00:01:26 <twelwe> banish desc is odd because you could potentially add `Attempts to` to every ability/spell with a failure chance
00:02:08 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.23-a0-890-gc1834021d8 (34)
00:33:23 <alexjurkiewicz> advil: exciting updates to seeding!!
00:33:54 <Eksell> Unstable branch on CRAWL.XTAHUA.COM updated to: 0.23-a0-890-gc183402 (34)
01:15:52 <Cheibriados> 03NormalPerson702 {ebering} 07* 0.23-a0-891-g16004d8: Make yellow draconians' rAcid a true mutation (11381) 10(20 hours ago, 7 files, 30+ 9-) 13https://github.com/crawl/crawl/commit/16004d8ef05e
01:28:06 <Gretell> Unstable branch on crawl.develz.org updated to: 0.23-a0-890-gc183402 (34)
01:52:56 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.23-a0-891-g16004d8ef0 (34)
01:58:50 <Gretell> Windows builds of master branch on crawl.develz.org updated to: 0.23-a0-891-g16004d8
02:56:54 <alexjurkiewicz> advil: i removed the #ifndef DGAMELAUNCH around pregen_dungeon in options.h and startup.cc, but I still can't get a webtiles build to pregenerate the dungeon with -extra-options-after pregen_dungeon=trie
02:56:54 <alexjurkiewicz> am I missing something?
02:59:43 <Gretell> Monster database of master branch on crawl.develz.org updated to: 0.23-a0-891-g16004d8
03:03:56 <alexjurkiewicz> oh i missed initfile.cc
03:41:41 -!- amalloy is now known as amalloy_
03:47:33 <alexjurkiewicz> it works!
03:48:16 <alexjurkiewicz> man I'm glad I listened to advil and didn't jump the gun and announce this before pregen dungeon worked,
04:01:53 <gammafunk> well I would be careful still
04:02:18 <gammafunk> it's still a WIP for the webtiles side, you might want to talk with him again about remaining issues
04:02:54 <gammafunk> and in fact didn't you already announce it in multiple places!
05:09:20 <Cheibriados> 03gammafunk02 07* 0.23-a0-892-g479ecb9: Adjust an Okawaru altar vault 10(5 hours ago, 1 file, 5+ 9-) 13https://github.com/crawl/crawl/commit/479ecb9a15b6
05:09:20 <Cheibriados> 03gammafunk02 07* 0.23-a0-893-g5d1f0fd: Remove a placement limitation on a Lair vault 10(2 hours ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/5d1f0fd2dd0c
05:09:20 <Cheibriados> 03gammafunk02 07* 0.23-a0-894-gc028996: Adjustments to a Lair vault 10(2 hours ago, 1 file, 9+ 10-) 13https://github.com/crawl/crawl/commit/c028996d07d0
05:09:20 <Cheibriados> 03gammafunk02 07* 0.23-a0-895-g27a1b2c: Adjustments to a Lair end vault 10(5 hours ago, 1 file, 5+ 7-) 13https://github.com/crawl/crawl/commit/27a1b2c1d850
05:09:20 <Cheibriados> 03gammafunk02 07* 0.23-a0-896-g250879e: Adjustments to a runed door vault 10(5 hours ago, 1 file, 8+ 13-) 13https://github.com/crawl/crawl/commit/250879e67528
05:44:46 -!- muffindrake2 is now known as muffindrake
06:02:15 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.23-a0-896-g250879e675 (34)
10:23:15 <advil> no disastrous crashes so far
10:25:57 <gammafunk> &versions
10:26:20 <gammafunk> %git HEAD~5
10:26:20 <Cheibriados> 07NormalPerson702 {ebering} * 0.23-a0-891-g16004d8: Make yellow draconians' rAcid a true mutation (11381) 10(29 hours ago, 7 files, 30+ 9-) 13https://github.com/crawl/crawl/commit/16004d8ef05e
10:26:23 <gammafunk> %git HEAD~6
10:26:23 <Cheibriados> 07advil02 * 0.23-a0-890-gc183402: swap true and false (vt) 10(11 hours ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/c1834021d887
10:26:24 <Sequell> CAO: 0.23-a0-860-g8ff03fb, CBRO: 0.23-a0-891-g16004d8, CDO: 0.23-a0-890-gc183402, CKO: 0.23-a0-896-g250879e675, CPO: 0.23-a0-896-g250879e, CUE: 0.23-a0-767-gdc19b93d68, CWZ: 0.23-a0-729-g94ec030, CXC: 0.23-a0-890-gc183402, LLD: 0.23-a0-860-g8ff03fb2f7
10:26:47 <gammafunk> seems so
10:27:19 <advil> really the rng stuff, which is the bulk of what's enabled online, is pretty innocuous, except that it did involve a change in save format
10:27:54 <gammafunk> we also have almost two weeks before release, thankfully
10:28:00 <advil> yeah
10:28:59 <advil> the pregen stuff ... could have some quirks. (But nothing like when I tried to get a staged version working that only pregenerated as needed, and the lair exit led to lair:6 for reasons I still don't understand!)
10:29:32 <gammafunk> weird
10:30:49 <advil> it turns out that when that happens, qw gets very confused
10:33:12 <advil> I think for a test that we can leave enabled I might make a cheating version of qw that uses &G, &{, and &H instead of all that complicated slow stuff
10:54:23 <Cheibriados> New branch created: pull/970 (1 commit) 13https://github.com/crawl/crawl/pull/970
10:54:23 <Cheibriados> 03mgdelmonte02 07https://github.com/crawl/crawl/pull/970 * 0.23-a0-847-gaab08a6: added more monster info to Lua -- AC, EV, MR, and MaxHP -- optional "full description" option for desc() -- defeat_mr(spell, evoked) gives your chance to defeat monster's MR 10(2 minutes ago, 1 file, 117+ 4-) 13https://github.com/crawl/crawl/commit/aab08a6ff878
12:27:56 <Henzell> Unstable branch on crawl.akrasiac.org updated to: 0.23-a0-896-g250879e (34)
12:34:56 <Cheibriados> Most buttons on random character selection screen are treated as 'y' 13https://crawl.develz.org/mantis/view.php?id=11844 by pedant
12:50:29 <ebering> speaking of changelogs
12:51:11 <advil> I need to add something about seeding I suppose
12:56:47 <ebering> yep; and we should get it updated
12:56:59 <ebering> probably post a last trunk-updates post of the cycle
13:29:53 <gammafunk> ebering: I request you include a few joke changelog entries
13:29:59 <gammafunk> a tradition I think PF really first started
13:30:20 <gammafunk> you know, like you make some about a relatively minor change and make it sound weird or funny
13:30:34 <gammafunk> usually said commit was already a bit humorous
13:30:44 <gammafunk> people reading the updates like those, at least
13:38:50 <Cheibriados> 03NormalPerson702 07https://github.com/crawl/crawl/pull/966 * 0.23-a0-840-ga5c7d1f: Use keys for paralysed_by and petrified_by player props (ebering) 10(55 seconds ago, 4 files, 18+ 10-) 13https://github.com/crawl/crawl/commit/a5c7d1f57516
16:34:13 -!- aidanh_ is now known as aidanh
16:35:33 <switch__> i'm pretty sure a sleeping ugly thing (that i confused before-hand) just sleepwalked
16:35:59 <switch__> is that a thing?
16:36:57 <Pinkbeast> I'd check the ttyrec first
16:37:24 <geekosaur> sleepwalking while confused is a thing
16:41:48 <switch__> translucent runed doors make yuif look kinda strange
16:41:52 <switch__> running around in his little hut
16:44:13 <switch__> https://i.imgur.com/eN9l1rs.png lol, does this vault normally have a potion of agility in it?
17:01:42 <alexjurkiewicz> no
17:07:44 <switch__> well, that one did :P
17:58:38 <Cheibriados> Long filepaths can overflow and overlap text 13https://crawl.develz.org/mantis/view.php?id=11845 by CanOfWorms
18:13:00 <advil> bleh the windows trunk build isn't doing the same thing as linux/os x on a seed
18:13:16 <advil> it does have the same arrival vault but seems to diverge after that
18:14:33 <Cheibriados> Bribing behaviour issues 13https://crawl.develz.org/mantis/view.php?id=11846 by CanOfWorms
18:14:53 <advil> !flip windows
18:14:54 <Sequell> (╯°□°)╯︵ sʍ°puᴉʍ
18:15:17 <advil> well, at least now it's a bug to fix rather than a new feature to add!
18:23:34 <alexjurkiewicz> :D weird that it only slowly diverges. Does your "10 random numbers from a seed"test diverge?
18:24:35 <advil> I'm not sure how to run that on windows...I doubt it does though, I would guess based on the initial behavior (temple vault chosen is also the same) that the rng is fine
18:24:41 <advil> it's something like the ordering bug I already fixed
18:24:56 <advil> in fact, probably windows does something insane with filenames and they get sorted differently
18:56:07 <switch__> what's different about the windows build?
18:56:36 <switch__> (obviously a lot of things, but, do you have an idea what might be relevantly different?)
18:59:18 <alexjurkiewicz> you'd think sorting a bunch of ascii file names would work the same between linux and osx and windows though...
18:59:30 <alexjurkiewicz> do you need a windows dev machine btw?
19:00:20 <CanOfWorms> windows is weird with number sort
19:00:49 <CanOfWorms> explorer will sort 00002.png after 1.png because it interprets the first as "2"
19:01:15 <alexjurkiewicz> %git 641d409e0c94b4334e4be646ac7d5871fbf8daf5
19:01:15 <Cheibriados> 07advil02 * 0.23-a0-880-g641d409: Use a stable order for loading vault files 10(2 days ago, 3 files, 13+ 5-) 13https://github.com/crawl/crawl/commit/641d409e0c94
19:01:24 <alexjurkiewicz> not explorer :)
19:04:31 <CanOfWorms> no, I'm saying that the behaviour I described is exactly what windows explorer does
19:46:49 <advil> alexjurkiewicz I can do windows dev stuff in a vm, why, do you have access to something?
19:47:03 <advil> it's just a bit of a slog
19:47:16 <advil> as to what's different, all I could see with my quick check was the results (different vault choices), so I'm not sure
19:47:53 <advil> I spot-checked the .des files for anything obviously suspicious, but this is probably pretty subtle
19:55:19 <alexjurkiewicz> i was playing with an aws workspace that I won't be using, but I have to pay for the rest of the month
19:55:39 <alexjurkiewicz> so there's like 10 days of it left
19:55:58 <gammafunk> oh does aws offer windows servers now?
19:57:29 <advil> oh, I see...I think I'm probably ok with a vm
19:57:45 <advil> thanks though
19:59:06 -!- amalloy_ is now known as amalloy
20:04:36 <alexjurkiewicz> gammafunk: it has for yearrrrs
20:04:59 <alexjurkiewicz> they sell windows desktops (a la citrix) as of a couple of years, I was trialling that
20:06:15 <gammafunk> alexjurkiewicz: oh, I guess I don't know how those work or are priced differently than e.g. the linux instances
20:06:35 <gammafunk> like are they full virtual servers or what; guess I can go read more about that, but good to know, thanks
20:16:12 <switch__> i've used their windows servers for something like 5 years now
20:29:57 <alexjurkiewicz> EC2 can run windows server, you pay a bit extra per hour for the windows license. Workspaces runs presumably on top of EC2 but the pricing is different: monthly upfront + hourly fee (more for Windows license)
20:37:43 <minmay> nearly every webtiles screenshot I see has the tiles mutilated by scaling to 40x40 or so instead of their original 32x32. why is this?
20:38:12 <gammafunk> are people playing with browser zoom on?
20:38:49 <gammafunk> maybe it's something a popular screen cap software is doing
20:39:51 <minmay> it doesn't look like browser zoom because there's no interpolation
20:40:18 <minmay> consistently no interpolation
20:41:05 <CanOfWorms> the browser is rescaling the view so it fits the window
20:41:08 <minmay> some of the pixels just turn into 2x2 squares
20:44:55 <gammafunk> CanOfWorms: I believe the client rendering code changes the number of tiles displayed and I think adjusts padding so that all the components fit whatever window size; I don't think tile size changes as the window size does
20:45:46 <gammafunk> so if you have a larger window you just see more tiles out of LOS, not bigger tiles, generally speaking
20:45:47 <CanOfWorms> no, it'll scale tiles
20:45:49 <CanOfWorms> I just checked
20:46:05 <gammafunk> I mean, my display is not scaling tiles
20:46:31 <CanOfWorms> it'll definitely adjust the number of tiles if you mess with the width:height ratio but if you shrink the window the tiles will shrink
20:47:01 <gammafunk> yeah, but that's shrinking
20:47:40 <gammafunk> that wouldn't lead to the kind of screenshots being talked about here, right
20:47:51 <gammafunk> I think as it becomes larger it's just adding tiles
20:48:05 <minmay> right, that's what confuses me, this hasn't happened to me when I've tried webtiles, in either firefox or chrome
20:48:16 <gammafunk> once you meet a size where it doesn't have to downscale, that is
20:48:41 <minmay> they stay 32x32
20:48:41 <minmay> yes
20:49:13 <|amethyst> perhaps these people have tile_cell_pixels set?
20:49:13 <CanOfWorms> well, I can't test what happens past 32x32
20:49:20 <CanOfWorms> my window's maximum tile size is 28x28
20:49:48 <gammafunk> you can't see tiles larger than 28x28? what kind of display do you have?!
20:49:54 <|amethyst> gammafunk:  and, yeah, CSZO for the tournament should be fine, but I'd keep everything backed up just in case
20:49:58 <advil> people might have huge displays
20:50:04 <gammafunk> |amethyst: oh, great, thanks
20:50:07 <CanOfWorms> oh wait if I fullscreen I can get 32x32
20:50:14 <CanOfWorms> any I imagine it still has to scale past 32x32
20:50:22 <CanOfWorms> apple devices have a higher resolution
20:50:25 <CanOfWorms> *anway
20:50:28 <CanOfWorms> *anyway
20:50:28 <gammafunk> and will do
20:50:33 <gammafunk> oh right, mobile
20:50:47 <advil> well, apple devices are hidpi which usually behaves a bit differently
20:51:04 <advil> on some devices hidpi screenshots might look odd when viewed on a non hidpi device though
20:52:12 <gammafunk> CanOfWorms: right, what I'm saying is that instead of scaling it generally just adds more tiles; I'm not sure what would make it decide to start scaling and stop adding tiles
20:52:26 <gammafunk> maybe it has a limit to how many out of los tiles it'd display or something?
20:52:34 <CanOfWorms> if it stops scaling the view will be tiny on mobile
20:52:39 <|amethyst> also, we (by default) explicitly turn off interpolation when tiles are scaled
20:53:08 <|amethyst> s/interpolation/smoothing/
20:54:00 <gammafunk> no, I'm talking about *starting* scaling up to e.g. 40x40 when the display gets larger; obviously it might need to scale down
20:54:00 <gammafunk> if the screen is too small
20:54:01 <|amethyst> I mean, if the player does ctrl-+
20:54:17 <|amethyst> I think the thing we do to turn off smoothing would affect that, too
20:54:30 <gammafunk> yeah, but minmay was saying that this didn't look like browser zoom since non-interpolated (which I guess browsers do if you zoom?)
20:54:45 <CanOfWorms> I mean the mobile resolution can have retina mode
20:54:57 <CanOfWorms> which is double? the normal resolution
20:55:14 <|amethyst> gammafunk:  right, but I think the way we turn off smoothing would also affect that
20:55:17 <CanOfWorms> so if you stop scaling each tile will be very small on such a screen
20:55:22 <|amethyst> this.ctx.imageSmoothingEnabled = options.get("tile_filter_scaling");
20:55:35 <|amethyst> and likewise for webkitImageSmoothingEnabled and mozImageSmoothingEnabled
20:55:40 <ebering> |amethyst: thanks for offering CSZO for the tournament
20:56:18 <minmay> ah ok yeah that would affect browser zoom
20:56:38 <|amethyst> I'm not sure why that option defaults to false though
20:56:44 <|amethyst> speed of local tiles?
20:56:55 <|amethyst> poor smoothing algorithm in local tiles?
20:57:25 <gammafunk> didn't advil do work for hidpi devices that affected this recently?
20:57:38 <advil> mostly on local tiles
20:57:43 <minmay> I think the idea is that smoothed pixel art looks awful
20:57:49 <advil> i did fix some hidpi issues on webtiles but I think that was mostly bugfixing
20:58:01 <minmay> I can't say the alternative I see in these screenshots is actually an improvement over smoothing though
20:59:56 <gammafunk> I will set up the tournament page on cszo like we did for last tournaments, probably some time tomorrow or wednesday in terms of getting a proper page
21:00:29 <gammafunk> ebering: if you want a url to give people, it will be http://dobrazupa.org/tournament/0.23/ ; that won't work until I get the initial setup going, however, and will be 404 until then
21:01:21 <gammafunk> I usually make local commits and push to https://github.com/crawl/dcss_tourney
21:01:53 <gammafunk> and pull on cszo, with some changes just being local and outside the repo, if you're getting familiar with any of that stuff
21:02:18 <gammafunk> but if you don't want to manage it much, that's fine, since it doesn't require too much hands-on
21:02:40 <gammafunk> and of course if we want any rule changes, those have to be agreed upon and implemented
21:02:43 <ebering> gammafunk: ok; I'm merging some more bugfixes now and then will start in on trunk updates and changelog
21:03:10 <gammafunk> cool, and no rush; you can mention url or not mention; I can always make a quick follow-up post about the tourney info
21:03:18 <ebering> yeah
21:04:36 <gammafunk> only devs who have completed a gauntlet will be eligible for the elite dev clan
21:04:42 <gammafunk> so elliptic's application may be rejected outright
21:05:18 <gammafunk> because he is fixed on the past, where Labs still exist
21:05:44 <minmay> my time to shine
21:11:34 <Cheibriados> 03NormalPerson702 {ebering} 07* 0.23-a0-897-g779a845: Include petrification information in death messages (11374) 10(2 days ago, 3 files, 13+ 0-) 13https://github.com/crawl/crawl/commit/779a845ef75d
21:11:34 <Cheibriados> 03NormalPerson702 {ebering} 07* 0.23-a0-898-g0a62fed: Use keys for paralysed_by and petrified_by player props (ebering) 10(8 hours ago, 4 files, 18+ 10-) 13https://github.com/crawl/crawl/commit/0a62fede9633
21:37:18 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.23-a0-898-g0a62fede96 (34)
21:44:37 <Cheibriados> 03Martin Staffa02 {ebering} 07* 0.23-a0-899-gbbae0a3: Further distinguish light skill drain 10(3 days ago, 1 file, 8+ 1-) 13https://github.com/crawl/crawl/commit/bbae0a335adb
21:47:08 <Kelbell> Unstable branch on crawl.kelbi.org updated to: 0.23-a0-899-gbbae0a335a (34)
22:09:39 <floraline> i think i'm able to reproduce my problem with console games rarely not showing up in the lobby, or being able to be spectated
22:10:36 <floraline> i'm getting some mysterious "Exception in i/o handler" when tornado tries to connect to the socket
22:12:22 <floraline> along with Errno 111 Connection refused
22:12:22 <floraline> does anyone have tips on how i can find out more about that exception? is there a tool i can run to catch it in the act?
22:12:36 <floraline> i don't know what my next steps are here
22:31:57 <gammafunk> floraline: you can run the python debugger, pdb, to break under relevent conditions, allowing you to inspect program state, but that may not work well for you
22:32:41 <gammafunk> you can also sort of add debugging output to relevant sections of the server if you know from where in webtiles the exception is generating
22:33:02 <gammafunk> but again you'd have to know what to print and have some idea what you're looking for
22:34:55 <floraline> i do know which line is generating the exception, but i'm not sure i would know what to look for
22:36:18 <alexjurkiewicz> a random feeling is that it's an error meaning "try again later" for that particular socket
22:36:38 <alexjurkiewicz> might be able to upload the python code to swallow the exception and continue, and things will just work when it next tries
22:37:21 <gammafunk> webtiles server is definitely do for an overhaul on the networking side
22:37:37 <gammafunk> even webtiles-changes didn't change this substantially
22:38:16 <gammafunk> it would be cool to port the codebase to a more recent tornado, but it's a lot of work that I certainly wouldn't have lots of time for in the near term
22:38:30 <gammafunk> we need edlothiol or a nu-edlothiol
22:38:49 <ebering> floraline is very close to florian
22:40:21 <floraline> i'm not familiar with that name :P
22:42:16 <ebering> florian wrote the initial webtiles prototype
22:56:58 <gammafunk> yeah their handle is edlothiol
22:57:08 <gammafunk> they did more then wrote the prototype really
22:57:16 <gammafunk> I think they basically wrote all of webtiles
22:57:29 <gammafunk> and they wrote the core parts of the webtiles-changes overhaul as well
22:58:27 <gammafunk> we've basically just been tweaking it over the years
22:58:49 <gammafunk> but they wrote the client, the server, and made lots of changes to both over the years as well
22:59:13 <gammafunk> but edlothiol moved on not too long after I became dev
22:59:34 <gammafunk> last commit in 2014
22:59:45 <gammafunk> heh
22:59:50 <gammafunk> one of their last commits:
22:59:59 <gammafunk> Webtiles: Do device pixel scaling better, fixing Firefox rendering.
22:59:59 <gammafunk> 
22:59:59 <gammafunk> Non-integer cell sizes are bad, which this amounted to with
22:59:59 <gammafunk> devicePixelRatios like 1.3333333730697632 in zoomed-in Firefox.
22:59:59 <gammafunk> 
23:00:02 <gammafunk> Gruuunt
23:12:00 <advil> I think I did work on something related to that commit
23:13:07 <advil> %git 4706f7efea59d
23:13:07 <Cheibriados> 07advil02 * 0.20-a0-1025-g4706f7e: Scale cell widths/heights by devicePixelRatio (wheals) 10(1 year, 9 months ago, 1 file, 6+ 2-) 13https://github.com/crawl/crawl/commit/4706f7efea59
23:46:50 <advil> l i n k i n g