00:25:59 what was the proposal? 00:29:18 03gammafunk02 07* 0.25-a0-284-g26646b4: Clean up some DES syntax for a vault 10(16 minutes ago, 1 file, 38+ 39-) 13https://github.com/crawl/crawl/commit/26646b4e49d4 00:29:18 03gammafunk02 07* 0.25-a0-285-gcba6f66: A swamp worm tunnel vault (mikee) 10(15 minutes ago, 1 file, 57+ 0-) 13https://github.com/crawl/crawl/commit/cba6f664f805 00:32:25 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-285-gcba6f664f8 (34) 00:38:47 gammafunk: can't take stairs if you are damaged during the delay 00:39:26 and an alternate approach based on Brogue (which I haven't implemeted yet) 01:37:44 Unstable branch on crawl.develz.org updated to: 0.25-a0-285-gcba6f66 (34) 02:08:45 Windows builds of master branch on crawl.develz.org updated to: 0.25-a0-285-gcba6f66 02:59:20 Monster database of master branch on crawl.develz.org updated to: 0.24-a0-443-g80245de 03:12:18 Unstable branch on crawl.beRotato.org updated to: 0.25-a0-285-gcba6f66 (34) 03:31:02 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-2777-g6b872f2bae 04:45:56 is it possible to have a player tile that animates in the same way altars do? 06:33:22 -!- amalloy is now known as amalloy_ 07:49:01 Hi just a heads up b/c I don't know where to report there is a player on CAO at the moment with the n-word in their name 08:27:37 Minor superfluous code: The thrown effect for MI_THROWING_NET is listed twice in tilepick.cc; the second listing does nothing (it will have returned a value from the first listing before reaching it. lines 2794 (first listing) and 2826 (second listing); the second one can (again) be removed safely with no actual effect. 09:40:19 server admin question: what actually is the best way to disable a dgl account? 09:40:19 in the past I've just changed the password to line noise 09:41:25 (I'm going to ban the player that ratatosk mentions) 10:35:24 advil, there was a discussion about this in a pull request 1034: https://github.com/crawl/crawl/pull/1034 10:35:32 Rawlins at least seems to do the same thing 10:52:04 pblur: Rawlins = me ;-) 10:52:20 Oh. lolol. 10:52:37 Well, turns out you do the same thing as yourself I guess. :P 10:52:43 probably a good sign 10:53:04 so the question is really, is there a better option that I don't know about 10:53:25 until/if that PR becomes reality 11:22:15 Stable (0.23) branch on underhound.eu updated to: 0.23.1-91-gf373564dc4 13:38:00 -!- amalloy_ is now known as amalloy 13:43:36 03Byrel Mitchell02 {amalloy} 07* 0.25-a0-286-g8ac0393: Add new swamp ending vault with Fedhas and Yred theming. 10(18 hours ago, 1 file, 60+ 0-) 13https://github.com/crawl/crawl/commit/8ac039397e9d 13:48:37 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-286-g8ac039397e (34) 14:16:36 Gutso (L13 DECj) Crash caused by signal #11: Segmentation fault (Lair:3) 14:16:45 !crashlog 14:17:08 21483. lepoulpe303, XL14 MfFi, T:31740 (milestone): https://crawl.xtahua.com/crawl/morgue/lepoulpe303/crash-lepoulpe303-20191211-000408.txt 14:17:45 !crashlog Gutso 14:17:47 No milestones for Gutso (crash). 14:17:52 !crashlog Gutso sprint 14:17:54 No milestones for Gutso (sprint crash). 14:20:20 !lm Gutso 14:20:21 5092. [2019-12-11 19:17:42] Gutso the Thaumaturge (L13 DECj of Fedhas) killed Fannar on turn 22818. (Lair:3) 14:20:27 !lm Gutso -2 14:20:28 5091/5092. [2019-12-11 19:16:10] Gutso the Thaumaturge (L13 DECj of Fedhas) Crash caused by signal #11: Segmentation fault (Lair:3) 14:20:35 !crashlog Gutso 14:20:36 1. Gutso, XL13 DECj, T:22809 (milestone): https://crawl.kelbi.org/crawl/morgue/Gutso/crash-Gutso-20191211-191610.txt 14:23:53 if anyone wants to look at my scoring optimization push, it's https://github.com/crawl/scoring/compare/11ec8e30d6b304fbc1e740e72814eaaa22249950..050d1ba7474aa 14:24:01 I think I might actually be done with this thing for a while 14:24:23 once the current page rebuild finishes 14:25:15 I've also figured out a workflow for building the db on a dfifferent machine and just loading it directly on cao 14:25:21 *different 14:26:37 Gutso (L13 DECj) Crash caused by signal #11: Segmentation fault (Lair:5) 14:35:38 ??ignite poison 14:35:38 ignite poison[1/3]: Damages poisoned (but not merely poisonous) creatures (excluding you) and turns meph/poison clouds into flame. L3 Fire/Tmut/Poison. 14:40:34 03advil02 07* 0.25-a0-287-g56a5b49: Don't crash when checking if fedhas protects player from igniting 10(2 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/56a5b4989272 14:48:25 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-287-g56a5b49892 (34) 15:16:07 advil: nice, I imagine that will help a ton with the db rebuild situations, and thanks for doing that scoring work 15:16:26 I'll hopefully have time to sit down and read it and learn a few things; do you think this will be useful or revamping tournament stuff? 15:16:35 I'm not sure how similar the two codebases are 15:16:47 I think they were similar at one point, but have diverged 15:17:00 some of it might be, it would be nice if they were more similar...it could be too late though 15:17:14 I think some of the concerns for a smaller-scale db are just different 15:17:28 right; I asked since I noticed that they seemed to have some similarity when I briefly looked at the CAO scoring code 15:18:15 it's possible one even started from the other 15:18:25 yeah, would seem plausible that they just started with the scoring code and adapted it for the tournament 15:19:16 the tournament scripts are not very stable in that 1) outages on CWZ would seem to hang the log fetching indefinitely and 2) certain logfile lines didn't get correctly parsed, I think ghost-related ones 15:19:25 oh hm 15:19:29 I already had to fix the ghost parsing code when ghost names containing spaces were introduced 15:19:34 pretty sure both of those issues happened on cao and got some kind of fix 15:19:46 but even with that fix sometimes it would throw and exception and then fail to continue any updates 15:20:05 and the wget issue likewise caused it to not continue, just at the logfile fetching phase 15:20:29 so I had to disabled cwz at times and generally watch the scripts carefully as they needed a restart every day or so 15:21:30 perhaps something I can play with once we get closer to next tournament 15:21:38 yeah, definitely worth taking a look 15:52:17 oh, weird. my trunk game has an issue someone reported a while ago, but i don't remember figuring out the issue. "Orc (0/2) D:11 D:12" 15:52:38 GO0 takes me to an up-stair on D:12, claiming that i am already at the entrance to orc (which is actually on D:11) 15:55:18 er, the opposite. it takes me to an up-stair on D:11, but orc is actually on D:12 15:57:32 advil: I do the same account lockout procedure as you. I set the password to something that will never match and remove any login tokens from the other table 15:57:48 where's the table to remove login tokens? 15:57:59 it looked like that was just in memory 15:58:03 though I didn't look very hard 15:58:36 also in the same sqlite database. You can run .schema to see all tables 15:58:41 maybe it's a webtiles-changes thing though 16:02:57 wow, the autotravel stuff is just all messed up in this game. when i'm on D:12, pressing G< (to go to D:11) warns me that i'll have to go through D:11 to get there 16:03:46 maybe seeding will make it easier to replicate the initial conditions 16:04:11 that's the problem with these travel cache issues I've looked at, once it's broken it's usually kind of hard to figure out why that happened in the first place 16:06:24 perhaps. Game seed: 8757853671676576576, levelgen mode: deterministic. it looks from my log like some weird stuff happened around those floors: i was shafted from D:11 to D:13, then climbed back up to 11 (through 12, of course) 16:45:14 amalloy: did you go in any of the portals around that level range? 16:45:18 !lm amalloy 16:45:24 8108. [2019-12-11 21:28:13] amalloy the Fighter (L14 NaFE of Ru) reached level 2 of the Orcish Mines on turn 27017. (Orc:2) 16:45:45 !lm amalloy x=src 16:45:46 8108. [2019-12-11 21:28:13] [src=cao] amalloy the Fighter (L14 NaFE of Ru) reached level 2 of the Orcish Mines on turn 27017. (Orc:2) 16:56:03 advil: no, no portals. would you like the save backup? 16:56:30 sure...must be something different with whatever trunk version I'm on 16:56:45 though if it's trunk it could just be game upgrades 16:56:58 yeah, game upgrades seem like a good explanation 16:58:21 https://drive.google.com/file/d/1M3PFfY6tTmIIIgiwYW-wd5yyEktr4TOr/view?usp=sharing 16:58:42 i remember there was a gauntlet, but i was shafted before i could find it 16:59:09 yeah, there was a gauntlet on D:11, but also portals on the other two levels 17:00:14 hm, i don't remember those 17:00:45 ah this has seen a bunch of versions including ones much older than what I'm on 17:01:26 how does pregen work with version upgrades, anyway? 17:02:07 incremental pregen just uses whatever the current version gives it, so upgrades mid-game can change seeds 17:02:24 you can still replicate things if you do the upgrades in sync with the game, but it's obviously much more annoying 17:02:55 it's really intended for seeding on a single version, e.g. a stable version or a version used for tournament 17:03:24 it's still possible to do full pregen so for example uv4 still recommends that for gotm games 17:04:30 actually for this game all of D was generated on the same version, I think, because it should have been generated when you entered temple 17:05:40 is full pregen an option on servers, or just local? i'd happily switch to full pregen - the extra few seconds at startup are pretty irrelevant when i start like 4 games a year 17:06:14 it's still disabled (unless I did something wrong) because it's kind of heavy on servers while it's running 17:06:44 I say unless I did something wrong because seeding is accidentally enabled, and I didn't notice until someone asked me how to find the morgues 17:33:19 <|amethyst> btw, any thoughts on incremental pregen doing Temple before D? Would eliminate the noticeable delay when entering the temple (which is about 20% of games), with what I think would be a negligible hit to start-up time for all the other games 17:33:40 yes, it's probably a good idea to do that 17:33:52 the temple map is even already selected so I don't think it would change anything 17:35:08 Wouldn't starting off in Temple have implications for the backgrounds starting with a god? Since then, effectively they all would. Aside from those starting with non-temple gods, of course 17:35:26 <|amethyst> Not starting in the temple, just generating temple before D 17:35:46 <|amethyst> so that when you enter the temple, you don't get the delay from having to generate the rest of D first 17:37:21 Oh, sorry, misinterpreted 17:37:41 I mostly popped in here to ask if dazzling flash (in trunk) is supposed to affect demons. It does, but says it only affects "living" targets 17:39:41 <|amethyst> advil: hmm... I wonder just how much special-case code for Temple could be eliminated with such a change 17:40:08 it definitely does some pretty weird stuff at the beginning 17:40:14 a lot of that is for overflow vaults too 17:40:17 <|amethyst> advil: probably not as much as I'm hoping, since you still have to pick the levels for the overflow altars 17:40:25 yeah 17:45:33 !lg sharkfinsouperman cjr spen x=game_key,cv 17:45:35 3. [game_key=sharkfinsouperman:cjr:20180018191906S;cv=0.20] sharkfinsouperman the Acrobat (L27 SpEn of Fedhas), escaped with the Orb and 3 runes on 2018-10-23 00:27:59, with 1398231 points after 133461 turns and 1d+4:57:49. 17:45:49 hm, I'm *still* missing some CJR games that are in sequell 17:46:34 <|amethyst> some of the log entries may have been overwritten, particularly back in the beginning 17:47:17 yeah...there were some extra log files that gammafunk found that cao scoring didn't have, but they aren't complete I guess 17:47:57 !lg sharkfinsouperman cjr spen x=game_key,cv,src 17:47:57 3. [game_key=sharkfinsouperman:cjr:20180018191906S;cv=0.20;src=cjr] sharkfinsouperman the Acrobat (L27 SpEn of Fedhas), escaped with the Orb and 3 runes on 2018-10-23 00:27:59, with 1398231 points after 133461 turns and 1d+4:57:49. 17:47:57 <|amethyst> I seem to recall this happening more with CKR actually 17:48:00 <|amethyst> but maybe I'm misremembering 17:48:28 <|amethyst> incremental downloads are great, until the file gets replaced on the server 17:48:47 yeah 17:49:03 G-Flex: demons are living, but not natural. nonliving just covers stuff like statues and golems 17:49:10 though on the other hand, incremental downloads mean that everything is cached when something gets removed 17:50:21 <|amethyst> scoring could do better by fetching and checksumming the first 100 bytes or so of the file before determining whether to do an incremental download or to rotate and re-download instead 17:51:23 <|amethyst> but that's an extra HTTP request (and, because we're calling out to external curl, an extra connection) for every file we would download 17:51:23 the actual implementation for mons_can_be_dazzled ignores undead/nonliving/plant, and also anything specially marked M_UNBLINDABLE 17:51:36 yeah, the wgets already take a while 17:51:44 which is like...apparently you can't blind bats because they use echolocation 17:52:10 <|amethyst> which is dumb 17:52:15 <|amethyst> because Crawl bats are megabats 17:52:28 <|amethyst> which don't generally echolocate 17:52:39 amalloy: Okay, thanks, just wanted to make sure. So an angel would also count? 17:52:43 (as living) 17:52:44 |amethyst: alexjurkiewicz was suggesting something more api-based, where a client could request "all games since X game_key" or something along those lines, which would solve this problem 17:53:18 or "all games since X time" 17:53:18 <|amethyst> advil: it would... and servers already have to have sqlite anyway 17:53:18 yes, angels are MH_HOLY with no special tags 17:53:29 |amethyst: i did not know megabat is a real thing 17:53:55 <|amethyst> amalloy: let me tell you about sporulation 17:53:56 also microbats 17:54:40 <|amethyst> My DM recently used the word "sporulate"; I was so proud 17:56:08 so now, aside from whatever missing CJR(/CKR,etc) games there are, scoring has all problems that I know of fixed; should be no duplicate games, streaks and counts are correct, etc 17:56:22 <|amethyst> thanks for doing this btw 17:57:04 no problem...it was fun in the way that these things can be 17:57:22 <|amethyst> I looked into the scoring code a few times, tweaked it once or twice (nearly always with unanticipated side effects), and I still have nightmares about it 17:57:29 heh 17:57:58 tbh I doubt that anything I have done has made it easier 17:58:39 code for efficient bulk database reads is probably always less clear than the more straightforward code it replaces 17:58:52 <|amethyst> ah, but now you have made it easier for everyone else by being the person they turn to when there are problems :P 17:58:56 haha 18:22:16 Unstable branch on underhound.eu updated to: 0.25-a0-287-g56a5b49892 (34) 18:55:54 about pre-generating temple: instead, make temple replace a normal dungeon floor instead of being a special branch :D so players go d4->temple->d6->... 18:57:27 the temple maps are so beautiful. And often players never even enter the branch. It's such a shame 18:59:21 advil: the q&a threads have been a really successful initiative. Nearly 150 comments in the most recent one. I think they've been responsible for a big uptick in subreddit participation. thumbs up for landing such a successful community feature 18:59:53 i can imagine some issues with a branch whose levels are noncontiguous 19:00:39 it could be purely cosmetic, internally refer to the floor as d5 19:06:54 <|amethyst> alexjurkiewicz: you'd have to deface all those beautiful maps by adding extra stairs :P 19:08:38 should we bring back santa hats for elves? 19:16:18 %git faf6667daaae57768ed8a82a17eac6883e728d37 19:16:18 07bh02 * 0.18-a0-895-gfaf6667: Bring back Christmas 10(4 years ago, 46 files, 126+ 90-) 13https://github.com/crawl/crawl/commit/faf6667daaae 19:16:23 absolutely 19:16:55 %git 1d3b463666d87237ef3dab843e9c4e0833c67810 19:16:55 07bh02 * 0.18-a0-947-g1d3b463: Back out Christmas Easter Eggs. 10(3 years, 11 months ago, 47 files, 92+ 123-) 13https://github.com/crawl/crawl/commit/1d3b463666d8 19:17:03 the other commit IDs referenced here don't seem to exist... 19:18:26 alexjurkiewicz: they look to me like they exist, e.g. https://github.com/crawl/crawl/commit/96bdeba11489fe4e89068b4853c948175d010968 19:19:23 oh yeah. Don't grep for them with a trailing `.` then they show up 20:12:44 -!- amalloy is now known as amalloy_