00:05:01 Unstable branch on crawl.akrasiac.org updated to: 0.22-a0-769-g178e555 (34) 00:06:10 ??refactoring 00:06:10 I don't have a page labeled refactoring in my learndb. 00:06:13 ??refactor 00:06:13 I don't have a page labeled refactor in my learndb. 00:06:16 aw, rip 00:06:22 there used to be a learndb 00:10:14 god i cannot wait to merge this shit 00:10:17 Unstable branch on crawl.jorgrun.rocks updated to: 0.22-a0-772-gac08f70f71 (34) 00:16:02 if the changes in hellcrawl are mostly game play related, the conflicts shouldn't be all that numerous 00:20:20 New branch created: pull/820 (1 commit) 13https://github.com/crawl/crawl/pull/820 00:20:20 03alexjurkiewicz02 07https://github.com/crawl/crawl/pull/820 * 0.22-a0-758-g0a1ab5b: Add chequers_guarded_unrand_robe 10(3 minutes ago, 1 file, 43+ 0-) 13https://github.com/crawl/crawl/commit/0a1ab5bea376 00:21:43 >A randart robe is possible so that it's not trivial to identify the robe of misfortune. 00:21:46 i hate this 00:26:25 leaving it trivial to identify that there's a 50% chance it's the robe of misfortune 00:27:31 I think the removal and rework of various things will still make merging into hellcrawl non-trivial 00:27:34 but certainly still doable 00:30:57 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.22-a0-772-gac08f70 (34) 00:44:17 The build was broken. (master - ac08f70 #9936 : advil): https://travis-ci.org/crawl/crawl/builds/404721941 00:52:41 minmay: nothing's perfect, hopefully it's rare enough people forget 00:59:00 you need to give some of your monsters other equipment 00:59:19 meliai and probably blood saint 00:59:28 not sure what weapons the latter usually have 01:01:14 this also doesn't need no_item_gen nor patrolling 01:03:28 oh and it shouldn't really be placing blood saints at all, well I can get to making edits to that later 01:59:56 meliai by default have mundane robes or leather armour, I wanted to force robes 02:00:09 but i could make them non good_item if you want? 02:00:26 i thought blood saints were good to make the monsters scary, with call of chaos in the back 02:02:43 alexjurkiewicz: no I mean you have to give them weapons 02:02:55 if you give them equip like that, they'll only get that equip 02:03:36 as far as scary stuff goes, I generally would just stick with relatively branch-appropriate stuff and bloodsaints are extended monsters where you have plenty to choose from in dungeon/orc that are good choices 02:03:39 oh, annoying 02:03:57 hm, maybe titans then? airstrike is smite targeted 02:04:18 in orc I'd make that a box of orc sorc, orc high priests, deep elf mages/knights maybe 02:04:37 you're placing in D; it shouldn't use such out of branch stuff generally speaking 02:04:41 look at the dungeon ood list 02:04:44 !source pop_d 02:04:45 1/1. https://github.com/crawl/crawl/blob/master/crawl-ref/source/mon-pick-data.h#L1 02:06:06 I think just a box with robe wearing dungeon spellcasters is ok though; maybe rakshasa use robes? 02:06:06 those I believe are in ood 02:06:27 oh they're not even ood 02:06:41 yeah and robe users 02:07:16 also I would randomize the monster types more instead of placing fixed numbers in rows; the monsters will wake up and move around so placing them like that won't be effectively generally 02:08:00 it's sort of a relatively uninteresting box as far as the vault itself; nothing exactly wrong with that but if you'd like interesting groupings, it's probably better to try making more structure in the vault 02:08:10 I wanted a few monsters with smite spells, which is what the meliai and blood saints give 02:09:01 well blood saints are just not in-branch at all and there's no need for those; they're sort of close to tier 1 or 2 demons 02:09:02 huh. titan isn't even in D ood 02:09:20 you have a massive list of dungeon monsters to choose from so you really don't need to go to such extremes 02:09:49 and in orc pulling in meliai and other stuff I don't think is good for such a weak theme (players aren't going to recognize these as "robe monsters") 02:09:59 fair enough. I just figure runed door = extremely dangerous 02:10:26 sure, it's ok to be very dangerous 02:11:02 I guess I'd sort of prefer if the vault had some transparency for the monsters if you're trying to be fairly extreme since it's only dungeon, but we allow blind runed doors either way 02:11:39 but yeah maybe take a look at the D ood list and pull in some things that you like, just keep in mind players won't really see the "robe" connection too much 02:12:00 ok, i'll take another look 02:12:01 at most they'll just think a bunch of spellcasters, which is sort of the same thing 02:12:06 thanks for the ideas 02:12:38 np 02:14:37 btw, why don't you like patrolling? 02:15:02 with this no-los type vault, I figured it would be a good way to keep the vault from being so easily kited 02:24:41 patrolling doesn't prevent kiting, just makes it more annoying; it can help monsters stay near thematic terrain a bit instead of dispersing before player gets there and it can be alright if there's some kind of time pressure like in hells 02:25:03 but in this case the player opens the vault right there and the monsters themselves carry loot 02:26:05 minotaur needs patrolling so he doesn't get lost in the maze, stuff like that 02:38:00 can the minotaur find its way back to the vault if it gets teleported? 02:38:09 yeah 02:38:17 i've had that happen a time or two and it didn't seem like it could, but i might have just not waited long enough 02:38:27 well, monster pathfinding is not that great 02:38:54 well, they don't typically need great pathfinding either 02:39:00 I think though that once it loses foe memory of you it will pathfind ok when it wanders 02:39:08 ah 02:39:29 yeah let me test that actually 02:40:03 time ot use the ultimate hex, teleport other 02:40:59 yep, I teleported him and he wandered back into los after resting a few times 02:41:10 back into los of the stairs from where I teleported him 02:46:39 ah cool 02:46:50 i probably just didn't wait long enough 03:20:08 Long artefacts wrap text on small displays in the % screen 13https://crawl.develz.org/mantis/view.php?id=11533 by kitchen_ace 03:25:25 Crash when trying to use ` to enchant equipment with several scrolls 13https://crawl.develz.org/mantis/view.php?id=11534 by kitchen_ace 03:47:58 -!- amalloy is now known as amalloy_ 04:35:19 I've just noticed that in Makhleb's ability screen failure chance column, the word "hostile" is cut short to "host" 05:45:18 -!- eee_ is now known as fontfao 06:38:01 heh, that long artefact name text wrapping bug has been there since forever 06:39:15 NP7: unfortunately I don't have access to a windows machine to fix the glitch you reported 06:40:13 could you bisect it to the exact commit? it's very likely to be one of the earlier ones 06:41:04 it's likely the one that adds console resizing detection for windows: if so, I can likely revert it directly 06:48:07 aidanh: btw, I checked that monster description scrolling issue, and it seems I still can't scroll e.g. the desc of a lich except with mouse in tiles 06:48:28 I looked at your html scroller commit but wasn't sure if it's the same issue or not 06:48:56 the specific issue is that arrow/pagedown keys don't work for scrolling 06:49:17 if you do something like ?/M then "lich" 06:49:27 it might be the same in x-v 06:49:29 webtiles? yeah, unfortunately that will require a separate fix 06:49:40 yeah webtiles 06:50:38 I'll try to get to it soon, but I'm going to tackle the "evoke from inventory" bug first 06:52:03 and I'm not immediately sure if it'll be possible to synchronize scrolling between webtiles and console for those kinds of views, since they can contain arbitrary html 06:52:21 no worries at all, just wanted to make sure you knew since I didn't make a bug report 06:52:25 which maybe I should still 07:14:35 aidanh - I can certainly try to narrow it down, but it will take a while given how long crawl takes to compile 07:15:11 sure I'll check that commit first 07:19:10 %git 3f26980 07:19:10 07Aidan Holm02 * 0.22-a0-621-g3f26980: Add CK_RESIZE to console input handling 10(9 months ago, 4 files, 57+ 15-) 13https://github.com/crawl/crawl/commit/3f26980e4ead 07:19:15 this one I presume 07:23:35 NP7, yup; if not that one, then one of the commits that touches libw32c.cc 07:29:51 nope, it's fine in 3f26980 07:34:30 NP7: if you're not using git bisect, it will make finding the commit a lot faster 07:35:02 ah interesting 07:35:06 ccache will speed up recompilation considerably as well 08:07:44 of course I then get compile errors in particular builds 08:08:01 s/builds/versions 09:02:18 Faydane (L27 HuFi) ERROR in 'menu.cc' at line 197: ASSERT failed: index of 32 out of range 0 (0) .. (int)m_menu->items.size() (0) (Depths:3) 09:07:08 aidanh I've narrowed it down to three commits now, and it looks like changes to startup.cc in the main menu screen are the issue 09:07:23 by far the most likely commit to have caused the issue now is... 09:07:36 %git 6f189af 09:07:36 07Aidan Holm02 * 0.22-a0-700-g6f189af: Wrap the main menu in a popup widget 10(3 months ago, 3 files, 299+ 193-) 13https://github.com/crawl/crawl/commit/6f189af7ca97 09:07:51 !crashlog 09:07:53 19803. Faydane, XL27 HuFi, T:106790 (milestone): http://crawl.akrasiac.org/rawdata/Faydane/crash-Faydane-20180717-130217.txt 09:08:05 I'll complete the final 2 iterations of git bisect for completeness 09:13:04 Indeed 6f189af is the first bad commit 09:24:56 On Makhleb's ability screen in the failure column, the word "hostile" is cut short to "host". 13https://crawl.develz.org/mantis/view.php?id=11535 by NormalPerson7 09:32:49 uh oh 09:32:55 R I P Sequell 09:49:10 NP7: hrm, ok, that's more troubling; that indicates the UI framework itself isn't working on windows console 09:52:10 NP7: one thing you could try is removing the calls to set_getch_returns_resizes() in ui.cc 09:52:44 if that doesn't work, I may be able to obtain access to a windows environment in a few days 10:00:32 -!- muffindrake1 is now known as muffindrake 10:04:23 -!- NP7 is now known as NormalPerson7 10:05:20 -!- NormalPerson7 is now known as NP7 10:06:22 well aidanh I can tell you that removing the calls to set_getch_returns_resizes actually works 10:08:10 cleeve (L22 GnFi) ERROR in 'menu.cc' at line 197: ASSERT failed: index of 51 out of range 0 (0) .. (int)m_menu->items.size() (0) (Lair:1) 10:09:43 :) great! unfortunately committing that as-is will break resizing for non-windows terms, so I'll undo the libw32c.cc changes in 3f26980 instead 10:11:35 and it immediately reveals another far less serious graphical error, as in this image https://imgur.com/a/mV2V5wq 10:16:26 !crashlog 10:16:27 19804. cleeve, XL22 GnFi, T:57916 (milestone): http://crawl.akrasiac.org/rawdata/cleeve/crash-cleeve-20180717-140808.txt 10:17:29 New branch created: fix-windows-console-glitch (1 commit) 13https://github.com/crawl/crawl/tree/fix-windows-console-glitch 10:17:29 03Aidan Holm02 07[fix-windows-console-glitch] * 0.22-a0-766-g98cd069: Disable resize detection on windows console (NormalPerson7, #11532) 10(2 minutes ago, 1 file, 2+ 8-) 13https://github.com/crawl/crawl/commit/98cd0694a067 10:18:24 NP7: could you please revert the changes to ui.cc, and then test that commit that I just pushed to the fix-windows-console-glitch branch? 10:19:14 sure 10:19:19 thanks 10:19:57 separately, does the graphical error you just posted a screenshot of start with commit 6f189af? 10:20:22 I'd have to test that separately 10:22:10 98cd069 works 10:24:10 03Aidan Holm02 07* 0.22-a0-773-g9746042: Disable resize detection on windows console (NormalPerson7, #11532) 10(9 minutes ago, 1 file, 2+ 8-) 13https://github.com/crawl/crawl/commit/9746042cceb3 10:24:24 thanks, merged it 10:25:05 thanks to you too, now I can test things again without having to work out how to compile tiles! 10:31:50 :) 10:35:57 The build failed. (fix-windows-console-glitch - 98cd069 #9938 : Aidan Holm): https://travis-ci.org/crawl/crawl/builds/404925807 10:38:43 tiles failed? 10:43:31 github is timing out 10:46:33 compiling one commit before 6f189af to test the graphical bug now 10:56:29 Game crashed while going up the stairs to swamp 1 13https://crawl.develz.org/mantis/view.php?id=11536 by Cakaith 10:56:53 The build failed. (fix-windows-console-glitch - 98cd069 #9938 : Aidan Holm): https://travis-ci.org/crawl/crawl/builds/404925807 10:59:03 oh right, that's definitely going to fail because I already deleted that branch 10:59:20 oh haha 10:59:46 it's fine before 6f189af 11:01:06 thanks, I'll take a look at it sometime in the next few days 11:03:09 aidanh but it's also fine in 6f189af 11:03:15 so 6f189af didn't break it 11:03:20 I guess I can do another bisect 11:04:30 that'd be much appreciated 11:10:17 Unstable branch on crawl.jorgrun.rocks updated to: 0.22-a0-773-g9746042cce (34) 11:32:12 The build was fixed. (master - 9746042 #9939 : Aidan Holm): https://travis-ci.org/crawl/crawl/builds/404928726 11:41:30 PKrockin (L18 DsFE) ERROR in 'menu.cc' at line 197: ASSERT failed: index of 24 out of range 0 (0) .. (int)m_menu->items.size() (0) (Elf:1) 11:43:30 -!- amalloy_ is now known as amalloy 12:41:08 %git e1e3f6e 12:41:09 07Aidan Holm02 * 0.22-a0-728-ge1e3f6e: Remove BlackWhiteHighlighter 10(8 weeks ago, 4 files, 0+ 86-) 13https://github.com/crawl/crawl/commit/e1e3f6e481f0 12:41:30 boxmenuhighlighter definitely doesn't do the same thing aidanh :) 12:41:42 this is the problematic commit 12:42:19 d'oh! 12:42:56 in fairness, it does do the exact same thing on unix consoles 12:43:04 !blame3 windows 12:43:04 wiiiiindooooows 12:43:12 welp 12:44:21 I blame unix :p 12:48:56 03Aidan Holm02 07* 0.22-a0-774-g12c0d5b: Fix crash when scrolling+switching in webtiles menus 10(15 minutes ago, 1 file, 4+ 0-) 13https://github.com/crawl/crawl/commit/12c0d5b0f5f4 12:48:56 03Aidan Holm02 07* 0.22-a0-775-ga7d8001: Revert "Remove BlackWhiteHighlighter" (NormalPerson7) 10(2 minutes ago, 4 files, 86+ 0-) 13https://github.com/crawl/crawl/commit/a7d80019b58b 12:50:11 thumbs up 12:50:34 all working smoothly? 12:57:37 fwiw in Cakaith's crash, the unaccounted-for monster seems to be...i'm not 100% sure on this, but i think a glowing shapeshifter who briefly became an orc, converted to follow the player under beogh, and then turned into a fire bat? 12:58:25 don't know why this would cause the monster not to be in the mid cache 13:00:01 amalloy: yes; the sequence of events seems to be: 13:00:03 1. he fought an apocalypse crab around turn 34000 in swamp 1. morguth got converted to a shapeshift (along with another named follower). 13:00:31 2. went back to lair, recalled followers. Morguth recalled as a regular orc, though the other named follower recalled as a shapeshifter 13:01:03 I am guessing that much later he returned to swamp 1, and recalled again? 13:01:24 at some point morguth the shapeshifter lost its god companion flag, too 13:02:17 so far I just don't know why morguth recalled as an orc in step 2, I can't replicate it, but it probably involves some complicated sequence of level changes and recalls? 13:02:20 aidanh - all fine now, yep 13:03:09 amalloy: if you want to watch, `!lm Cakaith -tv:>:channel=debug` shows those two events 13:03:36 (morguth was a regular orc before the apocalypse crab) 13:04:40 at the point in the save, there are two monsters with mid 308 on swamp 1, one is the shapeshifter, one is a regular instance of morguth the orc 13:05:58 &versions 13:06:18 CAO: 0.22-a0-769-g178e555, CBRO: 0.22-a0-614-g34fbaa0, CDO: 0.22-a0-757-gdba82d8, CJR: 0.22-a0-773-g9746042cce, CPO: 0.22-a0-773-g9746042, CUE: 0.22-a0-769-g178e55541d, CWZ: 0.22-a0-422-g9628fc8, CXC: 0.22-a0-772-gac08f70, LLD: 0.22-a0-621-g3f26980e4e 13:07:35 I take it trunk rebuild is still broken on cbro 13:07:44 I wonder if johnstein actually knows that? 13:10:18 Unstable branch on crawl.jorgrun.rocks updated to: 0.22-a0-775-ga7d80019b5 (34) 13:21:05 Unstable branch on crawl.akrasiac.org updated to: 0.22-a0-775-ga7d8001 (34) 14:01:20 !tell gammafunk thanks for the tvs but I prefer sequell tells to discord mentions 14:01:21 ebering: OK, I'll let gammafunk know. 14:12:00 my great hope is that gammafunk replies to clarify that he prefers to receive discord mentions 14:14:17 hisssss 14:14:41 also I am back from an interval of no internet 14:14:45 when's feature freeze 14:29:35 03ebering02 07https://github.com/crawl/crawl/pull/809 * 0.22-a0-609-gffe50d3: fixup! Document the rest of clua in ldoc 10(8 days ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/ffe50d39700f 14:56:01 03ebering02 07https://github.com/crawl/crawl/pull/809 * 0.22-a0-776-gd3a23a3: Switch from LuaDoc to LDoc 10(12 days ago, 5 files, 55+ 769-) 13https://github.com/crawl/crawl/commit/d3a23a32c6a9 14:56:01 03ebering02 07https://github.com/crawl/crawl/pull/809 * 0.22-a0-777-g713e0dc: De-Doxygen-ify 10(12 days ago, 18 files, 2+ 92-) 13https://github.com/crawl/crawl/commit/713e0dc92d64 14:56:01 03ebering02 07https://github.com/crawl/crawl/pull/809 * 0.22-a0-778-g8fcbca3: Switch over and expand old luadoc 10(12 days ago, 2 files, 432+ 146-) 13https://github.com/crawl/crawl/commit/8fcbca3b6fc4 14:56:01 03ebering02 07https://github.com/crawl/crawl/pull/809 * 0.22-a0-779-g4deae7e: Document dlua additions to the crawl module 10(12 days ago, 1 file, 49+ 1-) 13https://github.com/crawl/crawl/commit/4deae7ee78c3 14:56:01 03ebering02 07https://github.com/crawl/crawl/pull/809 * 0.22-a0-780-gccc63dc: Stubs for dlua only modules. 10(12 days ago, 21 files, 57+ 18-) 13https://github.com/crawl/crawl/commit/ccc63dc1df12 14:56:01 03ebering02 07https://github.com/crawl/crawl/pull/809 * 0.22-a0-781-g988edb2: Document the rest of clua in ldoc 10(12 days ago, 16 files, 1676+ 106-) 13https://github.com/crawl/crawl/commit/988edb20955c 14:56:01 03ebering02 07https://github.com/crawl/crawl/pull/809 * 0.22-a0-782-gd751f66: Sort lua api table of contents. 10(11 days ago, 1 file, 6+ 5-) 13https://github.com/crawl/crawl/commit/d751f66a7fc4 14:56:01 03ebering02 07https://github.com/crawl/crawl/pull/809 * 0.22-a0-783-g2e8abf5: Clean up internal comments (aidanh) 10(10 minutes ago, 11 files, 12+ 54-) 13https://github.com/crawl/crawl/commit/2e8abf55ea30 15:24:58 zluc (L18 GnWz) ERROR in 'menu.cc' at line 197: ASSERT failed: index of 19 out of range 0 (0) .. (int)m_menu->items.size() (0) (Snake:4) 16:45:50 eyesofgods1 (L27 DsFi) ERROR in 'menu.cc' at line 197: ASSERT failed: index of 3 out of range 0 (0) .. (int)m_menu->items.size() (0) (Orc:2) 16:46:50 eyesofgods1 (L27 DsFi) ERROR in 'menu.cc' at line 197: ASSERT failed: index of 3 out of range 0 (0) .. (int)m_menu->items.size() (0) (Orc:2) 16:59:03 Damn it, do I really have to restart my webtiles instance to update my banner.html? :/ 17:01:37 Bleh, I'll restart it in the morning 17:01:40 :/ 18:02:13 I think I've finally replicated the mid cache bogosity thing 18:02:57 if a shapeshifter follower changes shape while you are taking the stairs and it's not following, you get a clone 18:07:48 also, the shapeshifter follower needs to have moved into the level by following you, not by being recalled 18:08:08 the first level, that is 18:22:01 Unstable branch on underhound.eu updated to: 0.22-a0-775-ga7d80019b5 (34) 18:22:35 interesting 18:22:59 I think I had a crash of that kind in a HO^Beogh game the made it unwinnable, and we had to edit the save so I could progress 18:23:17 I had a follower who had a clone who got turned into an AST 18:23:28 maybe no shapeshifter was involved and it was unrelated though 18:23:56 but I couldn't return to an earlier level of vaults without crashing the game; had to kill the clone 18:24:21 after warping to the level in wiz mode 18:24:43 probably this isn't the only way to get a code 18:24:49 was it a mid cache bogosity error? 18:24:54 s/code/clone 18:24:55 hrm, wonder if I can find that out 18:24:59 !crash . ho 18:25:06 !crashlog . ho 18:25:08 13. gammafunk, XL21 HOFE, T:76738 (milestone): http://dobrazupa.org/morgue/gammafunk/crash-gammafunk-20150208-172120.txt 18:25:08 13. gammafunk, XL21 HOFE, T:76738 (milestone): http://dobrazupa.org/morgue/gammafunk/crash-gammafunk-20150208-172120.txt 18:25:15 ah 18:25:16 that's it 18:25:27 and yeah it was mid cache bogosity 18:27:00 looks similar 18:27:36 there's definitely other cleanup code around for some clone issues though, so it all depends on how the clone came about 18:28:21 I still haven't really figured out the mechanism here 18:30:42 yeah and I wasn't really clear how my clone came about, maybe it was a chaos cloud but I couldn't be sure 18:31:04 that's probably one of the more likely culprits since chaos spawns are common as summons 18:31:17 either af_chaos or a chaos cloud 18:31:45 oh, they probably don't leave chaos clouds if they're summons though 18:59:17 New branch created: pull/821 (2 commits) 13https://github.com/crawl/crawl/pull/821 18:59:17 03Pekkekke02 {GitHub} 07https://github.com/crawl/crawl/pull/821 * 0.22-a0-766-g1fba96a: Remove references to 'p'raying at altars 10(30 minutes ago, 1 file, 4+ 4-) 13https://github.com/crawl/crawl/commit/1fba96af3c19 18:59:17 03Pekkekke02 {GitHub} 07https://github.com/crawl/crawl/pull/821 * 0.22-a0-767-g2fdb696: Remove reference to 'p'raying 10(6 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/2fdb6964861b 19:28:53 just wanted to give props to whoever contributed to the ui update, and whoever is bugfixing 19:29:21 ui update is sweet, and bugfixes are comin in hot 19:39:33 !cheers jboons 19:39:33 * Sequell slides a cask of mead across the bar to jboons, courtesy of gammafunk. 19:41:03 can i ask here about if a bug that i've noticed is already acknowledged? 19:41:17 or is my best bet to just submit a bug on mantis regardless 19:44:36 ask away 19:45:48 on item list screens, if you only use pg-up and pg-down to navigate, certain line items are basically cut out due to the gradient effect on the top and bottom of these windows 19:46:56 or it seems that way anyway 19:47:14 the last item is difficult to read and the whole point of those popouts is to read every item 19:47:44 i appreciate a good gradient but these seem to take away from it 19:53:56 that's a fair observation: I'll look into either reducing the effect or adjusting page up/down to take it into account 19:55:11 i like the second option but it might not be worth the effort 19:56:36 I'm absolutely sure that the second option will lead to bug reports about page up/down scrolling slightly less than a full page 19:57:31 yeah i think you're right 19:58:06 and i said item list screens but it's any screen that follows the same style 19:59:23 libraries books and shops are others i can think of 19:59:31 i think shops anyway, but def libs and books 20:00:48 also the workaround to the problem is to just scroll or use up/down arrow 20:00:54 for me anyway 20:01:13 so not critical! 20:12:56 -!- Bammboobies is now known as Bammboo 20:26:32 yes, I was noticing issues with the gradient 20:26:55 it was showing up incorrectly when using ctrl-f in memorize, for example, but it wasn't obvious what the exact conditions were 20:30:19 augh my recipe for producing extra followers stopped working 20:30:28 why did it work several times in a row 20:38:08 followers on what 20:40:08 orcs 20:46:53 -!- _krator44 is now known as krator44 20:49:35 hm, I wonder if this is one of those ones where printing all these diagnostic messages is actually affecting the problem 20:56:24 <|amethyst> advil: is this related to the perennial orc follower mcache bug? 20:56:38 it's at least one of them 21:02:16 why do we die() on a debug scan that the mcache is broken, instead of rebuilding the mcache? 21:05:30 is it really a cache that could be rebuilt, or does it have un-reconstructable information? 21:13:08 I think at least some cases it's not clear what to do, but yeah, this is a recipe for broken saves 21:13:24 this particular crash is when there is a monster in the level whose mid is mapped to 0 21:13:29 well to nullptr 21:14:09 one way it arises is from two monsters with the same mid on the same level, one of them dying on that level 21:14:22 (I think there's other ways because I don't think that's exactly how the crash came about) 21:14:37 so in general I'm not sure what the right thing to do is for this sort of messed up mid cache 21:15:17 I'll probably write a fix for this case that just deletes any friendly non-god-aligned bogus monsters, since that picks out this bug relatively specifically 21:15:33 also, I was *completely* wrong about the circumstances, lol 21:15:42 it is triggered by changing into a meliai 21:16:05 (so unfortunately, this can't really be a very old instance of an orc clone bug) 21:26:24 <|amethyst> In general, there are lots of places where we ASSERT() even though we could just fix the problem 21:27:00 <|amethyst> usually that's not so bad, because we'd prefer to get a bug report 21:27:25 <|amethyst> it's a problem in cases like this when it breaks the save 21:27:30 <|amethyst> advil: why specifically a meliai? 21:32:57 <|amethyst> (re breaking the save, I think there are at least some cases where we crash, then silently or not-so-silently fix the bug when loading the save) 21:50:30 |amethyst it's because meliai are marked as priests 21:51:15 (where is that set?) 21:51:31 advil: mon-data.h, there's a priest flag I believe? 21:52:03 interesting, they don't have it 21:52:04 doesn't seem to be -- it's implemented as a prop 21:52:15 ("priest") 21:52:21 but I'm not seeing where that gets set 21:52:22 oh 21:52:26 It's the spells 21:52:29 they have priestly spells 21:52:34 that gives them the priest flag 21:52:38 ah 21:53:27 also, how many people know that bushes are half-opaque? 21:53:33 in that two of them in a row block LOS 21:53:50 seems odd to me, since a single plant blocks line of fire, but never blocks los 21:54:03 bushes allow line of fire but block los with two 21:54:26 I guess it's a thing you could argue makes sense, but I'm sort of wondering why it's a thing 21:55:12 <|amethyst> gammafunk: because no one ever changed that when they changed trees? 21:56:25 <|amethyst> gammafunk: I guess a bush is supposed to be bigger than a plant, which raises the question of why plants but not bushes block lof 21:57:05 <|amethyst> does anything half-opaque remain other than bushes and (some) clouds? 21:57:25 statues? 21:57:28 not statues 21:57:33 <|amethyst> statues never block LOS 21:58:05 new monster: half-opaque golem 21:58:20 <|amethyst> tree mimic 22:09:38 03ebering02 07https://github.com/crawl/crawl/pull/764 * 0.22-a0-776-g83b5ea7: Recolour awoken forests in console. 10(8 weeks ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/83b5ea78ea2f 22:15:09 03advil02 07* 0.22-a0-776-ge6d7efa: Don't overwrite god with GOD_NAMELESS on poly to a priest (11536) 10(11 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/e6d7efa92cb0 22:16:15 I think that's my new favorite bugfix 22:21:54 that commit doesn't fix the save, but I'll do that tomorrow 22:47:05 gammafunk: I've tracked down what's half baked about the red trees patch 22:47:37 random_element_colour expects that the list be weighted pairs with weights totalling 120. The extra weight is taken by the default colour which is BLACK, but this gets turned to GREEN for glyphs. 22:47:58 The only two element colours that get this wrong are awoken trees and wu jian 22:48:06 though it's unclear if wu jian was intentional 22:49:02 everyone remembers green fire giants fondly, manifestation of a similar issue before I redid how monster colors were stored 22:49:51 doubt wu jian was intentional, but not sure what was said in the commit adding that colour 22:52:53 the commit was the original SN commit 22:54:32 yeah, but I'm talking about e.g. the commit message and whether anything was said there 22:55:02 if not, it was probably just chosen and SN wasn't aware of how that weighting work 22:55:47 ya, nothing in the commit message, only the one comment mentioning the colours chosen to match the theme 22:56:12 not sure if those colors match the tile 22:56:54 I see yellow and red in the tile 22:57:01 sadly this doesn't look like an easy place to use choose_random_weighted 22:57:39 yeah yellow, red, and white in the etc color, which matches the tile well, so I'd say no green was intentional 22:59:39 so guess you can just do 53, 53, 14 in the existing def to fix wu jian's 23:00:10 or just 50, 50, 20 would be fine as well 23:00:45 fr: ETC_MOON 23:02:11 ok, red trees is ready to go 23:02:27 I guess I'll add on to that pr a wu jian fix 23:02:53 03ebering02 07https://github.com/crawl/crawl/pull/764 * 0.22-a0-776-gdea5ac9: Recolour awoken forests in console. 10(8 weeks ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/dea5ac9b038a 23:02:57 and a comment describing the unexpected choice of total weight 23:03:18 I'll sponsor the debat of that PR in the crawl senate 23:03:57 But be warned that if Senator amalloy deploys his Recite filibuster strategy, there's little I can do to help 23:04:15 *debate 23:04:21 !recite 23:04:44 hm. todo make that work right 23:10:19 Unstable branch on crawl.jorgrun.rocks updated to: 0.22-a0-776-ge6d7efa92c (34) 23:10:32 it's not super clear where to put the comment 23:11:22 new plan is to re-factor random_element_colour to not use a hard-coded weight... tomorrow 23:13:27 03ebering02 07https://github.com/crawl/crawl/pull/764 * 0.22-a0-777-g217ac37: Fix Wu Jian elemental colour. 10(3 minutes ago, 1 file, 3+ 3-) 13https://github.com/crawl/crawl/commit/217ac37344c9