00:04:09 <geekosaur> noit sure it should be searching everything you've ever seen
00:06:42 <minmay> I guess staff acquirement is happy to give you exact duplicates all the time
00:17:47 <johnstein> gammafunk: some sort of shared memory quirk. |amethyst gave me a magic spell that works. I just gotta go find the file (I should turn it into a gist)
00:30:46 <gammafunk> johnstein: yeah, if you let me know what you actually did to fix the shared memory file, I can mention to zibudo, thanks
00:43:19 <johnstein> gammafunk:https://gist.github.com/jmbjr/35fa100da59a920460d2f077880c99d7
00:53:00 <johnstein> though I did this and still not working...
01:06:27 <johnstein> I'm bummed
01:06:38 <johnstein> jmbjr@www: (master *+%) ~$  ipcs -m
01:06:38 <johnstein> ------ Shared Memory Segments --------
01:06:38 <johnstein> key        shmid      owner      perms      bytes      nattch     status
01:06:38 <johnstein> 0x00000000 0          crawl      644        33648      3          dest
01:07:46 <johnstein> the dest status should mean that the shared memory block will be deleted once all processes close.  but why is it 0x00000000 shmid 0?
01:08:28 <johnstein> when I started tonight it was 0x52290cd6. every time I ipcrm -M  0x52290cd6, that one clears. but then when I test in console, it crashes again and the key shows back up
01:25:58 <johnstein> !tell |amethyst, on cao in dgldir, the inotify log is pretty huge. is there a safe maintenance process to address that? -rw-rw-rw-  1 games games    3526631830 Aug  6 22:24 crawl-inotify-where.log
01:25:59 <Sequell> Sorry johnstein, I don't know who |amethyst, is.
01:26:19 <johnstein> !tell |amethyst on cao in dgldir, the inotify log is pretty huge. is there a safe maintenance process to address that? -rw-rw-rw-  1 games games    3526631830 Aug  6 22:24 crawl-inotify-where.log
01:26:20 <Sequell> johnstein: OK, I'll let |amethyst know.
01:42:14 <johnstein> ??rebuild
01:42:14 <Sequell> rebuild[1/2]: http://crawl.akrasiac.org/rebuild/ http://underhound.eu:81/rebuild/ http://crawl.berotato.org/crawl/rebuild/ http://crawl.xtahua.com/rebuild/ https://crawl.jorgrun.rocks/rebuild/ Bug |amethyst or Nap.Kin for CDO. Use your powers wisely.
01:58:55 <Henzell> Stable (0.22) branch on crawl.akrasiac.org updated to: 0.22-b1-26-g9dbaab1
02:01:01 <johnstein> ok. this is weird. after I kill a test char on cao, the death screen isn't coming up
02:01:13 <johnstein> just freezes. on webtiles and console.
02:01:17 <johnstein> &versions
02:01:54 <johnstein> it's probably some stupid permissions issue I'm forgetting but I've gone through my notes several times
02:01:54 <Sequell> CAO: 0.23-a0-49-gd9f7a49, CBRO: 0.23-a0-49-gd9f7a49, CDO: 0.23-a0-49-gd9f7a49, CJR: 0.23-a0-49-gd9f7a49668, CPO: 0.23-a0-49-gd9f7a49, CUE: 0.23-a0-49-gd9f7a49668, CWZ: 0.22-a0-834-g05cf915, CXC: 0.23-a0-49-gd9f7a49, LLD: 0.23-a0-33-g1e3a2e4e2f
02:03:21 <johnstein> hm. working now.
02:05:25 <johnstein> oh weird. maybe it's been updated so that only 'return' will take you to the high score screen. I swear I used to be able to click any key or mouseclick.
02:05:36 <johnstein> cao should be ready to go
02:15:25 <johnstein> !tell gammafunk https://github.com/crawl/sequell/pull/77
02:15:26 <Sequell> johnstein: OK, I'll let gammafunk know.
02:23:44 <johnstein> !tell advil I sent you a link to the bones file in pm
02:23:45 <Sequell> johnstein: OK, I'll let advil know.
02:24:53 <alexjurkiewicz> johnstein: the inotify thing is probably only safe to kill by stopping all crawl games, then truncating the file
02:25:09 <alexjurkiewicz> but amethyst might have a less disruptive approach
03:02:14 <Rast2-> !gamesby mdonais
03:02:16 <Sequell> mdonais has played 941 games, between 2016-11-11 02:57:23 and 2018-08-06 15:36:15, won 46 (4.9%), high score 20104965, total score 190142861, total turns 12003006, play-time/day 1:50:15, total time 48d+13:00:58.
03:02:24 <Rast2-> !gamesby mdonais t
03:02:25 <Sequell> No games for mdonais (t).
04:38:35 <joy1999> hello. I found a bug.
04:38:53 <joy1999> ru's sacrifice health is raising piety abnormally.
04:39:29 <joy1999> Perhaps break in line 5311 of god-abil.cc is missing.
04:57:56 <joy1999> m
05:23:20 <Yermak> joy1999, you're right
05:23:29 <Yermak> you can report bugs on crawl.develz.org/mantis
05:33:57 <joy1999> thanks you
05:39:15 <Cheibriados> ru's sacrifice health is raising piety abnormally 13https://crawl.develz.org/mantis/view.php?id=11594 by joy1999
05:57:56 -!- TabMasher is now known as Apple
05:58:25 -!- Apple is now known as Guest62433
05:58:42 -!- Guest62433 is now known as TabMasher
06:40:43 -!- TabMasher is now known as CoffeeIV
06:44:26 -!- CoffeeIV is now known as FutureShock
06:56:19 -!- amalloy is now known as amalloy_
06:57:43 -!- amalloy_ is now known as amalloy
09:10:07 <Cheibriados> 03advil02 07* 0.23-a0-50-g799c399: Fix Ru sacrifice health (joy1999 / 11549) 10(84 seconds ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/799c399c3811
09:35:09 <advil> aidanh: http://dpaste.com/2Q7PVE4 we may need to find a workaround for 2f468e5a82e4 that doesn't call redraw_screen :-P
09:35:09 <Sequell> advil: You have 1 message. Use !messages to read it.
09:35:12 <advil> !messages
09:35:13 <Sequell> (1/1) johnstein said (7h 11m 29s ago): I sent you a link to the bones file in pm
09:35:38 <aidanh> impressive
09:36:50 <advil> it'll keep going
09:37:30 <advil> I was wondering why the dig targeter was entirely filling the message log when it did work (I didn't think I'd done it *that* ineffeciently) but the recursion seems to include the beam construction, so I guess that explains that
09:39:12 <advil> I think this recursion comes up if there's a badly timed more (guaranteed on fulldebug), so I wonder if it could explain the difficulty reproducing other recent targeter weirdness
09:40:23 <advil> it's sort of amazing that anything works, but it turns out that _send_map is accidentally structured so it works recursively as well as iteratively, because it carefull sets dirty flags as it proceeds
09:40:29 <advil> *carefully
09:41:40 <advil> btw the recursion doesn't actually include the redraw_screen call, so I think the conditions for it were already there
09:44:04 <advil> in fact, maybe it was already happening before that change?
09:44:22 <advil> the change might just be a multiplier
09:45:21 <advil> oh, no it is triggered by that, was misreading line numbers
09:47:52 <aidanh> right, the whole cascade was initiated by my innocent redraw_screen() call
09:51:01 <advil> yeah, it seems to be send_cell calling beam stuff, but the beam stuff leads to a more that calls tiles redraw
09:51:12 <advil> which then calls send_cell
09:51:39 <advil> so this might be triggerable by any badly timed more during direction chooser as well
09:52:08 <gammafunk> is direction_chooser at all to blame for this?
09:52:11 <gammafunk> cuz I hate direction_chooser
09:52:50 <advil> I barely know what to blame
09:54:15 <advil> I'm still amazed that this mostly works except in edge cases where the json string is too big or something, this is an incredibly intricate multiple-recursion calculation that if I were to try to write intentionally would take a ton of effort to get right
09:54:39 <aidanh> that stacktrace has all my favourite parts of crawl insanity: getch() causing a redraw, the message window, direction chooser, hell, it's even got beam.cc
09:55:26 <advil> yeah, I should let it go a few more steps and then |amethyst can post a screenshot on @crawlcode
09:58:02 <advil> although now I've done too many more and I can't get it to all fit on screen at once
09:59:56 <advil> |amethyst: https://imgur.com/a/fdZ1qEp @crawlcode candidate?
10:01:54 <advil> I'm sort of at a loss as to what a really good fix for this would be, one obvious thing is that _send_map should probably be inert to re-entry but it doesn't seem like that is the only problem here
10:02:34 <advil> possibly should suppress more / message window output in direction chooser too
10:03:01 <advil> might be hard to do without breaking the direction chooser though
10:03:21 <advil> I could finally delete that stupid beam dprf too
10:03:44 <advil> possibly the single most annoying thing about running a fulldebug build
10:06:19 <advil> I guess probably we need minimal intervention here given the release schedule, so I shouldn't ask questions like, why does getch trigger a redraw?
10:07:22 <aidanh> you really don't want to ask that question, regardless
10:08:17 <advil> heh
10:10:17 <Jorgrell> Unstable branch on crawl.jorgrun.rocks updated to: 0.23-a0-50-g799c399c38 (34)
10:11:25 <aidanh> while looking at the ?# issues, I realized that currently ?# lines are clipped on console, whereas before they were wrapped
10:11:53 <aidanh> is there any preferred behaviour here? the aptitudes table is semi-broken when wrapped, but info is lost if it's just clipped
10:12:40 -!- G-Flex| is now known as G-Flex
10:16:06 <advil> I think that screen might be designed to be wrapped
10:16:27 <advil> a non-stable version number is enough to push the first line over 80 chars
10:16:42 <advil> also the message log needs to be wrapped
10:20:43 <aidanh> er, I meant the action/skill tables; everything else is fine
10:26:45 <advil> ah
10:28:53 <aidanh> I could wrap the rest and allow horizontal scrolling, but that seems kludgy
10:29:51 <aidanh> an alternative is to split those tables into two chunks; the first for early-mid levels and the second for mid-late levels
10:29:56 <advil> looking at a real game example I don't think that table was designed for 80 chars at all
10:30:16 <advil> more for viewing in a browser
10:31:26 <gammafunk> I'm going to make the xp tracking table just track vault vs non-vault xp instead of spawns
10:32:05 <gammafunk> I don't think there's any reason why we'd really want to track spawn xp in the future, it would only be for spawns in pan (extremely minor and probably will become and unthing at some point) and spawns in the abyss
10:32:15 <gammafunk> and we don't really care about spawn xp from the abyss
10:32:29 <gammafunk> but if anyone has a problem with that, speak now or forever hold your peace
10:32:48 <aidanh> it clearly wasn't designed for it, but do console players use it?
10:33:25 <advil> people do look at that, judging from the bug reports I've seen
10:33:44 <advil> I think a lot of in game use of ?# is from players who take notes
10:34:18 <advil> I definitely look at my xp table while playing, but I always forget that ?# exists and use # to do it in a browser
10:35:09 <advil> so I guess it ought to be fixed somehow, maybe the table break solution (for ?# specifically, not chardumps in files) seems like the best solution?
10:35:21 <advil> that seems like the most annoying one to implement
10:35:21 <gammafunk> oh, well the xp table would still exist, to be clear
10:35:31 <gammafunk> oh I see what you mean
10:35:38 <gammafunk> yeah it would no longer dump before game end
10:35:44 <gammafunk> well I suppose it should?
10:35:46 <advil> yeah sorry not the xp gain table, the skill progression table
10:36:07 <advil> I'm just getting confused
10:36:07 <gammafunk> oh ok, yeah
10:36:17 <aidanh> perhaps a future UI project would be to move those to their own specific screen? a subscreen of % or m, perhaps
10:36:20 <advil> cleaning up the xp gain table seems good to me, it's sort of const now
10:36:47 <advil> er, spawn xp table
10:36:49 <aidanh> such a screen could scroll horizontally, perhaps; or be otherwise improved
10:37:06 <gammafunk> only change I'm planning is the vault counting since that could be useful for balance purposes, although even then that's a bit theoretical
10:37:17 <advil> or have ?# with submenus or something
10:37:17 <gammafunk> but certainly more useful than spawns at least
10:38:02 <aidanh> well, for now I'll just wrap the text
10:55:11 <advil> !source webtiles_send_last_messages
10:55:11 <Sequell> 1/2. https://github.com/crawl/crawl/blob/master/crawl-ref/source/message.cc#L934
10:56:44 <advil> %git 4297a6b4e
10:56:45 <Cheibriados> 07gammafunk02 * 0.17-a0-1797-g4297a6b: Removed unused code for designating old player messages in WebTiles 10(2 years, 11 months ago, 1 file, 2+ 5-) 13https://github.com/crawl/crawl/commit/4297a6b4eec1
10:57:16 <advil> gammafunk: I realize this was years ago, but any particular reason you didn't also remove the n parameter to webtiles_send_last_messages? it's unused now
10:58:56 <advil> I'll just collapse those two functions if not
11:06:04 <gammafunk> advil: no idea why I didn't
11:06:58 <advil> ok, it doesn't seem like there's any reason to send partial message backlogs any more so I'll just clean that up
11:07:39 <gammafunk> vaguely wonder what led me to investigate this code
11:18:38 <Cheibriados> 03Aidan Holm02 07* 0.23-a0-51-gdc02b5f: Fix monster abjuration description UI (NormalPerson7, #11589) 10(8 minutes ago, 2 files, 2+ 2-) 13https://github.com/crawl/crawl/commit/dc02b5f656dd
11:28:02 <aidanh> there's a mantis issue requesting an amnesia confirmation yesno; accidental amnesia doesn't matter with sif, but with scrolls perhaps a prompt is in order?
11:35:00 <gammafunk> hrm, but it prompts you for the spell in the first place, no?
11:35:17 <gammafunk> are they asking for confirmation even after you've selected a spell?
11:35:39 <gammafunk> maybe the memorization interface already works that way
11:36:04 <gammafunk> yeah I guess it does
11:36:32 <gammafunk> I suppose it would be fine to have said confirmation
11:36:53 <gammafunk> but would defer to anyone who had a strong opinion otherwise
11:39:24 <aidanh> here's the issue in question: https://crawl.develz.org/mantis/view.php?id=11581
11:40:06 <aidanh> they mentioned fat-fingering the UI as a concern
11:40:28 <gammafunk> yeah, the spell memorization interface prompts you
11:40:37 <gammafunk> so it seems it'd be consistent to do so for amnesia
11:40:53 <gammafunk> since it's a very similar problem in terms of potential loss of spell levels for a time
11:41:02 <gammafunk> if you only have a limited number of scrolls
11:41:23 <gammafunk> accidental amnesia with sif certainly also matters since that ability costs piety
11:53:54 <aidanh> ah, right
12:10:18 <Jorgrell> Unstable branch on crawl.jorgrun.rocks updated to: 0.23-a0-51-gdc02b5f656 (34)
12:42:25 -!- FutureShock is now known as Gwen_Dolin
12:47:25 <Cheibriados> 03advil02 07* 0.23-a0-52-gbb52cfb: Don't do extra map redraws in direction chooser 10(53 minutes ago, 3 files, 39+ 1-) 13https://github.com/crawl/crawl/commit/bb52cfba1566
12:47:25 <Cheibriados> 03advil02 07* 0.23-a0-53-g4ff6b17: Make some webtiles _send... calls inert to reentry 10(49 minutes ago, 4 files, 25+ 8-) 13https://github.com/crawl/crawl/commit/4ff6b1725dfc
12:47:25 <Cheibriados> 03advil02 07* 0.23-a0-54-gae6112e: Improve dig targeter for fulldebug builds 10(3 minutes ago, 2 files, 25+ 9-) 13https://github.com/crawl/crawl/commit/ae6112e33f99
12:52:08 <advil> aidanh: I'm not sure bb52cfba1566 is much better than what you did, but there seem to be some weird/awkward interactions with general map redraw code and what direction_chooser does for redraws, so this is another placeholder-y fix for the evoke from inventory redraw issue
13:10:19 <Jorgrell> Unstable branch on crawl.jorgrun.rocks updated to: 0.23-a0-54-gae6112e33f (34)
13:22:54 <Henzell> Unstable branch on crawl.akrasiac.org updated to: 0.23-a0-54-gae6112e (34)
15:43:11 <NormalPerson7> aidanh, I've just noticed that XG is broken (at least in trunk), it's possible that's similar to the earlier bug we had with Xv?7
15:46:40 <NormalPerson7> s/XV?7/XV ?/
16:46:11 <amalloy> specifically that is broken in webtiles but not console
17:32:03 <ZiBuDo> !messages
17:32:04 <Sequell> No messages for ZiBuDo.
17:33:15 <gammafunk> ZiBuDo: I haven't heard from |amethyst about that trunk menu problem, but johnstein mentioned that the problem with cbro console spectating being broken was something wrong with the shared memory file
17:33:20 <gammafunk> maybe a permissions issue?
17:33:23 <gammafunk> if that rings any bells
17:33:34 <gammafunk> not sure if it's related to cjr
17:33:54 <ZiBuDo> gammafunk: i can look into that at some point, but that gives some clues, so ill mess around with some stuff
17:34:18 <gammafunk> I'm happy I can start the tutorial to get to my trunk game on cjr, at least
17:34:26 <ZiBuDo> lol
17:34:49 <gammafunk> it's no laughing matter, I had to play through all of lair using TILES!
17:34:55 <gammafunk> the humiliation was unbearable
17:35:23 <ZiBuDo> thats why the picture for webtiles is a little kid playing with a car
17:35:44 <gammafunk> it all makes sense
18:21:38 <Lantell> Unstable branch on underhound.eu updated to: 0.23-a0-54-gae6112e33f (34)
18:36:30 <gammafunk> advil: heh, interesting thing that's happened due to your permastore name generation
18:36:35 <gammafunk> ghost names can now contain spaces
18:36:40 <gammafunk> this broke the tourney script
18:36:45 <gammafunk> should be an easy fix though
19:16:57 <alexjurkiewicz> a nice UI project could be to accumulate all rarely-used game screens under a single button, so they're more easily discoverable for newbies, in the same way that inventory allows discovery of rqPTw etc
19:17:11 <alexjurkiewicz> thinking screens like % ^O \ E etc
19:23:02 <gammafunk> rip sequell
19:23:14 <gammafunk> wb Sequell
19:23:20 <gammafunk> !lg * 0.22
19:23:21 <Sequell> No games for * (0.22).
19:23:25 <gammafunk> doh
19:23:50 <gammafunk> greensnark: are you working on Sequell, I noticed that 0.22 games don't seem to be showing up in it yet
19:23:58 <gammafunk> despite being in the sources
19:24:05 <greensnark> Hm let me look
19:24:13 <greensnark> Yes, I just updated the dependencies for a security issue
19:24:50 <gammafunk> ah, yeah I saw those github notifications
19:25:16 <greensnark> !lg * cao s=-file
19:25:17 <Sequell> 2892038 games for * (cao): 2667x cao/logfile-lorcs, 30258x cao/logfile11, 52063x cao/logfile12, 56074x cao/logfile13, 64315x cao/logfile10, 67775x cao/logfile14, 68363x cao/logfile08, 68617x cao/logfile09, 80359x cao/logfile06, 111235x cao/logfile18, 113468x cao/logfile17, 117290x cao/logfile19, 121836x cao/logfile15, 122223x cao/logfile16, 139122x cao/allgames.txt, 141674x cao/logfile20, 142913x ...
19:25:47 <greensnark> !lg * cao file=~logfile22
19:25:47 <Sequell> No games for * (cao file=~logfile22).
19:26:23 <gammafunk> cxc 0.22 seems to properly be in sequell I noticed
19:26:25 <gammafunk> just not the others
19:26:33 <gammafunk> well the branch file is
19:26:56 <gammafunk> !lg * file~~0\.22 s=file
19:26:57 <Sequell> 1281 games for * (file~~0\.22): 1281x cxc/crawl/meta/0.22/logfile
19:27:09 <gammafunk> !lg * file~~0\.22 s=cv
19:27:10 <Sequell> 1281 games for * (file~~0\.22): 1281x 0.22-a
19:27:44 <greensnark> It seems to be reading the logs after the restart, I'm not sure the config rescanning is working properly
19:27:45 <gammafunk> hrm, maybe the beta still shows up as 0.22-a
19:28:06 <greensnark> !lg * cao file=~logfile22
19:28:07 <Sequell> 370. barghouti the Skirmisher (L1 TrAK of Lugonu), blasted by a wretched star (orb of energy) on Abyss:1 on 2018-08-07 23:22:25, with 0 points after 60 turns and 0:00:18.
19:28:11 <gammafunk> cool
19:28:12 <greensnark> !lg * cao file=~logfile22 x=cv,v
19:28:14 <Sequell> 370. [cv=0.22-a;v=0.22.0-b1] barghouti the Skirmisher (L1 TrAK of Lugonu), blasted by a wretched star (orb of energy) on Abyss:1 on 2018-08-07 23:22:25, with 0 points after 60 turns and 0:00:18.
19:28:21 <gammafunk> ah, so that will be 0.22-a, ok
19:28:48 <gammafunk> !lg * v=0.21.0-b1 x=cv
19:28:49 <Sequell> 22320. [cv=0.21-a] lakebodom the Chopper (L6 MiBe of Trog), slain by an orc (a +0 trident) on D:4 on 2018-08-07 18:05:08, with 437 points after 2596 turns and 0:10:49.
19:43:34 <alexjurkiewicz> !lg * cpo 0.22
19:43:35 <Sequell> No games for * (cpo 0.22).
19:44:31 <alexjurkiewicz> !lg . HoFi xl=12
19:44:32 <Sequell> 5. chequers the Severer (L12 HOFi of Wu Jian), slain by Snorg on Lair:3 on 2018-07-31 09:33:05, with 17200 points after 14879 turns and 0:19:31.
19:44:46 <alexjurkiewicz> !lg . HoFi xl=12 x=cv,v,file
19:44:47 <Sequell> 5. [cv=0.22-a;v=0.22.0-b1;file=cpo/dcss-logfiles-0.22] chequers the Severer (L12 HOFi of Wu Jian), slain by Snorg on Lair:3 on 2018-07-31 09:33:05, with 17200 points after 14879 turns and 0:19:31.
19:44:53 <alexjurkiewicz> !kw 0.22
19:44:54 <Sequell> Built-in: 0.22 => cv=0.22
19:45:23 <alexjurkiewicz> oh, cv 0.22-a
20:28:53 <Cheibriados> ctrl-[ throws an error 13https://crawl.develz.org/mantis/view.php?id=11595 by Yermak
21:32:31 <Cheibriados> in the description screen for Sling Bullets, skill training is Throwing.png 13https://crawl.develz.org/mantis/view.php?id=11596 by RoGGa
21:38:14 <hellmonk> is it intended that you aren't listed for the lower tiers of a banner if you get the highest tier of that banner? There are some banners where getting the tier III does not imply that you got the tier I and/or II banners
21:41:53 <alexjurkiewicz> I think it is, because the awards are tiered
21:48:53 <hellmonk> dang, rip
21:49:32 <hellmonk> I guess my tournament plan if mibe does not come back will just be all tier 3 banners then, too lazy to track the lower ones
21:49:43 <hellmonk> not worth the extra effort
23:50:20 <Cheibriados> Shop's item list sometimes goes to 2 columns during purchase 13https://crawl.develz.org/mantis/view.php?id=11597 by RoGGa