00:32:10 <Eksell> Unstable branch on CRAWL.XTAHUA.COM updated to: 0.24-a0-84-g25a2d9f (34)
00:35:37 <Eksell> Stable (0.23) branch on CRAWL.XTAHUA.COM updated to: 0.23-b1-84-g6588a6b
00:52:09 -!- name is now known as nosuni
00:52:31 <Cheibriados> Wand of digging reports indestructible wall is out of range 13https://crawl.develz.org/mantis/view.php?id=11860 by Sage1234
01:06:31 <switch___> since it was brought up; ebering, forgive me if you already answered this last time we spoke of it, but when did you expect to have time to take a look at the cloud PR?
01:12:45 <Gretell> Stable branch on crawl.develz.org updated to: 0.23-b1-84-g6588a6b (34)
01:55:43 <johnstein> ??rebuild
01:55:44 <Sequell> johnstein: You have 1 message. Use !messages to read it.
01:55:44 <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.kelbi.org/rebuild/ Bug gammafunk, advil, |amethyst, or Nap.Kin for CDO. Use your powers wisely.
02:09:35 -!- amalloy_ is now known as amalloy
03:03:29 -!- amalloy is now known as amalloy_
03:32:17 <Kelbell> Stable (0.23) branch on crawl.kelbi.org updated to: 0.23-b1-84-g6588a6bd43
03:46:41 <Kelbell> Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-1682-g99b902e015
08:10:33 -!- muffindrake1 is now known as muffindrake
08:45:28 <advil> if a post-place hook fails, it vetos the entire map. Anyone know if that's what ought to happen?
08:45:48 <advil> the way in which it happens is pretty indirect, so that's one reason why I'm wondering
08:46:17 <advil> also because the way it is done throws a dgn_veto across lua code which has pretty bad interaction with setjmp/longjmp
08:48:06 <advil> I think I should at least catch this exception before returning to lua
08:56:28 <advil> the easiest fix would cause failures on post-place hooks not to veto
08:57:00 <advil> I think this is probably fine for serial maps, it probably happens when you still get 5/6 maps or something
08:57:26 <advil> e.g. my test case rolls like 8 serial forest maps, and the last one can't fit
09:01:24 <advil> actually, given that without this change, dgn.place_maps effectively doesn't respect die_on_error, maybe it is what was distantly intended
09:02:22 <advil> aha yeah
09:02:44 <advil> can restore the previous veto behavior if you pass die_on_error in your post_place hook
09:30:02 <Cheibriados> 03advil02 07* 0.24-a0-85-g091e5a1: Catch dgn_veto_exception when placing maps from lua 10(10 minutes ago, 1 file, 18+ 6-) 13https://github.com/crawl/crawl/commit/091e5a18feeb
09:31:00 <advil> would also like to push that to 0.23 but might be good to get some input if anyone has concerns about some other way post-place hooks are used that I'm not thinking of
09:32:19 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-85-g091e5a18fe (34)
09:35:09 <advil> oops, ignore the spurious cherry-pick message with a non-existent ref in that commit msg
10:46:16 <Cheibriados> 03advil02 07* 0.24-a0-86-g56b2405: Adjust one post_place hook 10(6 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/56b24058a9e1
10:51:03 <advil> still interested in feedback on the change to post_place behavior, but I just looked systematically through all the post_place hooks, and given how they work I think it's fine
10:51:47 <Cheibriados> 03advil02 07[stone_soup-0.23] * 0.23-b1-85-g06d10bc: Catch dgn_veto_exception when placing maps from lua 10(2 hours ago, 1 file, 18+ 6-) 13https://github.com/crawl/crawl/commit/06d10bc0c134
10:51:47 <Cheibriados> 03advil02 07[stone_soup-0.23] * 0.23-b1-86-gba62f8e: Adjust one post_place hook 10(11 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/ba62f8e09226
10:54:51 <advil> weird, github swallowed my issue reference in the "Catch..." commit
10:55:19 <advil> (or I did it)
10:55:21 <gammafunk> advil: line beginning with #?
10:55:21 <advil> should say "I have left a note in #981"
10:55:24 <advil> oh yeah
10:55:31 <gammafunk> that's burned me a couple times
10:55:35 <advil> must be it, got reformatted when I did gqap
10:56:00 <gammafunk> advil: I know that the ilsuiw shoals vault often fails to place and prints a user-visible error when it does so
10:56:08 <gammafunk> don't know how related to your change that is
10:56:11 <advil> yeah, that one already set die_on_error
10:56:18 <gammafunk> right, hence it's visible, ok
10:56:39 <advil> well, it might have been visible for the wrong reasons, but I think its behavior should be unchanged from the player perspective
10:57:38 <advil> there may be some compiler dependence here too, given that this is all about exactly how the lua stack works in cases where it isn't supposed to work
10:59:07 <gammafunk> yeah, re: that note in your second commit
10:59:19 <gammafunk> it seemed to me that there's no good reason why the error is printed if the level is vetoed
10:59:49 <gammafunk> however we probably want to take a look at maps that do frequently generate vetoes
10:59:54 <advil> yeah, that seems right form the cases I looked at
11:01:02 <advil> ah, here's a nice 1-sentence summary of the situation from a SO discussion: "setjmp()/longjmp() completely subvert stack unwinding and therefore exception handling as well as RAII (destructors in general)."
11:01:24 <advil> probably every time there's a lua `error()` call currently we have some risk of memory leaks
11:01:43 <advil> or c++ stack corruption more generally
11:01:49 <ebering> yikers
11:01:58 <advil> indeed
11:02:25 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-86-g56b24058a9 (34)
11:02:25 <ebering> clua can call error(); if that can corrupt the stack we might have a vulnerability in the clua interpreter :/
11:02:41 <advil> I actually had no problems compiling with contrib lua in c++ mode (and it avoided the crash that I just fixed by catching the exception), but I'm not going to commit that now
11:02:56 <ebering> something for 0.24 :)
11:03:00 <ebering> speaking of
11:03:44 <advil> I am once again in ignorance of any disastrous but obscure problems that need an immediate fix
11:03:59 <advil> so it could be a good time to tag :D
11:04:01 <ebering> switch___: I didn't set a specific timeline for the clouds PR; it's something I'd like to do in the 0.24 cycle. My doing it will be made much easier if you add some git hub annotations to the PR linking to the specific parts of crawl's lua that this is replacing, as well as a more detailed description of the problem with desolation clouds as best you understand it
11:04:39 <gammafunk> can't tag before I push my demigod rework
11:05:47 <gammafunk> I need to push a quick update to the debian control stuff at least
11:07:12 <gammafunk> oh
11:07:18 <gammafunk> any any last minute changelog adjustments?
11:07:44 <gammafunk> might be a good time to take a look
11:07:51 <gammafunk> I rewrote the seed section, hopefully it's not inaccurate (the last bullet point about what people should report as a bug might be)
11:09:14 <ebering> under highlights: * Supported for seeded dungeon play
11:09:19 <ebering> should be "Support"
11:09:25 <gammafunk> oh, thanks
11:09:27 <ebering> and the title line needs a release date
11:09:36 <gammafunk> yes, that part I intentionally left out
11:09:55 <gammafunk> since I wasn't sure when it would be
11:09:57 <ebering> also I dislike the verb rework but not so strongly as to edit it out :)
11:10:13 <gammafunk> oh feel free to suggest another verb
11:10:20 <gammafunk> revamp? overhaul? ruin?
11:10:26 <ebering> for traps I like overhaul
11:11:15 <advil> burn to the ground
11:11:16 <ebering> maybe reimagined for nemelex :P
11:11:21 <gammafunk> heh
11:11:25 <ebering> !!!1
11:11:32 <ebering> Nemelex Xobeh has been redrawn
11:11:37 <gammafunk> that is a *little* grandiose
11:11:56 <advil> reshuffled
11:11:59 <advil> for nem
11:13:02 <gammafunk> also I'm going for proper bullet points with the highlights
11:13:02 <gammafunk> so not complete sentences
11:13:24 <gammafunk> although I guess you could argue that full sentences are clearer
11:13:50 <gammafunk> but this way seems a little better for grabbing attention
11:14:29 <gammafunk> Nemelex Xobeh overhaul? I'm avoiding outright jokes and going for the just-the-facts style we use for changelogs
11:14:45 <gammafunk> We're not the Discord app, where literally every UI message is some kind of joke!!!
11:18:03 <Cheibriados> Lua error handling vs c++ stack 13https://crawl.develz.org/mantis/view.php?id=11861 by advil
11:18:21 <Cheibriados> 03gammafunk02 07[stone_soup-0.23] * 0.23-b1-87-g195ea7e: Update the Debian changelog for 0.23 10(2 minutes ago, 1 file, 5+ 0-) 13https://github.com/crawl/crawl/commit/195ea7e421b6
11:18:24 <gammafunk> we are also probably doing our debian changelog wrong, what with having it UNRELEASED
11:18:32 <Cheibriados> 03gammafunk02 07* 0.24-a0-87-gf756fb0: Update the Debian changelog for 0.23 10(3 minutes ago, 1 file, 5+ 0-) 13https://github.com/crawl/crawl/commit/f756fb0b9a79
11:18:45 <gammafunk> tbh I'm not sure how you're actually supposed to manage your debian directory in a repo in general
11:21:54 <advil> gammafunk I think if we'd spent a bit more time it would be worth adopting literally every one of minmay's rule change suggestions, but since we didn't get to it I'm not sure about some of them at the last minue. But possibly the nem choice changes would be the easiest low hanging fruit there?
11:21:58 <advil> "Either only award points for a player's first 5 Nemelex' Choice wins, or have a new Nemelex' Choice every 2 days instead of every time the previous one is won. In the latter case, award the points no matter how many times that combo has already been won"
11:22:05 <Lantell> Stable (0.23) branch on underhound.eu updated to: 0.23-b1-86-gba62f8e092
11:22:10 <advil> maybe also "No clan points for unique deaths, killing ghosts, killing other players with ghosts, number of high scores, or first to kill all uniques."
11:22:27 <advil> a streak point limit also could be a good idea
11:23:23 <advil> well, the only one I don't like is "only award points for each rune once" because that works really differently for weaker players, who might not be able to win but can get runes
11:24:34 <advil> putting a cap on those points that is not 1 seems good though
11:25:32 <gammafunk> advil: advil: I thought about applying those, yes, although the thing that put me off is that fixed scoring for species
11:25:38 <gammafunk> need per-species scores
11:25:58 <gammafunk> advil: actually imo nchoich scoring is not a major problem due to the stepdown elliptic put in place
11:26:11 <gammafunk> far more a problem are streaks and just the god/species/class/high scores
11:26:22 <advil> oh, didn't realize there was a stepdown
11:26:28 <advil> I wonder how many people are aware of that
11:26:32 <gammafunk> yeah, check out the rules for details
11:26:51 <gammafunk> it's all explained rather well and in detail
11:26:53 <gammafunk> under Scoring
11:27:10 <gammafunk> "Important note for scoring" describes the stepdown
11:27:30 <advil> this seems a bit like assuming people will actually read the manual ;-)
11:27:37 <gammafunk> sorry?
11:27:53 <advil> oh I just mean that probably people don't read the rules, in the same way they don't read the manual
11:27:53 <gammafunk> I mean, it's the tournament rules page and we're changing the tournament rules
11:28:04 <gammafunk> sure, but are those people going for top scores
11:28:18 <gammafunk> people spending 8 hours a day playing are definitely reading and gaming this
11:28:35 <advil> so what about capping the streak points at 5-streaks?
11:28:55 <gammafunk> possible, might be a bit aggressive given we have a banner stopping right at 5
11:29:23 <advil> 10-streaks would be fine with me
11:29:40 <advil> are there any past tournament scoring pages that are up right now?
11:29:47 <gammafunk> https://crawl.develz.org/tournament/0.23/
11:29:51 <gammafunk> the current one is up
11:30:00 <gammafunk> you want to see the scores? I think dobrazupa is not up atm
11:30:13 <gammafunk> there are ones on CAO but they are really ancient
11:30:13 <advil> yeah, was just trying to refresh myself on what people did
11:30:46 <gammafunk> yeah, re: the rules in general though, I think we do have to actually assume people are reading these rules and carefully applying them for the cases we're trying to change
11:31:02 <gammafunk> just to be clear about that, detailed discussions between the top players for these rules are very common
11:31:21 <gammafunk> so it's ok if we have these relatively complicated rules, so long as they're well-described
11:31:51 <gammafunk> maybe we could design a vastly simpler tournament and it'd be good, but if we want to mostly preserve the format, we're going to have this complicated rules situation, basically
11:32:12 <advil> yeah, seems true
11:32:13 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-87-gf756fb0b9a (34)
11:33:11 <gammafunk> the one thing that made me hesitate about applying all of these rules was the need for per-species/class score weights
11:33:36 <gammafunk> we could target just streaks and maybe eliminate combo high scores?
11:33:39 <advil> ok I'd vote for either capping streak points at 10-streaks, or applying a stepdown to streaks as well (not sure if the existing stepdown would work though)
11:33:56 <gammafunk> yeah, could do I guess
11:34:02 <gammafunk> re: applying stepdown
11:34:11 <gammafunk> I'm actually not sure if that would fix the problem as well as a cap might
11:34:41 <advil> right, it might just make the grinding out points even more tedious
11:34:54 <advil> probably someone would still try to get a longest streak, we can/should leave those scoreboard things up, it just wouldn't incentivize it with points
11:35:02 <gammafunk> we could cap points for both streaks and nchoice certainly
11:35:09 <gammafunk> yeah...hrm
11:35:25 <gammafunk> well, if only we'd started in on this a couple weeks before instead of the day before, but you know how that goes
11:35:37 <advil> all the best things in 0.23 are last minute ;-)
11:35:42 <advil> for combo high scores, it would be nice to know what the distribution of those over non-competitive players is
11:36:15 <advil> for example, iirc I got a combo high score last t, which was fun
11:36:45 <advil> it was obviously a combo that wasn't played much
11:36:46 <gammafunk> yeah, I guess awarding them is fun; the points part is sort of bad.
11:36:57 <advil> I guess we could also just not award points, but keep the markers
11:37:03 <gammafunk> |amethyst: any chance you could kick dobrazupa.org so we could take a look at the tournament pages?
11:37:07 <gammafunk> actually...
11:37:23 <gammafunk> ah, ssh not working, nm
11:37:58 <gammafunk> right, I think some of that might be a bit misleading for players
11:38:34 <gammafunk> tracking but not awarding points; I guess it's not too confusing for casual players (those would be the ones who really might just not read the rules page at all)
11:38:45 <advil> yeah, maybe
11:39:17 <advil> could just show them in a list on the player page and keep the aggregate combo high score board
11:40:21 <advil> my other ideas mostly involved adding a couple of scoreboards, which wouldn't change point allocations or rules
11:41:00 <advil> I got caught up in insane lua stuff so I don't yet know how realistic it is to think they could be added v quickly
11:43:23 <gammafunk> well if we weren't also moving to a new hosting server with limited cpu
11:43:39 <gammafunk> I might be a bit more excited to try bolder changes, but yeah even then it's a little last minute
11:43:47 <gammafunk> I don't want to interrupt the start of the tournament, basically
11:43:50 <advil> right
11:44:19 <gammafunk> I don't fully know how cdo will do in the opening hours, but I think it will be fine
11:44:51 <gammafunk> it's very convenient to host it on CDO, so I hope it works out
11:45:51 <NP7> I just got a report of morgue files' vault lists being broken, see this morgue https://crawl.kelbi.org/crawl/morgue/Amao/morgue-Amao-20190207-145437.txt . You can see the header "Levels and vault maps discovered" in the file and then there's nothing below it. You can clearly see that a vault spawned because there's a ghost in the notes.
11:46:19 <advil> hm
11:54:58 <advil> yeah, I can replicate that
11:55:03 <advil> undoubtedly my fault somehow
11:56:37 <advil> man, I meant to test that online once it got pushed and just forgot
12:09:02 <advil> ugh, something is kind of messed up here
12:09:38 <gammafunk> I'm going to push the final changelog but will hold off on tagging
12:10:13 <Cheibriados> Jewellery autopickup behavior changed after monster inventory ID change 13https://crawl.develz.org/mantis/view.php?id=11862 by Goratrix
12:10:17 <Cheibriados> 03gammafunk02 07* 0.24-a0-88-gb3eb09e: Final changelog for 0.23 10(3 minutes ago, 1 file, 7+ 7-) 13https://github.com/crawl/crawl/commit/b3eb09ec81ca
12:10:30 <Cheibriados> 03gammafunk02 07[stone_soup-0.23] * 0.23-b1-88-gd2508bd: Final changelog for 0.23 10(3 minutes ago, 1 file, 7+ 7-) 13https://github.com/crawl/crawl/commit/d2508bd6c9b4
12:12:57 <advil> oh
12:17:11 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-88-gb3eb09ec81 (34)
12:19:59 -!- amalloy_ is now known as amalloy
12:25:53 <Henzell> Unstable branch on crawl.akrasiac.org updated to: 0.24-a0-87-gf756fb0 (34)
12:33:35 <Cheibriados> 03advil02 07* 0.24-a0-89-gce0dbf4: Fix vault printing in morgues after game end (NP7) 10(14 minutes ago, 2 files, 9+ 3-) 13https://github.com/crawl/crawl/commit/ce0dbf48b24c
12:34:10 <Cheibriados> 03advil02 07[stone_soup-0.23] * 0.23-b1-89-g91c98b2: Fix vault printing in morgues after game end (NP7) 10(15 minutes ago, 2 files, 9+ 3-) 13https://github.com/crawl/crawl/commit/91c98b23f63d
12:34:12 <advil> there was a brief 10 minute window where I thought I had introduced save corruption 1 day before a release
12:34:19 <advil> but no, you.save is just deleted at the point where a post-game morgue file is getting printed
12:47:28 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-89-gce0dbf48b2 (34)
13:13:31 <gammafunk> advil: looks like travis is (partially) ok with that, am I clear to tag?
13:13:36 <advil> yeah
13:31:13 <gammafunk> %git stone_soup-0.23
13:31:13 <Cheibriados> 07advil02 * 0.23.0-1-g91c98b2: Fix vault printing in morgues after game end (NP7) 10(72 minutes ago, 2 files, 9+ 3-) 13https://github.com/crawl/crawl/commit/91c98b23f63d
13:31:31 <gammafunk> oh, oops
13:31:37 <gammafunk> damn
13:31:40 <gammafunk> how did I miss that
13:31:42 <advil> heh did you miss one?
13:31:49 <gammafunk> yeah seems so
13:31:50 <advil> fastest 0.23.0 version ever
13:31:53 <gammafunk> I guess I did forget to pull
13:31:55 <advil> I think you could safely retag it though
13:31:59 <gammafunk> right
13:32:00 <advil> as long as you do it quickly
13:32:10 <advil> maybe doesn't even matter
13:33:11 <gammafunk> %git stone_soup-0.23
13:33:11 <Cheibriados> 07advil02 * 0.23.0: Fix vault printing in morgues after game end (NP7) 10(74 minutes ago, 2 files, 9+ 3-) 13https://github.com/crawl/crawl/commit/91c98b23f63d
13:33:21 <advil> ncie
13:33:22 <advil> er
13:33:23 <advil> nice
13:33:31 <gammafunk> hopefully that doesn't break something...
13:33:36 <gammafunk> I had to force push the tag
13:33:39 <advil> ah
13:33:54 <gammafunk> silly mistake, oh well
13:33:55 <advil> no servers pulled in the interim
13:34:11 <advil> I once accidentally force pushed master when no one was looking (or updating) and it was ok
13:34:40 <gammafunk> if we break cpo, I consider it a victory
13:34:43 <advil> haha
13:36:33 <ebering> woooo tagged
13:36:40 <advil> just sent a discord message to johnstein re 0.23, in case he hasn't been paying attention here
13:37:22 <gammafunk> Yeah I sent one to him yesterday
13:37:34 <gammafunk> hopefully he'll have time to do that soon
13:37:35 <gammafunk> oh
13:37:52 <gammafunk> well, looks like cwz has 0.23
13:37:54 <gammafunk> so that's good
13:38:30 <gammafunk> advil: I'm starting up builds, so of course feel free to do the ones for OS X whenever
13:38:39 <gammafunk> if it will take you a day or two I can always mention in post that they'll be avail later
13:39:23 <advil> I can probably do it today
13:44:18 <NP7> Yay 0.23 has been tagged! Thanks everyone, I know you guys have worked really hard this release and I approve of the changes! :D
14:03:43 <gammafunk> %git 6588a6b
14:03:43 <Cheibriados> 07ebering02 * 0.23-b1-84-g6588a6b: Update CREDITS.txt 10(16 hours ago, 1 file, 5+ 0-) 13https://github.com/crawl/crawl/commit/6588a6bd4328
14:04:04 <gammafunk> ebering: that's what CWZ eventually ended up getting, so pretty close to actually 0.23.0 at least
14:04:26 <ebering> nice
14:04:29 <gammafunk> alas another version will cwz will never fully see the latest stable version
14:37:22 <advil> oh god this is 2-space python
14:37:24 <advil> why
14:37:48 <amalloy> advil: is 4 more popular?
14:37:56 <advil> yes
14:37:57 <advil> very yes
14:38:26 <advil> https://www.python.org/dev/peps/pep-0008/
14:38:37 <amalloy> interesting. we use 2 at work, and i'm not exposed to any other python
14:38:47 <advil> it's pretty enshrined, given pep 8
14:40:21 <gammafunk> advil: yeah, I had the same gripe, but didn't feel like reformatting all of the code
14:40:31 <gammafunk> and I'm also looking at you 2-space lua in most of crawlcode
14:40:41 <advil> yeah, I hate it in lua too
14:40:52 <gammafunk> I make sure the stuff I write is 4-space if it's new, at least
14:40:54 <advil> for lua there at least seems to be a lot of people who do it
14:41:23 <advil> at least my editor draws little vertical bars for each indent
14:42:15 <gammafunk> are you trying rule changes, or just some new game displays?
14:43:10 <advil> just some displays
14:44:08 <gammafunk> just fyi, the way I tend to work is to push any config changes to the repo from my machine, and then pull from cdo
14:44:33 <gammafunk> and right now on cdo there's a local commit with cdo-specific changes (db auth stuff)
14:44:43 <gammafunk> and it autorebases for git pull
14:44:54 <advil> ok...I won't try to update cdo any time soon
14:45:01 <gammafunk> as opposed to pushing changing from cdo or just not pushing them at all
14:45:05 <advil> and I'd probably do a branch first
14:45:09 <gammafunk> s/changing/changes/
14:45:19 <advil> not sure how much I can get done, I'd hoped to work on this yesterday
14:47:14 <gammafunk> well you can try to add some during the tournament certainly
14:47:33 <advil> yeah
14:47:33 <gammafunk> it's not a problem to stop and restart the script
14:47:39 <gammafunk> it's just we want to avoid longer outages and most especially we don't want to break the db
14:47:50 <advil> one simple thing I'd like to do is have a scoreboard (both clans, players) that is just one weekend, probably the last one
14:48:00 <advil> so that literally won't update for another couple weeks
14:48:22 <gammafunk> is that to help get statistics or just because you think people would like to see it?
14:48:38 <advil> I think people would like to see it
14:49:04 <advil> one complaint many people have reported is that they don't have time to play during the week
14:49:12 <gammafunk> it would be pretty cool to see date range summaries in general
14:49:14 <gammafunk> hrm, yeah
14:49:18 <gammafunk> I suppose weekends are special
14:49:37 <advil> so this would highlight a much shorter range than the whole tournament for people to focus on, if they want
14:49:53 <floraline> is tag 0.23.0 supposed to be annotated?
14:50:03 <gammafunk> all release tags are, yeah
14:50:10 <floraline> it looks not that way
14:50:12 <advil> the other thing I'm playing with is a scoreboard for accounts that have played only a very small number of games, something like 10
14:50:19 <floraline> or am i wrong
14:50:30 <gammafunk> how are you trying to look at the annotation
14:50:38 <floraline> https://github.com/crawl/crawl/releases
14:50:38 <floraline> just here
14:51:05 <advil> oh, we don't usually do that..I only did it a few times when I needed to put that bones file download somewhere
14:51:16 <gammafunk> right, that's a special github thing
14:51:25 <floraline> oh ok
14:51:37 <advil> no good reason not to add a download there, or at least annotate it so it gets the newest release green thing
14:51:42 <advil> but we just don't usually
14:52:51 <advil> ok, I did it
15:09:54 <gammafunk> !echo %git 0.23.0~$(rand 963)
15:09:54 <Sequell> %git 0.23.0~650
15:09:54 <Cheibriados> 07gammafunk02 * 0.23-a0-339-gbd7f2c8: New tiles for the Sceptre of Torment (CanOfWorms) 10(4 months ago, 4 files, 2+ 1-) 13https://github.com/crawl/crawl/commit/bd7f2c837c05
15:09:58 <gammafunk> !echo %git 0.23.0~$(rand 963)
15:09:58 <Sequell> %git 0.23.0~754
15:09:58 <Cheibriados> 07advil02 * 0.23-a0-235-gf47f433: Temporarily revert html side of forget password commits 10(5 months ago, 1 file, 0+ 50-) 13https://github.com/crawl/crawl/commit/f47f43309fcf
15:10:08 <gammafunk> powerful new tech for choosing the best commit of 0.23
15:37:05 <floraline> are we out of beta now
15:38:40 <alexjurkiewicz> woo, congrats on 0.23!
15:40:59 <gammafunk> floraline: once 0.23.0 is tagged, that means no beta
15:41:13 <gammafunk> well more succinctly it means the release has happened, repository-wise
15:46:00 <alexjurkiewicz> advil: I'm going to stop weekly challenge during the tournament, but I was thinking that it might be a good way to dig out any weirder causes of seed instability. Is there any additional debugging data I could add to the system to help you track down seed bugs?
15:51:23 <alexjurkiewicz> floraline: fwiw here is the current state of weekly challenge, but it will get rebuild over the next couple of weeks to use the new functionality that landed yesterday https://github.com/alexjurkiewicz/dcss-weekly-challenge/
15:54:01 <floraline> cool, thank you! i'll keep an eye on it
15:54:30 <travis-ci> The build has errored. (0.23.0 - 91c98b2 #11113 : advil): https://travis-ci.org/crawl/crawl/builds/490188265
15:54:47 <gammafunk> nice try, travis
15:58:46 <gammafunk> CDO download page is mostly updated for 0.23.0
15:59:06 <gammafunk> will make release post and final update for OS X when those are available
16:00:09 <alexjurkiewicz> https://www.reddit.com/r/dcss/comments/ao8c4w/my_game_just_crashed_never_happened_to_me_before/
16:09:12 <advil> oh, I guess no one ever fixed that
16:09:20 <advil> it's some kind of bug with beogh god display
16:09:27 <advil> at least I'm pretty sure
16:09:36 <ebering> seems plausible based on those stack traces
16:10:04 <advil> there was one yesterday too
16:31:57 <Kelbell> Stable (0.23) branch on crawl.kelbi.org updated to: 0.23.0-0-g91c98b23f6
16:34:46 <advil> man my sql is rusty
16:37:07 <travis-ci> The build passed. (0.23.0 - 91c98b2 #11113 : advil): https://travis-ci.org/crawl/crawl/builds/490188265
16:46:48 <alexjurkiewicz> it's ~20hrs early, but please remember to add the tournament reminder to trunk
16:56:20 <advil> hm, probably if this special weekend thing were to be done it would make the most sense (given how this all works) to just run a separate tournament script instance for it, which seems non-ideal
16:56:42 <advil> it doesn't seem otherwise easy to limit all these queries to just a subset of games
17:03:11 <alexjurkiewicz> 0.23 is up on CPO, can someone add to Sequell? gammafunk? Let me know if you'd like a PR
17:03:29 <gammafunk> alexjurkiewicz: I'll do that in a bit, thanks for the heads up
17:03:39 <gammafunk> maybe I just forgot, since I think I already enabled it in tournament
17:03:45 <gammafunk> actually
17:03:51 <gammafunk> !lg * cpo 0.23
17:03:52 <Sequell> No games for * (cpo 0.23).
17:04:30 <gammafunk> alexjurkiewicz: yeah cpo already configured in sequell
17:04:34 <gammafunk> so I didn't forget
17:04:37 <gammafunk> !kw 0.23
17:04:37 <Sequell> Built-in: 0.23 => cv=0.23
17:05:18 <alexjurkiewicz> hm, there has been at least one death though
17:05:37 <gammafunk> yeah I see them
17:05:37 <alexjurkiewicz> !lm * cpo 0.23
17:05:37 <gammafunk> ah, you know
17:05:37 <gammafunk> the beta counts as alpha
17:05:37 <gammafunk> that's what it is
17:05:41 <gammafunk> !lg * cpo 0.23-a s=file
17:05:42 <Sequell> 15376 games for * (cpo 0.23-a): 15357x cpo/dcss-logfiles-trunk, 19x cpo/dcss-logfiles-0.23
17:05:46 <gammafunk> yeah, you're good
17:05:50 <alexjurkiewicz> ah, fine
17:05:55 <gammafunk> now that release has been tagged, they'll show up as 0.23
17:05:59 <alexjurkiewicz> so I can still get "my first 0.23 game is a win" then
17:06:12 <Sequell> No milestones for * (cpo 0.23).
17:06:16 <gammafunk> imagine chequers winning a char on his first try
17:06:20 <alexjurkiewicz> !lg . file=cpo/dcss-logfiles-0.23
17:06:20 <Sequell> No keyword 'dcss-logfiles-0.23'
17:06:38 <alexjurkiewicz> !lg . cv~~0.23
17:06:39 <Sequell> 39. chequers the Invulnerable (L26 MiBe of Trog), slain by an emperor scorpion on Zig:11 on 2019-01-30 11:03:23, with 593418 points after 65429 turns and 2:30:40.
17:06:49 <alexjurkiewicz> invulnerable my ass
17:06:56 <gammafunk> !lg * cpo file~~0.23 s=name
17:06:58 <Sequell> 19 games for * (cpo file~~0.23): 8x blight, TrashTierHero, chequers, chongming12, edsrzf, icthyon, madedhun, mattm, MrDizzy, olourkin, Phlogiston, shareduser
17:07:58 <alexjurkiewicz> !lg * cpo file~~0.23 max=sc
17:07:59 <Sequell> 19. chequers the Invulnerable (L26 MiBe of Trog), slain by an emperor scorpion on Zig:11 on 2019-01-30 11:03:23, with 593418 points after 65429 turns and 2:30:40.
17:08:00 <gammafunk> I guess I could grip how you made the url "dcss-logfile*s*-0.XX" when it's one logfile
17:08:08 <gammafunk> but water under the bridge
17:08:27 <alexjurkiewicz> i could make logfile work as well
17:08:32 <alexjurkiewicz> if you want to update sequell
17:08:46 <gammafunk> I could yeah
17:08:51 <gammafunk> might as well be consistent with other names
17:08:58 <alexjurkiewicz> but why is it logfile and milestone*s*
17:09:09 <alexjurkiewicz> 🤔
17:09:15 <advil> logsfile
17:09:20 <gammafunk> right, it's that filenames have different approaches to what they're even describing
17:09:26 <gammafunk> logfile is a terrible name I guess
17:09:31 <gammafunk> plural or singular
17:09:55 <alexjurkiewicz> ok, it's restarted
17:10:02 <alexjurkiewicz> s/restarted/available
17:10:11 <gammafunk> that works for past ones as well?
17:10:30 <gammafunk> I should have mentioned that, since you have one yml pattern entry to match them all
17:10:34 <gammafunk> actually we can just start a new one
17:10:35 <alexjurkiewicz> yes
17:10:55 <gammafunk> cool, thx
17:10:55 <alexjurkiewicz> both URLs work for all logfiles
17:19:25 <ebering> advil: I have some time to build the mac app zips if there's no signing or magic involved
17:19:42 <advil> There's an invocation
17:19:52 <ebering> just the mac-app-zips target with TILES=y and without, right?
17:21:00 <gammafunk> not sure what you all do with building those releases, but you also want to take some precautions so that any personal info isn't included or whatnot
17:21:00 <advil> I do more things than that
17:21:10 <advil> it's unclear to me how much is needed
17:21:33 <advil> I'll just do it now
17:21:33 <gammafunk> yeah, no rush really though
17:21:47 <ebering> ok
17:31:02 <advil> gammafunk: how do I rerun point assignment?
17:31:14 <advil> do I need to clear the db first?
17:31:35 <gammafunk> advil: that command in the README to recreate the db
17:31:40 <gammafunk> from the sql file in the repo
17:32:10 <advil> oh, I see, that must have the right drop table stuff in it?
17:32:10 <gammafunk> and yeah you have to clear it if e.g. rules change
17:32:51 <gammafunk> ebering: You mention "hundreds of fatal vaults" in the post, which is a rather wild estimate, but I saw you claiming large numbers of vaults had changed earlier
17:33:02 <gammafunk> so I'm not sure where you're getting your vault count from, or rather what you're actually counting
17:36:46 <gammafunk> it's a bit hard to count, but I don't think we quite reached "dozens"
17:38:44 <gammafunk> advil: yeah, it wipes any relevant table in the scoring db and recreates them
17:40:45 <ebering> gammafunk: my count might be a bit high yeah
17:41:52 <gammafunk> you did merge a bunch in that pr
17:41:55 <gammafunk> any idea how many that was?
17:42:11 <gammafunk> I'll just change that to "dozens"
17:42:19 <gammafunk> which is probably accurate
17:42:24 <ebering> definitely
17:42:26 <gammafunk> critical release details
17:43:57 <gammafunk> ok good
17:44:04 <gammafunk> changing the draft doesn't change the author
17:45:30 <gammafunk> https://crawl.develz.org/tournament/0.23/teams.html
17:45:34 <gammafunk> no real standout names yet
17:45:41 <gammafunk> I see ebering has been busy thinking of one though
17:46:10 <gammafunk> and dispersal traps already have a shoutout
17:51:24 <alexjurkiewicz> Mennas II Society is good
17:51:55 <ebering> !lm * uniq noun~~Bai cv>=0.22 s=br,cv
17:51:57 <Sequell> 1188 milestones for * (uniq noun~~Bai cv>=0.22): 1156x Crypt (677x 0.23-a, 449x 0.22, 30x 0.24-a), 28x Vaults (15x 0.23-a, 13x 0.22), Abyss (0.22), Snake (0.23-a), Spider (0.23-a), Swamp (0.23-a)
17:52:41 <ebering> Bai Suzhen's spawn seems to be messed up at some point along 0.23-a
17:52:47 <advil> gammafunk: what was your take on "remove background/combo high score bonuses"?
17:57:38 <ebering> !lm * uniq noun~~Bai cv>=0.22 s=br,vlong
17:57:40 <Sequell> 1188 milestones for * (uniq noun~~Bai cv>=0.22): 1156x Crypt (77x 0.23-a0-344-ge122d05, 71x 0.22.1-1-ga44c0e5, 60x 0.22.1-2-g9c5b331, 59x 0.23-a0-136-g722f094, 48x 0.23-a0-607-g760965b, 35x 0.22.1-1-ga44c0e56f4, 32x 0.23-a0-538-g7b100e3, 30x 0.23-a0-729-g94ec030, 25x 0.22.0, 21x 0.22.0-6-g71bd7aa, 21x 0.22.1-2-g9c5b331383, 21x 0.23-a0-295-g2e428d4, 19x 0.22.0-38-g04b72b8, 18x 0.22.1-6-g1668fe2, 15...
17:58:25 <gammafunk> advil: I guess it depends on the context; are you trying to push some rule changes for this tournament?
17:58:44 <advil> well, I'm doing the commits
17:58:44 <advil> we can decide whether or not to use them
17:59:05 <gammafunk> if we track those but don't don't give points...I wonder how more casual players will feel about that
17:59:15 <gammafunk> there is also the issue of us having banners related to stealing high scores
17:59:22 <advil> ah right
17:59:26 <gammafunk> I suppose those don't specifically cause a problem
17:59:37 <advil> heh didn't I already worry about casual players
17:59:44 <advil> and forgot
18:00:13 <gammafunk> yeah, it's a bit of a difficult balancing act
18:00:30 <gammafunk> catering to top tournament players trying to inch out every bonus from their games
18:00:37 <gammafunk> vs people who just want to earn some points
18:00:45 <advil> right, given that I don't want to drop those high scores now
18:01:38 <gammafunk> I think minmay's list is correct in identifying some of the worst offenders in terms of rewarding grinding games for high level tournament play; not sure if those changes will truly put it to rest
18:01:53 <gammafunk> like she mentioned, it might just reward grinding out a lot of speedruns/turncount runs to improve those points
18:02:20 <gammafunk> but I have not really sat down to look at those numbers carefully I guess
18:05:43 <advil> well, here's a few things: https://github.com/crawl/dcss_tourney/pull/3
18:06:37 <advil> what about clan points for # of high scores?
18:09:31 <gammafunk> hrm, good question
18:09:35 <gammafunk> similar issue for people in clans I guess
18:10:05 <advil> those only go to 3 clans, right?
18:10:17 <gammafunk> oh is that the case?
18:10:25 <gammafunk> if so, then that seems ok to go I guess
18:10:28 <advil> I'm thinking of the "200-100-50 clan points for the clans with the most high scores in species/background combos. "
18:10:41 <gammafunk> thought there were clan points for that for every clan
18:10:44 <advil> that's how I thought that worked, but we've seen that my detail-level knowledge of these rules is shaky
18:10:47 <advil> oh
18:10:54 <gammafunk> elliptic: What do you think of the ideas in https://crawl.develz.org/tavern/viewtopic.php?p=331589#p331589 from minmay change-wise for tournament scoring? It's sort of much more iterative in terms of trying limit some of the grinding out of games
18:12:05 <gammafunk> I guess we've already vetoed at least one of those changes: the removal of points for high scores
18:12:28 <advil> also I think it's fine to rely on the stepdown now where relevant
18:12:45 <advil> so that means holding off on the nchoice suggestion
18:13:10 <gammafunk> right, nchoice wasn't really a factor for top players due to stepdown
18:13:17 <gammafunk> it's perhaps still reasonable to cap it I guess
18:13:49 <gammafunk> yeah, I guess there are two sets of clan points wrt high scores: one just given if you're a plan in the top 3 for most high scores, and then top 3 for clans collectively with most high scores?
18:14:18 <gammafunk> having both of those seems a bit odd, since wouldn't there be overlap
18:15:15 <gammafunk> getting a link to minmay's post was such a thing due to that issue where a thanked post no longer has a convenient post link
18:15:23 <gammafunk> s/thing/pain/
18:20:39 <ebering> !lm * uniq noun~~Bai cv>=0.23-a0 s=br br!=Crypt
18:20:40 <Sequell> 18 milestones for * (uniq noun~~Bai cv>=0.23-a0 br!=Crypt): 15x Vaults, Snake, Spider, Swamp
18:21:00 <ebering> %git 402804a59a5
18:21:00 <Cheibriados> 07gammafunk02 * 0.22-a0-899-g402804a: Cleanups for uniques.des (amalloy) 10(6 months ago, 1 file, 10+ 5-) 13https://github.com/crawl/crawl/commit/402804a59a51
18:21:06 <advil> I wonder if there's something odd with multi-line DEPTHs
18:21:07 <advil> yeah
18:21:10 <ebering> there is
18:21:10 <advil> was looking at that one too
18:21:14 <ebering> !source dgn_add_depth
18:21:20 <advil> if so, that probably impacts a bunch of uniques
18:21:24 <Sequell> Can't find dgn_add_depth.
18:21:38 <ebering> !source l-dgn.cc:dgn_add_depth
18:21:38 <Sequell> Can't find l-dgn.cc:dgn_add_depth.
18:21:42 <ebering> or however you get that
18:21:47 <elliptic> gammafunk: I don't really think those changes will accomplish much
18:21:51 <Lantell> Unstable branch on underhound.eu updated to: 0.24-a0-89-gce0dbf48b2 (34)
18:21:52 <ebering> that function clears the depth-ranges vector
18:22:06 <ebering> so DEPTH: tags *should* reset
18:22:15 <ebering> the lone Snake, Spider, and Swamp I can't easily explain
18:22:21 <ebering> the 15x Vaults could be stairs
18:22:24 <elliptic> there's also probably some concern about players feeling forced to try for certain things
18:22:26 <advil> there might be vaults that explicitly place it
18:22:34 <ebering> ah, yes
18:22:39 <advil> also this is the kind of fix that we now can't do in a stable version without incrementing the version number because it will break seeding
18:22:48 <ebering> yeah. rip
18:22:52 <ebering> a fix for 0.23.1
18:23:18 <advil> so this is impacting placement of donald, mara, margery, asterion, aizul, and jorgrun as well
18:24:47 <ebering> while it might be tempting to fix it by changing the behavior of DEPTH: I'd caution against that
18:25:18 <ebering> since it would have weird interactions with default depths in the setting of depth ranges
18:25:18 <advil> hm yeah also there might be scripts that use this to override depth
18:25:39 <ebering> yeah, better to fix the file and maybe I'll have the energy to document dlua this version :P
18:25:49 <elliptic> gammafunk: at the moment there are a bunch of different competing ways to try to get lots of points - if you look at the actual people who do well, the distribution of their points can be significantly different depending on various factors
18:26:13 <elliptic> if you do changes like this, it's limiting the number of ways in which different people can get different numbers of points
18:26:19 <ebering> but I won't be pushing a fix any time soon, it's rehersal time
18:26:21 <gammafunk> yeah
18:26:31 <gammafunk> the more we try to remove one category of points, the more it forces players to try other things like realtime/turncount; I guess we just need to break out the categories
18:26:40 <alexjurkiewicz> rehearsal! What are you rehearsing?
18:26:52 <ebering> I play violin in an orchestra
18:27:17 <alexjurkiewicz> nice. You should post some recordings if any ever get made
18:27:26 <elliptic> I do still think that moving away from the "one person wins the tournament" paradigm is the best thing to do, though it's certainly more complicated to do that well than just to tweak some points
18:27:27 <gammafunk> he plays a dirge for all the players who've died in his vaults
18:28:18 <gammafunk> yeah, I guess there's also a concern about how to meaningfully score clans
18:28:34 <gammafunk> if you don't really combine points across categories, how do you score clans
18:31:02 <elliptic> one option is to have clans be about banner-type achievements rather than an actual ranking
18:31:45 <elliptic> can have a bunch of banners for the clan like "someone completed a zig" or "got great(er)player as a clan"
18:41:11 <Cheibriados> New branch created: pull/982 (1 commit) 13https://github.com/crawl/crawl/pull/982
18:41:11 <Cheibriados> 03Umer Shaikh02 07https://github.com/crawl/crawl/pull/982 * 0.24-a0-90-gba23b0d: Make god gifts interrupt & clear runrest delays only (11825) 10(26 minutes ago, 1 file, 3+ 2-) 13https://github.com/crawl/crawl/commit/ba23b0da4cf1
18:45:58 <advil> gammafunk: os x builds can be found at https://www.dropbox.com/sh/mjk2xoklkhmyg2n/AAAfljuzOz4nC3-wYZ4qdZMEa?dl=0
18:46:11 <gammafunk> thx
18:46:32 <advil> ebering if you want to see the incantation, there's a text file at that link that has it
18:46:33 <gammafunk> if only dropbox gave you wget-able urls
18:46:37 <gammafunk> I could copy directly to cdo
18:48:44 <gammafunk> is "mac osx" no longer a correct designation for the OS?
18:49:00 <gammafunk> we use "macosx" as a suffix in our filenames
18:49:55 <advil> those are the filenames that the build script generates
18:50:23 <gammafunk> yeah, I rename them for upload to cdo
18:50:29 <gammafunk> http://crawl.develz.org/release/current/
18:50:31 <advil> could probably change it to match the download names
18:50:45 <gammafunk> yeah, if you want to make that change, feel free
18:51:11 <gammafunk> by make that change, I mean in your build command or in the makefile
18:51:26 <gammafunk> those are uploaded, so let me go update download page and post that release post
18:51:30 <advil> I think it's in a makefile somewhere
18:51:30 <alexjurkiewicz> LTO. heh. has anyone built a PGO version of crawl?
18:55:25 <gammafunk> http://crawl.develz.org/wordpress/crawl-0-23-nemelex-needs-traps-badly
18:55:52 <alexjurkiewicz> elliptic: heh, just noticed you added faded altar support to qw. Does it help win rate?
18:58:40 <gammafunk> qw just loves chei and looks for any opportunity to worship chei that it can find
19:13:15 <alexjurkiewicz> ah. the new 'confirm on purple potion quaff' broke qw
19:18:46 <Kelbell> Webtiles server stopped.
19:18:55 <Kelbell> Webtiles server started.
19:20:18 -!- amalloy is now known as amalloy_
20:05:23 <elliptic> alexjurkiewicz: the faded altar support was just to get the banner for winning with faded altar in some previous tournament
20:06:48 <Henzell> acidosis1 (L1 GrFE) ASSERT(props.exists(VISITED_LEVELS_KEY)) in 'files.cc' at line 1435 failed. (D:1)
20:09:01 <gammafunk> the real question is how qw will evolve exciting new Gauntlet portal strategies
20:09:12 <gammafunk> advil: uh oh, crash above
20:09:34 <gammafunk> !crash
20:09:50 <Sequell> 20597. acidosis1, XL1 GrFE, T:0 (milestone): http://crawl.akrasiac.org/rawdata/acidosis1/crash-acidosis1-20190208-010646.txt
20:10:16 <advil> uh
20:10:20 <gammafunk> missing save compat?
20:10:29 <advil> no, there should be save compat, that I tested
20:11:29 <advil> oh hm
20:11:29 <gammafunk> they started their game, then saved?
20:11:29 <advil> I bet the save compat doesn't work if you save at turn 0
20:11:30 <gammafunk> hah
20:22:43 <advil> although I'm still confused, I thought I created that prop in the constructor in the end
20:25:26 <Cheibriados> 03advil02 07* 0.24-a0-90-ga7c8294: Improve save compat for visited level tracking 10(60 seconds ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/a7c82940b45b
20:25:49 <advil> luckily, that's not important for 0.23
20:26:27 <advil> oops
20:27:10 <Cheibriados> 03advil02 07* 0.24-a0-91-g4e4936a: fix the fix 10(27 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/4e4936a57423
20:32:23 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.24-a0-91-g4e4936a574 (34)
20:39:28 <gammafunk> who fixes the fixers
20:40:04 <Henzell> acidosis1 (L1 GrFE) ASSERT(props.exists(VISITED_LEVELS_KEY)) in 'files.cc' at line 1435 failed. (D:1)
20:40:09 <Henzell> acidosis1 (L1 GrFE) ASSERT(props.exists(VISITED_LEVELS_KEY)) in 'files.cc' at line 1435 failed. (D:1)
20:40:21 <Henzell> acidosis1 (L1 GrFE) ASSERT(props.exists(VISITED_LEVELS_KEY)) in 'files.cc' at line 1435 failed. (D:1)
20:40:48 <Henzell> acidosis1 (L1 GrFE) ASSERT(props.exists(VISITED_LEVELS_KEY)) in 'files.cc' at line 1435 failed. (D:1)
20:53:26 <Henzell> acidosis1 (L1 GrFE) ASSERT(props.exists(VISITED_LEVELS_KEY)) in 'files.cc' at line 1435 failed. (D:1)
20:56:48 <Henzell> acidosis1 (L1 GrFE) ASSERT(props.exists(VISITED_LEVELS_KEY)) in 'files.cc' at line 1435 failed. (D:1)
21:02:51 <advil> ??rebuild
21:02:51 <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.kelbi.org/rebuild/ Bug gammafunk, advil, |amethyst, or Nap.Kin for CDO. Use your powers wisely.
21:04:33 <alexjurkiewicz> how do I add flags to crawl compilation temporarily?
21:04:33 <advil> for example?
21:04:37 <alexjurkiewicz> -fprofile-instr-generate
21:04:57 <advil> honestly I just do stuff like that by sticking them somewhere plausible in Makefile
21:04:59 <alexjurkiewicz> I tried CFLAGS="-fprofile-instr-generate" make debug
21:05:06 <alexjurkiewicz> :)
21:06:53 <Henzell> Unstable branch on crawl.akrasiac.org updated to: 0.24-a0-91-g4e4936a (34)
21:09:12 <gammafunk> advil: I'll be around four tournament start, but you aware of how to stop/start the scoring script on cdo?
21:09:16 <gammafunk> s/four/for/
21:09:33 <advil> maybe
21:09:52 <advil> I know how to do it on my own computer at least
21:09:52 <gammafunk> to start the script: python taildb.py
21:09:52 <advil> the script is taildb.py
21:09:52 <advil> ?
21:09:52 <advil> right
21:09:56 <gammafunk> to stop it just kill that process
21:10:14 <advil> ok
21:10:17 <gammafunk> and taildb.log has the daemonized script's output
21:10:28 <gammafunk> which you can tail -f in a window or whatnot
21:10:56 <gammafunk> and to do a one-off processing of games if taildb.py isn't running and you want to test: python loaddb.py
21:11:19 <gammafunk> the output from that just goes right to your terminal
21:11:32 <gammafunk> the only thing we have left is cao/cbro setup
21:11:32 <advil> ok yeah, this all makes sense
21:11:44 <advil> johnstein has updated cbro and just needs to restart webtiles
21:11:49 <gammafunk> oh, great
21:11:56 <advil> he said he might get to cao tonight, if he doesn't fall asleep first
21:12:09 <gammafunk> heh
21:27:02 <gammafunk> hrm, I don't see the 0.23 meta dir on cbro
21:27:08 <gammafunk> so I guess he hasn't configured all of that
21:29:35 <alexjurkiewicz> wow. little toy lto prototype results in 20% better performance *on the same qw run I instrumented*
21:29:57 <alexjurkiewicz> at least it works. Now to test with abyss.stress or something that hits more functions
21:30:10 <advil> fireworks.rc
21:30:30 <alexjurkiewicz> although tbh a qw run dying to d:4 ogre is probably VERY representative of typical hot code paths :D
22:17:28 <gammafunk> ebering: I put a code of conduct on the tournament page with language adapted from your csdc one and the one already on said page (which I forgot we had)
22:29:33 <johnstein> gammafunk, oops forgot to symlink that
22:29:40 <gammafunk> oh, ty
22:33:18 <ebering> gammafunk: cool
22:33:47 <johnstein> gammafunk, cbro meta files should be live
22:33:57 <johnstein> I should get to cao later tonight
22:34:01 <gammafunk> johnstein: thanks, I can see them
22:34:18 <gammafunk> although I wonder why the games in those files are still beta
22:34:52 <gammafunk> maybe it just has from before the release and will update later
22:35:04 <gammafunk> either way the files are working
22:41:35 <alexjurkiewicz> hey! you didn't list me in the 0.23 contributor list!
22:42:48 <gammafunk> he did actually
22:43:21 <gammafunk> it's sorted by last name
22:43:36 <alexjurkiewicz> oh, ok
22:43:39 <gammafunk> well for names that have last names
22:43:50 <alexjurkiewicz> teaches me not to cmd-f
22:43:58 <alexjurkiewicz> thanks eb <3
22:51:28 <alexjurkiewicz> in test/stress/run, top line says: CRAWL=${CRAWL:-timeout 655 ./crawl -seed 1 -no-save -name test -wizard -no-throttle}
22:51:31 <alexjurkiewicz> what is -timeout??
22:51:44 <alexjurkiewicz> I don't see it in the codebase anywhere
22:52:05 <gammafunk> looks like fakepty
22:52:27 <gammafunk> you'll probably see that being defined as the start of CRAWL somewhere earlier
22:52:41 <gammafunk> alexjurkiewicz: ^
22:52:52 <alexjurkiewicz> aha
22:52:58 <floraline> i'm a first time contributor but i was not given the newcomer italics
22:53:09 <floraline> i'm guilty of stolen valor
22:53:26 <gammafunk> I never recieved the big twitch donation you promised
22:53:33 <gammafunk> hence you don't get the delux italics
22:53:39 <gammafunk> those cost money
22:53:51 <alexjurkiewicz> hm. fake_pty.c doesn't have any -timeout option either:         fprintf(stderr, "Usage: fake_pty program [args]\n");
22:54:08 <gammafunk> how is CRAWL defined?
22:55:04 <gammafunk> oh
22:55:09 <gammafunk> that's the "timeout" shell command
22:55:17 <gammafunk> that syntax does not include the '-'
22:55:27 <gammafunk> that's bash syntax for specifying a default value of a variable
22:55:45 <gammafunk> so it's unix shell command timeout
22:56:06 <alexjurkiewicz> bash has a timeout builtin??
22:56:06 <gammafunk> alexjurkiewicz: ^
22:56:06 <gammafunk> not builtin
22:56:10 <gammafunk> just a shell utility
22:56:11 <alexjurkiewicz> ohhh
22:56:15 <alexjurkiewicz> right, yes
22:56:38 <gammafunk> part of coreutils I guess
22:56:56 <gammafunk> floraline: actually I bet this isn't even your first version contributing!
22:57:02 <gammafunk> let's see the evidence...
22:57:35 <floraline> the password reset was my first in september
22:57:52 <floraline> that's when it was accepted, anyway
22:58:06 <ebering> I spent some time researching that with git log and show
22:58:37 <gammafunk> man it's annoying how git log will include pull request commits
22:58:43 <gammafunk> wonder if I can exclude that
23:01:25 <gammafunk> floraline: yeah your first password reset commit shows up as 0.22-a for me yet I think that was in 0.23-a
23:02:20 <gammafunk> I've added italics
23:02:28 <floraline> oh interesting
23:03:31 <gammafunk> maybe that's a PR commit
23:03:44 <floraline> my commits in the PR were prior to 0.22, but it was merged into master afterwards
23:12:44 <alexjurkiewicz> git log --author=floraline 99613e666e6e44d419c74304adb359f8b92996a6..0.23-a0
23:12:46 <alexjurkiewicz> vs
23:12:48 <alexjurkiewicz> git log --author=floraline 0.23-a0..HEAD
23:12:53 <alexjurkiewicz> (not sure if there's a simpler way to write that first one)
23:14:24 <gammafunk> git log --author=floraline --no-merges --first-parent
23:14:27 <gammafunk> worked fine
23:14:59 <ebering> ah
23:15:02 <ebering> dang
23:15:29 <gammafunk> it's obvious that floraline was trying to game the system
23:15:40 <gammafunk> by all rights he should be banned
23:16:09 <floraline> i admit nothing
23:24:53 <floraline> i will try to get a pr together eventually to meet the demands of alexjurkiewicz regarding webtiles config enhancements
23:26:51 <floraline> i'll move as little as possible into a new reloadable config file and just post it to see what feedback i get
23:28:54 <alexjurkiewicz> good. see that you do.
23:29:30 <alexjurkiewicz> oh god,playing 8 qw games onto the same terminal is hell
23:29:54 -!- amalloy_ is now known as amalloy
23:38:07 <johnstein> gammafunk, I bet it's because sometimes the build scripts don't pull tags
23:38:35 <johnstein> I just did a git fetch in the repo and now I have the tags
23:38:36 <gammafunk> interesting
23:38:55 <gammafunk> I guess technically they might have some vlong values that you wouldn't find on any other server
23:39:05 <gammafunk> but shouldn't cause any real problems I think
23:39:08 <johnstein> oh hm.
23:39:16 <johnstein> &versions
23:39:28 <gammafunk> well, I guess one thing is that they'd be marked as alpha and not stable maybe
23:39:40 <gammafunk> but yeah I'm not sure how the dgl scripts manage tags
23:39:51 <johnstein> I think it's happened before. I have a note somewhere that tells me to git fetch before updating to a new version but I forgot to do that this time
23:39:53 <gammafunk> oh, I need to add 0.23 cbro to sequell
23:39:57 <Sequell> CAO: 0.24-a0-91-g4e4936a, CBRO: 0.24-a0-84-g25a2d9f, CDO: 0.24-a0-73-g81bb1ad, CKO: 0.24-a0-91-g4e4936a574, CPO: 0.24-a0-91-g4e4936a, CUE: 0.24-a0-89-gce0dbf48b2, CWZ: 0.23-a0-729-g94ec030, CXC: 0.24-a0-84-g25a2d9f, LLD: 0.24-a0-89-gce0dbf48b2