01:34:40 -!- Tiobot is now known as Guest27494 02:48:26 Monster database of master branch on crawl.develz.org updated to: 0.23-a0-650-ga4d50d4 03:50:49 Experimental (bcrawl) branch on crawl.kelbi.org updated to: 0.23-a0-1347-g430c86cbe3 05:11:58 -!- Evablue_ is now known as Evablue 06:18:18 -!- Evablue_ is now known as Evablue 06:25:03 Sometimes my lightning spire attacks out of los. "A bolt of electricity appears from out of your range of vision." 06:36:38 Sometimes my Lightning Spire attacks from out of my los. 13https://crawl.develz.org/mantis/view.php?id=11818 by Yermak 09:59:00 has anyone been able to replicate this stale sprite bug? I've seen a lot of reports of it lately and tried to replicate it multiple times, but no luck 10:17:37 not reliably 10:55:44 Sleeping monsters don't wake up on taking damage from trees (Summon Forest). 13https://crawl.develz.org/mantis/view.php?id=11819 by Yermak 11:36:28 advil: I first started encountering it on Dec 6 on CKO (https://crawl.develz.org/mantis/view.php?id=11780). Then had no issue for a few weeks. 11:36:59 oh thanks, I'd missed that mantis issue 11:37:15 I think there might be multiple issues 11:37:15 Now, it seems to be mostly related to range weapons leaving icons that should have been erased. 11:37:28 agreed 11:37:39 there's something to do with explore delays, but there's something else that isn't 11:38:26 Would be interesting to find out if it is not happening on some servers. 11:40:09 seems to be related to lag that isn't excessive enough to stop playing 11:40:34 maybe, the detailed report I just got on reddit suggests it might be independent of lag 11:40:50 I was also using an xcode tool to simulate lag in some of my attempts to replicate it 11:41:54 well, I think this report has some clue in it because I replicated it instantly with their rc settings 11:48:17 great 11:53:02 this is so easy to replicate idk what I was doing before 11:54:36 ah, not using chrome would be one thing I was doing before 11:55:17 this thing is wild...the bug doesn't show on firefox, but if you spectate a char running on firefox, spectating chrome, it is visible to the spectator 12:38:55 I was getting it with a local instance. no ranged. thinks like 'o' would leave a sprite behind, even if not otherwse visbly rendered (e.g. either me or e.g. Hep ancestor would sometimes have a sprite at the edge of LOS when explore stopped) 12:39:23 I actually have not yet been able to replicate it locally 12:39:28 were you doing anything special? 12:39:48 no. I was using chrome 12:40:18 don 't think I have much in the way of special settings for tiles (some customizations for console gyphs which I ghtink shouldn't affeect webtiles at all?) 12:41:14 show_player_species = true 12:41:28 https://gist.github.com/geekosaur/055356a81ba3abd9cf54e1023f40453a 12:47:04 what build options? 12:51:53 LTO=y, otherwise defautls at that point (was originally using clang on linux but switched away to no effect) 12:52:22 I can'tt look more closely, that machine is currently inaccessible and this machine has a different build w/different optionss 13:18:14 -!- amalloy is now known as amalloy_ 13:19:26 I think for whatever reason it's not showing up on debug builds 13:19:54 I just replicated it locally on a non-debug build 13:20:08 I can't tell. I thogutth I'd been using a debug build on the other machine 13:20:15 debug-lite 13:55:58 I can't get it with debug-lite either 14:01:53 03advil02 07[stone_soup-0.22] * 0.22.1-3-g569bc69: Use contribs if appropriate when building rltiles/ via sub-make 10(4 weeks ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/569bc69aa185 14:01:53 03advil02 07[stone_soup-0.22] * 0.22.1-4-g520d174: Fix cross-compiling + libpng 10(3 weeks ago, 1 file, 6+ 1-) 13https://github.com/crawl/crawl/commit/520d174dca3e 14:01:53 03advil02 07[stone_soup-0.22] * 0.22.1-5-g62c59d9: Fix some issues with build-rltiles and webtiles 10(2 weeks ago, 1 file, 4+ 1-) 13https://github.com/crawl/crawl/commit/62c59d919a6e 14:41:00 well, now I 14:41:12 'm really confused, because I've replicated this bug in 0.21 14:41:34 it does seem rarer in 0.21 than in trunk 14:44:13 wonder if this results from a chrome change, rather than a dcss change 16:07:45 advil: is it that visual artifacts bug? 16:08:27 if so, yeah, looking at the timing of things and how the reporters all seemed to have chrome and nothing had changed in webtiles display 16:14:23 yeah 16:14:51 there are some misc console errors (independent of whether this bug shows) in chrome that I don't think happened before either 16:18:03 thanks for looking into that 17:12:17 I still have no idea what's going on really, but it appears that chrome is, unpredictably, receiving messages out of order 17:24:01 I can actually run the player on firefox and a spectator on chrome and see different message orders for travel 17:24:28 (the sometimes-errors at startup are also due to message ordering issues) 17:42:47 hrm, weird 17:55:53 yeah...I can see that the raw websocket messages are in the right order, but socket.onmessage seems to be getting called in the wrong order?? 18:11:08 advil: is this something happening via that comm.js library we use to shuttle messages to the right javascript handlers? 18:11:35 it's been a while since I've looked at that stuff, but I recall that lib was responsible for packaging up the websock json messages and calling registered handlers 18:12:05 but what you're describing sounds like it's taking place before that, which is weird 18:12:19 I *think* what I'm looking at is at the browser level 18:14:23 also, I think I misunderstood how onmessage is called, I believe it is called in message order but also in parallel 18:15:10 onmessage of a given WebSocket is called on a single thread only 18:26:39 advil, decode_utf8() in https://github.com/crawl/crawl/blob/master/crawl-ref/source/webserver/static/scripts/client.js looks suspicious 18:27:41 advil, try using the synchronous https://developer.mozilla.org/en-US/docs/Web/API/TextDecoder instead 18:30:55 or if you can't use that for browser compatibility reasons, consider some pure-JS implementation of a UTF-8 decoder 18:32:42 oh, yeah, that does look suspicious 18:54:45 03gammafunk02 07* 0.23-a0-651-ge027d9b: Initial rework of Killer Klowns 10(2 hours ago, 5 files, 44+ 23-) 13https://github.com/crawl/crawl/commit/e027d9b72c0f 18:54:45 03gammafunk02 07* 0.23-a0-652-gbb96f5f: Code support for restricted player polymorph 10(2 hours ago, 5 files, 19+ 13-) 13https://github.com/crawl/crawl/commit/bb96f5f2485b 18:54:45 03gammafunk02 07* 0.23-a0-653-gca735f5: A Throw Klown Pie ability for Killer Klowns 10(2 hours ago, 20 files, 206+ 2-) 13https://github.com/crawl/crawl/commit/ca735f56021e 18:58:35 !kw newklowns vlong>=0.23-a0-653-gca735f5 18:58:36 Defined keyword: newklowns => vlong>=0.23-a0-653-gca735f5 18:58:42 monster jesters? :p 19:00:15 we stopped short of making them worship nem and have xom wrath 19:00:18 ....for now 19:16:22 do we support IE? 19:22:24 Unstable branch on crawl.kelbi.org updated to: 0.23-a0-653-gca735f5602 (34) 19:30:55 yeah, so the FileReader callbacks that jilles pointed out run in a different order than they are queued up, for some reason only on chrome. But I don't think any guarantees are made about order for FileReader, so we probably shouldn't be relying on it 19:32:30 I'm not sure we've ever declared anywhere just what browsers we support 19:32:55 it would probably be ok to use TextDecoder except that then we need to get the output of inflate.js into Uint8Array which is kind of annoying to do (though it can be done easily with a FileReader!) 19:57:10 oh, no, I was just misunderstanding something, that wasn't too bad to get working 21:15:31 03advil02 07* 0.23-a0-654-g63bca24: Fix out-of-order messages (maybe 11780) 10(13 minutes ago, 1 file, 23+ 9-) 13https://github.com/crawl/crawl/commit/63bca24c0784 21:16:59 hm I probably shouldn't have used let 21:17:05 do we ever use that? 21:26:38 Unstable branch on crawl.kelbi.org updated to: 0.23-a0-654-g63bca24c07 (34) 21:27:03 advil: not sure, but would be good to backport that fix to 0.22 once you're happy with it 21:27:30 well, because it's in the static js for most servers it should apply automatically 21:27:47 I think? 21:28:19 right, I forget how that goes since it was a while sinc we last discussed it 21:28:31 and didn't it get changed somewhat recently, what the install does? 21:30:57 for cko it looks like it took effect for trunk and 0.22 immediately 21:33:19 03advil02 07* 0.23-a0-655-g3751e1c: Don't use `let` 10(40 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/3751e1c4cc42 21:34:29 because it's a webtiles static file it was copied on trunk update and will be used for all games 21:35:34 that applies for any server using the traditional dgl setup 21:36:23 cool 21:36:36 still to go cherry-pick to 0.22, but we can do that closer to release 21:36:43 s/to go/good to/ 21:37:17 i'm sure that the more customized servers are doing something similar, i can't imagine anyone is keeping multiple versions of the webtiles server 21:37:58 well, the other servers have to manually merge trunk whenever they can/feel like it 21:38:02 since they apply local patches 21:38:15 and then there's cpo. which is some weird version of webtiles-changes 21:39:05 so I think it may also have to manually merge that stuff as it has different lobby code and all that, which is stuff in static 21:39:58 cwz doesn't actually ever fully install stable versions; it just installs the beta before the release or possibly just the .0 release if said beta is not made 21:40:10 !lg * cwz 0.22 x=vlong 21:40:13 No games for * (cwz 0.22). 21:40:18 is there a reason for that? 21:40:27 !lg * cwz vlong~~0\.22 21:40:30 210033. jekm0323 the Vexing (L1 OgAM), slain by a kobold on D:1 on 2019-01-07 01:57:58, with 1 point after 32 turns and 0:02:26. 21:40:34 !lg * cwz vlong~~0\.22 x=vlong 21:40:35 210033. [vlong=0.22-b1-59-g0aedee7] jekm0323 the Vexing (L1 OgAM), slain by a kobold on D:1 on 2019-01-07 01:57:58, with 1 point after 32 turns and 0:02:26. 21:40:56 just that hong doesn't want to mess with managing stable versions 21:41:08 especially since he has to merge local patches for korean language things, I think 21:41:14 we don't really know what his process is 21:41:33 I think lld might have up to date 0.22 21:41:37 !lg * 0.22 lld x=vlong 21:41:39 2621. [vlong=0.22.1-2-g9c5b331383] WanderingMacrophage the Chiller (L6 OpIE of Kikubaaqudgha), slain by a centaur on D:4 on 2019-01-06 15:10:03, with 359 points after 6223 turns and 0:26:58. 21:41:50 %git stone_soup-0.22 21:41:50 07advil02 * 0.22.1-5-g62c59d9: Fix some issues with build-rltiles and webtiles 10(2 weeks ago, 1 file, 4+ 1-) 13https://github.com/crawl/crawl/commit/62c59d919a6e 21:42:06 so he's updating that periodically, at least 21:42:16 he being dplusplus, the lld admin 21:42:29 %git 9c5b331383 21:42:29 07advil02 * 0.22.1-2-g9c5b331: Revert complex bones install stuff, add dist_bones to repo 10(4 months ago, 36 files, 21+ 28-) 13https://github.com/crawl/crawl/commit/9c5b33138319 21:43:06 looks like maybe no patched applied on top of that; maybe lld doesn't do the custom patch thing directly in the branch like I think cwz does 21:43:14 !lg lld current alpha x=vlong 21:43:15 No games for lld (current alpha). 21:43:20 !lg * lld current alpha x=vlong 21:43:21 3307. [vlong=0.23-a0-650-ga4d50d4d25] mirui the Warrior (L17 MiFi of Hepliaklqana), blasted by a deep elf knight (shard of ice) on Elf:2 (elven_hall_of_blades_hangedman_original) on 2019-01-07 02:03:44, with 191166 points after 38033 turns and 4:02:39. 21:43:28 %git a4d50d4d25 21:43:28 07gammafunk02 * 0.23-a0-650-ga4d50d4: A Trog-themed ghost vault with wrathtful warriors 10(6 days ago, 1 file, 228+ 0-) 13https://github.com/crawl/crawl/commit/a4d50d4d25f0 21:44:05 yeah, so I think he's applying some things just not in the repo in a way that affects the version 21:44:20 whereas I believe cwz actually changes the repo, or at least it's definitely done so in the past 21:46:55 Unstable branch on crawl.kelbi.org updated to: 0.23-a0-655-g3751e1c4cc (34) 22:26:22 -!- muffindrake1 is now known as muffindrake