00:09:39 <Eksell> Unstable branch on CRAWL.XTAHUA.COM updated to: 0.21-a0-394-gdc75ea7 (34)
01:15:05 <aidanh> I had a look at the monster tiles, and most monsters don't have visible pupils; apart from the eye monsters, there's only lom lobon, sun demons, unseen horrors and big-eyed pan lords
01:16:54 <aidanh> It'd be possible to have their tiles be pupil-less, and then draw their pupil from a separate sprite; each monster would need some additional data to define the oval path the pupul should be placed upon
01:19:07 <aidanh> A much easier and more effective improvement would be to flip monster tiles horizontally so that they face the player
01:20:49 <Gretell> Unstable branch on crawl.develz.org updated to: 0.21-a0-394-gdc75ea7 (34)
02:00:50 <Gretell> Windows builds of master branch on crawl.develz.org updated to: 0.21-a0-394-gdc75ea7
02:55:18 <Gretell> Monster database of master branch on crawl.develz.org updated to: 0.21-a0-394-gdc75ea7
03:15:08 <Rotatell> Unstable branch on crawl.beRotato.org updated to: 0.21-a0-394-gdc75ea7 (34)
04:10:10 -!- amalloy is now known as amalloy_
04:13:36 -!- amalloy_ is now known as amalloy
04:42:20 -!- amalloy is now known as amalloy_
05:25:43 -!- amalloy_ is now known as amalloy
06:29:32 <Cheibriados> Ghoul Cold Resistance Missing 13https://crawl.develz.org/mantis/view.php?id=11262 by ojifijod
07:00:50 <Cheibriados> New branch created: pull/622 (1 commit) 13https://github.com/crawl/crawl/pull/622
07:00:50 <Cheibriados> 03nlavsky02 {GitHub} 07https://github.com/crawl/crawl/pull/622 * 0.21-a0-395-gcc855ca: Fix message displayed when Gh tries to quaff potion of mutation 10(7 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/cc855ca458e1
07:14:05 <Rotatell> ziggurnaut (L16 OpWz)  (Swamp:3)
07:16:31 <Cheibriados> 03nlavsky02 {GitHub} 07https://github.com/crawl/crawl/pull/622 * 0.21-a0-396-g661444b: Clarify function's description 10(2 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/661444b490bb
07:21:08 <Cheibriados> 03nlavsky02 {|amethyst} 07* 0.21-a0-395-g39bf9ee: Fix message displayed when Gh tries to quaff potion of mutation (#622) 10(7 seconds ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/39bf9ee3b35d
08:21:21 <Jorgrell> Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-395-g39bf9ee (34)
08:57:11 <advil> getting a lot of travis jobs timing out on even starting these days
08:57:11 <Sequell> advil: You have 2 messages. Use !messages to read them.
09:13:36 <Cheibriados> 03Aidan Holm02 {advil} 07* 0.21-a0-396-ga924934: Fix tiles UI not being redrawn on expose 10(10 hours ago, 1 file, 5+ 0-) 13https://github.com/crawl/crawl/commit/a924934980fa
09:18:12 <advil> hm, for github to detect & automatically close, I guess it needs to be not rebased at the command line?
09:36:48 <|amethyst> yeah... you can close in the commit message ("Closes #666") but that marks it as red rather than purple :/
09:38:16 <|amethyst> the only way I know of to rebase and still keep it purple is to use the buttons on the website ("Squash and merge" or "Rebase and merge")
09:38:43 <|amethyst> if you want to do editing and then rebase, I think you'd have to modify the pull request
09:53:34 <advil> if I do a no-ff merge after rebassing and get the trivial merge commit, do you know if it detects that?
09:53:55 <advil> *rebasing
10:05:23 <travis-ci> The build has errored. (master - a924934 #8836 : Aidan Holm): https://travis-ci.org/crawl/crawl/builds/289501538
10:08:44 <travis-ci> The build has errored. (master - a924934 #8836 : Aidan Holm): https://travis-ci.org/crawl/crawl/builds/289501538
10:09:29 <advil> hm, travis is getting 401 errors when downloading debian stuff
10:21:26 <Jorgrell> Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-396-ga924934 (34)
10:52:52 <advil> looks like the packagecloud account we were getting packages from (https://packagecloud.io/basho/riak/ubuntu) no longer exists? was that an official distribution? something crawl-specific?
10:55:37 <advil> hm, seems like that company may be dead
10:55:40 <advil> https://www.theregister.co.uk/2017/07/13/will_the_last_person_at_basho_get_the_lights_oh_too_late/
10:56:32 <advil> well, its assets were bought by some other company
12:12:55 <aidanh> So I have a WIP patch that flips monster sprites to face the player as necessary; it works wonderfully on animals like frogs and yaks
12:13:45 <aidanh> But I've realized to get this to work with pan lords and most humanoids, it'd be necessary to show weapons on the right-hand side of the tile, and shields on the left
12:14:18 <aidanh> Is there any reason that wouldn't be desirable?
12:18:24 <|amethyst> as long as you flip those tiles and their offsets, I don't see a problem
12:18:33 <MarvinPA> i'm not a tiles player but that doesn't seem like desirable behaviour in general, seems like it'd just hurt readability
12:18:48 <MarvinPA> situationally flipping things at all, that is
12:21:35 <aidanh> Right, that's why I'm asking
12:22:03 <|amethyst> probably you'll find more tiles players in places that aren't ##crawl-dev :)
12:22:08 <|amethyst> there are a few though
12:22:48 <|amethyst> could design a study to measure reaction times for monster identification in flipped vs non-flipped orientations
12:23:01 <|amethyst> or mixed vs non-flipped rather
12:23:03 <|amethyst> :)
12:29:18 <aidanh> Hm; any particular reason the devs don't use tiles?
12:29:48 <aidanh> Well, tiles is the sort of game where if you try to rush through, you're going to die anyway
12:30:13 <aidanh> In that sense, a slightly unfamiliar tile that makes you stop and think could be a good thing
12:31:57 <bluedave> i tend to check the list pane if i can't immediately identify a tile
12:38:37 <geekosaur> mostly, been playing roguelikes longer and are used to the console interface, I think
12:39:22 <aidanh> that makes sense
12:49:24 <aidanh> In any case, there's a number of possible variations on the idea: e.g. only flip monsters in the viewport (not the side pane), only flip visually-distinctive monsters, or have an option to enable/disable this
13:11:34 <Henzell> Unstable branch on crawl.akrasiac.org updated to: 0.21-a0-396-ga924934 (34)
13:44:13 <amalloy> !tv GoodDemon3 trhu shoals
13:44:14 <Sequell> 1. GoodDemon3, XL17 TrHu, T:29582 requested for FooTV: telnet://termcast.develz.org or http://termcast.develz.org.
13:48:09 <bluedave> strange, i tossed rocks at a worm and it walked away
13:49:19 <amalloy> bluedave: it was too dumb to path to you across the water and didn't want to let you throw rocks at it forever
13:56:33 <geekosaur> yep. orcs will do that too if they can;t figure out how to path to you and have no ranged attack options
13:56:55 <amalloy> @??donald
13:56:55 <Gretell> Donald (02@) | Spd: 10 | HD: 14 | HP: 71-96 | AC/EV: 13/6 | Dam: 26 | 10weapons, 10items, 10doors, fighter, amphibious, spellcaster | Res: 06magic(100), 12drown | XP: 1971 | Sp: might, haste | Sz: Medium | Int: human.
13:56:58 <geekosaur> and other monsters
13:58:08 <bluedave> what i thought was weird was that there should have been a path in its los, but maybe i thought that from having moved forward and then back.  thanks for answers though :)
14:00:25 <geekosaur> worms are *dumb*
14:00:42 <geekosaur> move-back-to-move-forward is not in their repertoire
14:01:20 <geekosaur> e.g. orcs will do it though
14:04:00 <geekosaur> (that's not even unrealistic. you can see it with dogs for example; some breeds are smart enough to realize they can back up and get a path, others will just keep trying to bull on through the fence or w/e)
14:15:28 <amalloy> donald's fighter flag shouldn't increase his melee damage at all, right? just his to-hit chance?
14:16:06 <amalloy> but might would increase his damage of course
14:17:18 <gammafunk> ??fighter[2
14:17:19 <Sequell> fighter[2/2]: Also a monster flag that increases the accuracy of attacks. Monsters with this flag add their HD to melee to-hit (25*hd/10 versus 15*hd/10 for most monsters). They also get 20% extra damage and to-hit to ranged attacks.
14:19:27 <geekosaur> (I started parsing that as appalachian english o.O )
15:07:38 <Yermak> Guardian golem was tweaked so that allies don't anger upon its explosion. Is it right to implement the same behaviour for explosions from immolation scroll?
15:08:06 <orbekk> Is there some way to subscribe to morgue files on crawl servers? I'm thinking of building a scoreboard
15:11:44 <johnstein> subscribe?
15:12:04 <johnstein> what is the expected use case?
15:12:37 <orbekk> I would like to download all morgue files once, and download all new files once per day
15:13:08 <johnstein> !lg * s=day(end) -graph
15:13:14 <Sequell> 7924231 games for *: https://shalott.org/graphs/a148e20d4c02f582484551ef1236461220f01189.html
15:14:54 <johnstein> there was an attempt several months ago by someone trying to get all the morgue files. it hammered the servers. not sure what the least invasive way to get the list of new files is
15:15:19 <johnstein> you can certainly watch the announcement bots in ##crawl but not all games are announced
15:15:47 <bluedave> orbekk - back to johnstein's question what would the scoreboard be used for?
15:15:53 <johnstein> and most (all?) of the servers don't publish a list of all morgue files as they appear
15:17:25 <amalloy> i imagine the right way to do it would be to watch milestones like Sequell does, and when you see a death milestone then download the corresponding log
15:17:38 <amalloy> but i don't know how hard that is, if you're not Sequell
15:19:27 <orbekk> thanks, just wondered if there was a good existing way. I don't have a strong usecase, just want to explore this because it's cool data
15:19:48 <orbekk> I think I can do it on an ad-hoc per-player basis for now to avoid hammering the server
15:22:58 <johnstein> agreed with amalloy. that's usually how it's done. could probably fork sequell or postquell to get a head start
15:26:56 <amalloy> https://github.com/crawl/sequell/blob/master/config/sources.yml has a list of all the places that servers publish logfiles to, as far as Sequell knows. the logfiles are really long, though - i don't know how Sequell works with the server to avoid requesting a really long file over and over
15:34:37 <johnstein> it only wgets the new lines
15:35:10 <johnstein> somehow remembers how many lines were pulled last time then starts at n+1 next time it wgets
15:37:49 <amalloy> lines, or bytes?
15:37:52 <amalloy> i could understand bytes
15:39:55 <amalloy> oh cool, wget -c just magically does that for you
15:40:58 <orbekk> nice! maybe I can use that. thanks! <3
15:41:51 <geekosaur> it's just using the file size. since the file is written in line units, you resume at the byte offset corresponding to the next line
15:44:31 <geekosaur> (relying on the http range header, and a hidden unix dependency because the byte offset would be different on windows with \r\n line terminators...)
15:54:29 <amalloy> hah
18:07:09 <Lantell> Unstable branch on underhound.eu updated to: 0.21-a0-396-ga924934 (34)
18:32:24 <Brannock> Hey guys, long time no see. How have things been?
18:47:58 <advil> hi Brannock!
18:56:14 <advil> uh I guess the apparent answer is that things have been quiet?
19:03:33 <Brannock> Hey advil! Seems like it, yeah. Lots of smoothing out rough code and I see Gnoll went in and got a really neat revamp.
19:03:56 <advil> yeah the gnoll thing seems very good, though i haven't yet tried one
19:03:58 <Brannock> They seem popular so far, though I haven't checked tavern yet for the pessimism
19:05:04 <gammafunk> not too much pessimism on tavern, they seem to have gone over well
19:05:13 <Brannock> Amazing :)
19:05:26 <Brannock> gammafunk, you were responsible for the revamp, along with a few others I think?
19:05:27 <advil> what pessimism there is there is like "maybe they're too easy" or "maybe they aren't varied enough after a few wins" or something
19:05:38 <advil> they're trying, but not quite getting to some really good pessimism
19:05:57 <gammafunk> I would say it was my impetus to see that elliptic/mikee had a good idea, so made the initial implementation
19:06:09 <gammafunk> but elliptic had to do some nice work on skilling to improve things
19:06:19 <Brannock> I see. Well it's a better idea than I could have ever come up with, I"m really impressed with the approach y'all took
19:06:32 <gammafunk> yeah it was nice how it worked out
19:06:47 <gammafunk> given how poorly recieved the previous iteration was :)
19:07:04 <gammafunk> we have a dev plan
19:07:05 <gammafunk> ??plan
19:07:06 <Sequell> plan[1/2]: https://github.com/crawl/crawl/wiki/0.21-Development-and-Release-Plan
19:07:13 <gammafunk> there are a few things that still need writing up
19:07:47 <gammafunk> Wu removal + moving at least one of his abilities to Usk, I think we were agreed that heaven on earth would be good
19:07:58 <gammafunk> and then maybe using lunge somehow, but that's less clear
19:15:14 <alexjurkiewicz> ASSERT(you.skills[sk] == 0 || !is_useless_skill(sk)) in 'skills.cc' at line 276 failed.
19:15:37 <alexjurkiewicz> https://crawl.project357.org/morgue/Bunnehwyld/crash-Bunnehwyld-20171018-225214.txt
19:15:40 <alexjurkiewicz> is this a known issue?
19:15:46 <alexjurkiewicz> oh heh, it's dcss-ca
19:17:26 <advil> fr darkness magic
19:17:31 <advil> also time magic
19:18:32 <gammafunk> space magic?
19:19:21 <advil> the skill values for dodging in that crash look completely screwed up, but yeah, I don't think I've seen that in mainline
19:20:18 <advil> just to check, it seems like `unscale_round_up` is actually a ceiling function?
19:20:23 <advil> heh it's only used in this one place
19:20:43 <advil> so probably no one else has ever heard of it besides me
19:20:58 <amalloy> advil: yes, that is what it looks like
19:21:12 <advil> wonder if it really makes sense to do a ceiling when doing this ev calculation
19:21:24 <alexjurkiewicz> advil: actually dcss-ca skills can go to 99
19:22:10 <advil> I mean that it has 0 skill points but 3 levels in dodging
19:22:31 <advil> I guess it's not obvious that that would trigger that assert
19:22:55 <alexjurkiewicz> ah heh
19:23:17 <alexjurkiewicz> i should play dcss-ca again sometime. it can lead to some very funny experiences
19:25:58 <Brannock> ghost vaults sound a lot more intriguing to me than current ghost status
19:27:50 <jmart_> How do I go about submitting a bugfix? New here
19:28:05 <Brannock> You can post a PR on the github or through mantis
19:28:25 <advil> github is usually a bit easier these days
19:28:39 <jmart_> roger roger
19:28:57 <advil> what's the bug?
19:29:18 <jmart_> Wizmode bug; you can't give yourself mutation resistance
19:29:27 <jmart_> because the way it's ignored is by resetting it when you leave the function
19:29:38 <jmart_> Not exactly important
19:30:46 <amalloy> good bug
19:31:06 <jmart_> Is there a list of extant bugs? I get the impression there aren't many
19:31:23 <advil> oh there are bugs
19:31:28 <Brannock> plenty
19:31:46 <advil> we use mantis for issue tracking, and there's tons of stuff there. you have to sort through it a bit though
19:32:00 <advil> https://crawl.develz.org/mantis/view_all_bug_page.php
19:32:12 <jmart_> Cool, thanks
19:32:33 <Brannock> some of the older bugs might have been solved indirectly by now
19:33:16 <amalloy> some others get reported anew every couple months because we don't know how to fix them
19:33:32 <amalloy> i would love it if someone figured out how to fix the ashenzari autoexplore bug
19:34:59 <jmart_> The autoexplore looping in shoals? marked as fix
19:35:59 <yesno> ??tv
19:35:59 <Sequell> tv[1/1]: <simmarine> imagine if they had tv in real life
19:36:00 <yesno> ??footv
19:36:00 <Sequell> FooTV[1/5]: telnet termcast.develz.org (and then select the letter for FooTV), or http://termcast.develz.org. Do not use SSH. Shows games requested using !tv. Request games by adding -tv to !lg or !lm. Cancel games by adding -tv:cancel to the same command-line, clear playlist with !lg * -tv:nuke
19:36:11 <yesno> ??footv[2
19:36:11 <Sequell> FooTV[2/5]: Use -tv:<N to start playback approx Nx before the default start (-tv:<3 to start 3x before normal start), and for milestones, -tv:>N to end playback Nx after normal end, -tv:>$ to end playback at end of ttyrec. You can combine both seek options as -tv:<2:>3
19:36:41 <yesno> ??footv[3
19:36:41 <Sequell> FooTV[3/5]: For good fooTV, ??embarrassing deaths, ??hilarious deaths, ??amazing games, ??great near misses
19:37:06 <yesno> !lm . -tv:>$
19:37:07 <Sequell> 2523. yesno, XL4 VpAs, T:3695 (milestone) requested for FooTV: telnet://termcast.develz.org or http://termcast.develz.org.
19:37:24 <amalloy> jmart_: no, the one where you can autoexplore past sleeping monsters sometimes
19:37:33 <yesno> oh sorry
19:37:48 <yesno> i'm so embarrassed
19:38:14 <yesno> sequell tab next to this one...
19:38:32 <amalloy> jmart_: https://crawl.develz.org/mantis/view.php?id=6503
19:40:00 <jmart_> I'll take a look
19:40:04 <jmart_> thanks all
19:40:41 <advil> hm, ROUND_RANDOM for stepdowns is literally never used
19:56:06 <advil> hm, this stepdown overload is not great
19:57:00 <advil> int, int and double, double both work, but the arguments mean completely different things
19:57:43 <advil> oh nm
19:57:49 <advil> they mean the same thing, I just can't read
20:42:03 <Eksell> pipel (L27 DsWr) ASSERT(desc) in 'describe.cc' at line 3166 failed. (Geh:7)
20:44:43 <advil> !crashlog
20:44:57 <Sequell> 17965. pipel, XL27 DsWr, T:163160 (milestone): http://crawl.xtahua.com/crawl/morgue/pipel/crash-pipel-20171019-004142.txt
20:45:32 <alexjurkiewicz> attempting to x-v a gateway to hell?
20:46:27 <alexjurkiewicz> !lm pipel
20:46:27 <Sequell> 3950. [2017-10-19 00:44:13] pipel the Conqueror (L27 DsWr of Makhleb) entered the Ecumenical Temple on turn 163441. (D:5)
20:46:38 <alexjurkiewicz> full level 27 in all skills by 163k turns. seems impressive
20:47:17 <alexjurkiewicz> 90 zigs completed
20:47:25 <advil> it was a monster description of some kind
20:47:47 <alexjurkiewicz> !seen pipel
20:47:48 <Sequell> I last saw pipel at Thu Jun 23 21:37:59 2016 UTC (about 1y 16w 6d 3h 9m 48s ago) joining the channel.
20:54:28 <advil> it looks like an upgrading issue, where a salamander with firebrand is still around somewhere
20:54:52 <advil> actually, any xv of a monster with a removed attack attribute should trigger a crash there
20:56:00 <advil> unless there was something else with firebrand
20:56:30 <advil> though there aren't any monsters in sight?
20:56:40 <advil> except a wraith
21:00:34 <advil> maybe they were looking something up in ?
21:01:57 <advil> ohh they were using X to look at geh 7
21:02:47 <advil> so my salamander firebrand theory is probably right
21:05:29 <advil> 143 removed MONS_CHAOS_CHAMPIONs
21:05:29 <advil> 4 removed MONS_SALAMANDER_FIREBRANDs
21:05:32 -!- mikee__ is now known as mikee_
21:06:12 <advil> I guess if they reentered geh 7 those monsters would probably just go away
21:14:01 <alexjurkiewicz> ah, but they're still in the 'last seen cache'
22:14:46 <alexjurkiewicz> hellmonk: the doom clock instantly killing you in speedrun mode is a little harsh IMO, I'd prefer massive drain or something (theoretically) survivable
22:14:59 <hellmonk> hmm, maybe
22:15:16 <hellmonk> the whole point is for it to be a super fast mandatory dive though
22:15:45 <hellmonk> could rework it into severe maxhp rot or something
22:16:57 <alexjurkiewicz> even if the limit became 900 with harsher penalities or something
22:17:15 <alexjurkiewicz> it's just that the binary ok/dead is going to result in some probably not-good deaths
22:19:09 <hellmonk> yeah, maybe maxhp rot every few turns would be better, give a tiny amount of buffer at least
22:19:31 <advil> is it a per level limit or an overall limit?
22:19:39 <hellmonk> per-level
22:20:05 <nikheizen> getting killed 1 turn from the stairs or while descending will be super sad
22:20:11 <hellmonk> imagine that its going to take a while to get good balance parameters on it
22:20:29 <hellmonk> but it's also a bit of an afterthought for me, it just sounded cool to do a mode where you have to dive hard
22:20:41 <advil> maxhp rot per turn with a log curve of some sort could be pretty neat, you could see the urgency ramping up
22:20:48 <alexjurkiewicz> IMO make the limit much smaller. with 1k turns I'm mostly exploring all of levels
22:21:20 <hellmonk> yeah, 1k turns with double exp might be too generous
22:21:37 <hellmonk> will push some changes before it hits servers
22:21:38 <advil> log growth of rot that is
22:21:57 <hellmonk> maybe like 600-700 turns with gradually worsening maxhp rot afterwards
22:22:05 <advil> er
22:22:14 <advil> I really just mean exponential growth
22:24:18 <alexjurkiewicz> ^ :)
22:24:27 <alexjurkiewicz> just quadratic
22:38:58 <advil> 2^(turns-1) where turns is the number of turns since the time limit
22:44:45 <alexjurkiewicz> hellmonk: wow, even portals work in speedrun mode, nice
22:45:57 <hellmonk> yeah I think everything works, I believe the abyss timer resets on repeat abyss visits but I'm not concerned about that
22:49:42 <alexjurkiewicz> damn you only get 1k turns to escape the abyss? that will be funny
22:52:24 <hellmonk> well in a normal turncount speedrun game you'd probably ctrlQ if you got a 1k turn abyss
22:52:30 <hellmonk> so I'm not too concerned
22:54:10 <alexjurkiewicz> yeah
22:58:22 <hellmonk> gonna try and code up a skeleton species for halloween
23:10:26 <alexjurkiewicz> You start resting.
23:10:27 <alexjurkiewicz> _HP restored.
23:10:27 <alexjurkiewicz> The dungeon collapses into oblivion!
23:10:27 <alexjurkiewicz> You die...
23:10:36 <alexjurkiewicz> ^ auto-resting on downstairs
23:11:07 <alexjurkiewicz> !tell hellmonk in speedrun mode I want to auto-rest on downstairs until 1 turn, then take the downstairs, if I don't have full hp (needs care in case monsters appear)
23:16:52 -!- molotove1 is now known as molotove
23:59:22 <Eksell> Unstable branch on CRAWL.XTAHUA.COM updated to: 0.21-a0-396-ga924934 (34)