point in the tree?
16:49:18 <floraline> the only thing i don't like is some script is putting archives of the source for each branch into the source directory, git clean will remove those archives
16:49:47 <floraline> i'm assuming they are useful somehow for debugging in some certain situations, so i would like to keep them. i need to figure out whatever is responsible for that
16:50:13 <floraline> other than that i think git clean -xdf is the way to go, i can add that as a step prior to git checkout
16:52:15 <amalloy> gammafunk: right
17:02:34 <Cheibriados> 03advil02 07* 0.24-a0-69-g55f8356: Invalidate LOS properly for level pregeneration 10(13 minutes ago, 2 files, 9+ 9-) 13https://github.com/crawl/crawl/commit/55f835606b54
17:13:24 <alexjurkiewicz> floraline:you might want to use git worktrees to deal with your duplication issue
17:13:41 <alexjurkiewicz> less expensive than multiple copies of the repo, but you get to keep a cache of build objects for each branch you build
17:13:59 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-69-g55f835606b (34)
17:38:09 <floraline> it looks like submodules are not compatible with worktrees
17:38:11 <floraline> that sounds interesting though
17:42:52 <floraline> thank you everyone for the advice, i'm still a git novice
17:43:19 <alexjurkiewicz> I use worktrees for all branches here, seems to be fine
17:44:19 <gammafunk> I seem to have weird problems with ccache seemingly not recognizing stuff that's cached when I build from a different repo
17:45:16 <alexjurkiewicz> advil: are portals the same in seeded pregen?
17:45:37 <Cheibriados> 03advil02 07[stone_soup-0.23] * 0.23-b1-69-g26107c8: Invalidate LOS properly for level pregeneration 10(56 minutes ago, 2 files, 9+ 9-) 13https://github.com/crawl/crawl/commit/26107c8d21c9
17:45:38 <advil> good question
17:45:38 <alexjurkiewicz> wondering about bazaar items and trove entry fees
17:45:38 <advil> I haven't even really tested
17:45:39 <advil> oh, the entry vaults should be
17:45:51 <alexjurkiewicz> right, the trove entry fee is lua on the master level, so that would be the same
17:45:55 <advil> the portals themselves are not pregenerated, but do use isolated rngs
17:46:20 <alexjurkiewicz> ah, so "yes if you are on the same platform, but maybe not if you compare windows/linux/osx"?
17:46:35 <advil> os shouldn't matter for that one
17:46:40 <advil> I would expect them to be the same as long as levelgen for the portals doesn't do anything that interacts with vault chance elsewhere, which might be true for troves at least
17:47:01 <advil> if some portal vaults can place a unique, that kind of thing, then you might get interactions
17:47:39 <gammafunk> nothing would limit a portal from placing a unique although i'm not sure any actually do
17:47:50 <advil> snorg can place in the sewers, but that one shouldn't matter
17:48:04 <gammafunk> right, and purgy
17:48:09 <advil> er, maybe I meant purgy
17:48:17 <advil> it's really if portal vaults are fighting over the same unique, then entering them in a different order might matter
17:48:22 <advil> pretty edge case stuff
17:48:24 <floraline> gammafunk: how do you know that ccache is misbehaving? if the file is changed shouldn't it miss the cache by design?
17:48:32 <advil> but I have also tested this 0.0 times
17:48:44 <gammafunk> floraline: well it's weird, just seems like my build in the other repo should be hitting cache
17:48:48 <advil> and this whole seeding thing is very much, a butterfly flaps its wings, and then...
17:48:50 <gammafunk> but it's not, it's fully recompiling
17:49:13 <gammafunk> time to reintroduce chaos butterfulies then?
17:49:22 <advil> heh
17:49:28 <gammafunk> *butterflies
17:49:43 <gammafunk> we need a unique that starts pulling in random vaults dynamically
17:49:49 <gammafunk> really put advil's work to the test
17:50:13 <alexjurkiewicz> advil:great to know, thanks
17:50:28 <advil> a xom unique that breaks the fourth wall and wrecks the seeding
17:50:56 <alexjurkiewicz> The chaos caller re-configures the level!
17:51:15 <alexjurkiewicz> a unique with lugonu's corrupt
17:51:33 <advil> maybe just a straight up xom effect that runs &ctrl-r
17:51:44 <advil> when you enter a new level
17:55:00 <alexjurkiewicz> you're joking, but that sounds like a fun addition to seeded challenge modes. 5% chance for each level to be unique only for you
17:55:29 <gammafunk> can you do that and make the seed "resume" to the shared results
17:55:52 <advil> with pregeneration yeah, that should be possible
17:57:07 <advil> man travis failed on this one random side and even with the seed # I just can't replicate it
18:16:30 <travis-ci> The build passed. (master - 55f8356 #11081 : advil): https://travis-ci.org/crawl/crawl/builds/489268661
18:24:03 <Lantell> Unstable branch on underhound.eu updated to: 0.24-a0-69-g55f835606b (34)
18:39:39 <advil> ok I can replicate it but it takes a 10 minutes long incredibly specific sequence of events
18:40:43 <amalloy> that's exciting
18:43:31 <advil> it's been how a lot of these bugs go, but this one has resisted bisecting
18:54:43 -!- nosuni is now known as name
19:10:55 <travis-ci> The build has errored. (stone_soup-0.23 - 26107c8 #11082 : advil): https://travis-ci.org/crawl/crawl/builds/489286702
19:17:46 <alexjurkiewicz> I'm trying to add a new file to compilation, but it's not being compiled. Do I need to add the file to a list somewhere? https://gist.github.com/alexjurkiewicz/401e66fca7f3c232e3ddd90a2bfea505
19:18:30 <amalloy> !source Makefile
19:18:30 <Sequell> https://github.com/crawl/crawl/blob/master/crawl-ref/source/Makefile
19:18:33 <advil> Makefile.obj?
19:18:57 <|amethyst> alexjurkiewicz: Makefile.obj, MSVC/crawl.vcxproj, and android-project/jni/src/Android.mk
19:19:08 <alexjurkiewicz> aha
19:19:13 <|amethyst> if you forget the last two, no one will notice until they do
19:19:37 <alexjurkiewicz> :D
19:21:33 <gammafunk> alexjurkiewicz: if you want to know were, just do a "git grep FILENAME" for an existing filename in the codebase
19:21:38 <gammafunk> a c++ one, that is
19:21:48 <gammafunk> s/were/where/
19:22:00 <gammafunk> that's what I always do
19:22:24 <alexjurkiewicz> works, thanks
19:32:27 -!- jfcaron_ is now known as jfcaron
20:08:26 -!- amalloy is now known as amalloy_
20:38:19 -!- amalloy_ is now known as amalloy
21:49:29 <alexjurkiewicz> json_append_member(branch_info, "levels", json_mkstring(std::to_string(it->numlevels).c_str())); <-- any better way to write this?
21:52:38 <gammafunk> is it->numlevels not already a string?
21:53:14 <gammafunk> oh, it's an int
21:54:09 <gammafunk> it seems a bit odd that levels is a string when what you're sending is an int
21:54:28 <gammafunk> I see json_mknumber(), for instance
21:55:12 <gammafunk> but I don't know the context of this data; why it's being sent
21:57:36 <alexjurkiewicz> i guess it makes sense to write as a number doesn't it
22:00:28 <Cheibriados> New branch created: pull/978 (1 commit) 13https://github.com/crawl/crawl/pull/978
22:00:28 <Cheibriados> 03alexjurkiewicz02 07https://github.com/crawl/crawl/pull/978 * 0.24-a0-70-gb03b054: Add -branch-json 10(16 hours ago, 7 files, 53+ 2-) 13https://github.com/crawl/crawl/commit/b03b05411e2b
22:24:27 <ebering> gammafunk: re fair crawling, sounds good
22:24:51 <ebering> might want to add some clarification in consult with some of the speedrunners
22:25:30 <ebering> there are concerns about whether or not macro tab to tab 8 times is too much of a technology assist to count as "normal play"
22:25:53 <ebering> it's a less clearcut case than the more significant upgrades of parabolic
22:26:02 <gammafunk> ah, yeah, I don't think we want to get into the business of policing that... but then we do mention cheating
22:26:05 <ebering> it's probably not something we want to police actively
22:27:47 <ebering> it's tough to draft a good guideline
22:28:11 <ebering> maybe "Macros that take multiple actions for the purpose of improving real-time speed are unsporting."
22:28:32 <ebering> maybe "the sole purpose"
22:35:04 <gammafunk> well, it's a little bit of an issue in the sense I'm not sure I want to encourage people to monitor/report that
22:35:27 <ebering> yeah
22:35:58 <ebering> might better be left as the "gentlemens agreement" among the top contenders
22:37:42 <gammafunk> yeah, it's something that we could at least mention
22:43:00 <alexjurkiewicz> i wouldn't add any rules about speedrunning. The speed running community has their own ruleset
22:43:17 <alexjurkiewicz> you can see the rules here if you click 'view rules' https://www.speedrun.com/Dungeon_Crawl_Stone_Soup
22:48:42 <ebering> well we give points for fastest realtime and personal best realtime
23:05:20 <ebering> !lg . won -3
23:05:21 <Sequell> 35/37. ebering the Meteorologist (L27 TeAE of Gozag), escaped with the Orb and 4 runes on 2019-01-21 00:46:24, with 1799226 points after 119932 turns and 11:58:56.
23:05:23 <ebering> oops
23:25:31 <ebering> !bug 11843
23:25:31 <Sequell> https://crawl.develz.org/mantis/view.php?id=11843
23:32:39 <Cheibriados> 03NormalPerson702 {ebering} 07* 0.24-a0-70-gf44d0c5: Fix dancing weapon description when multiple monsters enter LOS (11843) 10(11 hours ago, 1 file, 3+ 1-) 13https://github.com/crawl/crawl/commit/f44d0c5ad20c
23:34:45 <Cheibriados> 03NormalPerson702 {ebering} 07[stone_soup-0.23] * 0.23-b1-70-gd12f7cf: Fix dancing weapon description when multiple monsters enter LOS (11843) 10(11 hours ago, 1 file, 3+ 1-) 13https://github.com/crawl/crawl/commit/d12f7cfb269c
23:35:46 <alexjurkiewicz> wow, is ebering np7 or is he just 7 minutes faster than Cheibriados 🤔
23:36:54 <geekosaur> chei can lag by that much
23:37:21 <alexjurkiewicz> runed door bug i think https://www.reddit.com/r/dcss/comments/anmi88/i_fart_in_your_general_direction/
23:37:45 <ebering> glass bug
23:41:07 <ebering> slash awaken earth bug
23:41:15 <ebering> slash "intended behavior"
23:47:20 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-70-gf44d0c5ad2 (34)
23:51:35 <gammafunk> slash stupid remove