00:01:03 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.23-a0-473-gc8aadda (34) 00:26:06 ok, found that crash 00:26:27 changing the player's location from inside viewwindow() is contraindicated 00:29:05 is this a result of your changes in the branch? 00:29:09 glad you were able to find it! 00:29:19 yea 00:40:45 floraline: I don't know if you have that branch configured to auto-update 00:41:29 03ebering02 07https://github.com/crawl/crawl/pull/872 * 0.23-a0-445-g6766ed7: Refactor to prevent a crash 10(6 minutes ago, 5 files, 46+ 26-) 13https://github.com/crawl/crawl/commit/6766ed741024 00:41:32 if not, this update fixes your crash 00:44:56 ebering, could recent Ely commits change how potion of heal wounds work? Change its distribution, for example? I haven't seen roll as low as 6 earlier. learndb says it is 10+(3d28/3). 00:45:11 ebering: thank you for finding & fixing this so quickly! :) the branch will build nightly at 2 EST 01:04:24 ebering: is the dispersal trap supposed to look like a question mark? webtiles 01:09:13 New branch created: labrework (7 commits) 13https://github.com/crawl/crawl/tree/labrework 01:09:14 03gammafunk02 07[labrework] * 0.23-a0-474-g8f6edef: Remove old labyrinth code 10(3 weeks ago, 27 files, 47+ 1412-) 13https://github.com/crawl/crawl/commit/8f6edef70d84 01:09:14 03gammafunk02 07[labrework] * 0.23-a0-475-gb918e59: A basic map for a branching arena Lab portal 10(3 hours ago, 1 file, 79+ 4-) 13https://github.com/crawl/crawl/commit/b918e596293c 01:09:14 03gammafunk02 07[labrework] * 0.23-a0-476-g7a130e9: Lab arena monster sets and support code 10(2 hours ago, 1 file, 271+ 0-) 13https://github.com/crawl/crawl/commit/7a130e9cc9ae 01:09:14 03gammafunk02 07[labrework] * 0.23-a0-477-g64be07b: Lab arena subvaults 10(2 hours ago, 1 file, 256+ 0-) 13https://github.com/crawl/crawl/commit/64be07b03ee6 01:09:14 03gammafunk02 07[labrework] * 0.23-a0-478-g0c65ece: Set Lab random monster list based on monster sets 10(2 hours ago, 1 file, 23+ 0-) 13https://github.com/crawl/crawl/commit/0c65ece2fd48 01:09:14 03gammafunk02 07[labrework] * 0.23-a0-479-gc3d269f: A Lab minotaur loot function 10(2 hours ago, 1 file, 18+ 0-) 13https://github.com/crawl/crawl/commit/c3d269fb2956 01:09:14 03gammafunk02 07[labrework] * 0.23-a0-480-g6469968: Some initial Lab exit subvaults 10(2 hours ago, 1 file, 259+ 0-) 13https://github.com/crawl/crawl/commit/64699683d7dd 01:10:33 !tell ebering branch 'labrework' is up for when you have time to take a look. I'll continue working on it but when ping you again after I've looked over your maps and thought about that 01:10:34 gammafunk: OK, I'll let ebering know. 01:10:56 ugh, my rebuild cgi script cuts off and does not show all of the build output 01:24:36 Yermak: I'm not sure what it was that you saw, but no. 01:24:37 ebering: You have 1 message. Use !messages to read it. 01:24:43 !source potion.cc:203 01:24:44 https://github.com/crawl/crawl/blob/master/crawl-ref/source/potion.cc#L203 01:27:44 Yermak: I think I know what it is. There is currently a bug in webtiles that doesn't display rot in the HUD correctly 01:27:54 my guess is that you were rotted for 4 mhp and didn't know 01:32:47 Wow, yes, good guess 01:32:49 Thank you 01:33:12 This could be very well much more than 4. 01:46:13 contemplating changing the effect of confusion on monsters to something more like what players get 01:46:17 or maybe even weaker 01:47:30 specifically: a confused monster is still vulnerable to stabs, but on the monster's move it gets a chance to move normally (no abilities or spells (maybe natural abilities)) instead of moving randomly 01:52:11 Isn't this already a thing for the player? 01:53:20 yes 01:53:31 the player the chance is 1/3 01:53:45 I was thinking the monster might get an even bigger chance to move normally 02:00:46 ebering: Frosteey just fell through one of your shafts without crashing :) 02:01:14 they are at Dungeon:14 and have been shafted seven times which feels a bit high 02:01:50 only seven times?! 02:02:30 hey i'm just a novice, those are their words 02:03:09 floraline: yes... that does seem a bit high 02:03:40 14 times or it doesn't feel like you're in a real dungeon 02:05:08 i'm in lair and have only had one alarm 02:07:28 the concept feels good, traps feel more mysterious now and not just another tile on the ground. although i don't understand why the features were all made to be visible 02:11:44 you mean the traps on the ground? 02:12:37 yeah traps on the ground are now all visible 02:13:07 it's my understanding that that is a change made in this branch 02:13:07 yes 02:13:07 if they were not visible, it would be the best of both worlds 02:14:04 the purpose of them being visible is to make them effective tactical terrain 02:14:09 places to fight near or far from 02:14:37 the random effects creating tense situations are better off decoupled from specific floor tiles 02:14:55 i definitely agree with that, traps as a 'chance to encounter' feel really good 02:15:29 i ran into a vault earlier that had some alarm traps in the entryway 02:15:49 those vaults might need a look 02:15:50 so the surprise of being alarmed and marked was not there at all, i just had to walk into it 02:16:06 and it asked me, are you sure? and it felt pretty dull to say yes 02:16:23 there are alarms in the ant vaults but further back 02:16:41 with the change to alarms the ants now mark you if they step on them 02:17:35 it almost feels like some traps should be visible and some should not, in addition to traps as a chance to encounter 02:18:25 nah, the invisible ones lead to spoilers; if you recognize the vault you know the trap is there 02:18:25 better to revise the vaults 02:19:20 but if you don't recognize it it's a very enjoyable event 02:19:49 would a vault like the one i described just not exist anymore? 02:20:03 yes, but a crawl development goal is to get good replayability without spoilers 02:20:25 so, somethign that only works if you're unspoiled must go 02:20:38 in its current form a vault like that would not exist, no 02:20:51 or, if someone got around to revising it, that is 02:21:01 i guess so 02:21:03 it's hard to search the vault description files for trap placement 02:34:20 s/for trap placement// 04:54:14 -!- amalloy is now known as amalloy_ 07:26:18 New branch created: pull/911 (1 commit) 13https://github.com/crawl/crawl/pull/911 07:26:18 03Nikolai Lavsky02 07https://github.com/crawl/crawl/pull/911 * 0.23-a0-474-g29a656d: Fix messaging when Maurice steals a weapon (Salmon) 10(21 minutes ago, 1 file, 68+ 69-) 13https://github.com/crawl/crawl/commit/29a656d013bd 10:32:01 03ebering02 07https://github.com/crawl/crawl/pull/872 * 0.23-a0-446-gfbeec78: Tone down the explore trap rate 10(5 minutes ago, 1 file, 24+ 17-) 13https://github.com/crawl/crawl/commit/fbeec78c918b 11:21:00 re the previous conversation about traps, don't forget about the uhh 11:21:04 sand place vaults 11:21:08 with the alarm traps in the doorway 11:21:14 ossuary 11:21:44 also personal opinion, being shafted more than one floor is complete bullshit and shouldn't be in the game 11:22:42 related personal opinion, being shafted sucks 11:22:58 !lg * current s=job / won 11:23:01 6296/489781 games for * (current): 1536/109524x Fighter [1.40%], 655/38411x Berserker [1.71%], 437/28925x Gladiator [1.51%], 378/32077x Monk [1.18%], 347/21214x Wanderer [1.64%], 300/14892x Hunter [2.01%], 229/18564x Wizard [1.23%], 221/23824x Fire Elementalist [0.93%], 190/14214x Earth Elementalist [1.34%], 181/17067x Skald [1.06%], 180/32013x Transmuter [0.56%], 161/15805x Conjurer [1.02%], 147/... 11:23:15 I gotta get better about my irssi split screens 11:23:46 can you do that same command but sort by win%? 11:24:19 switch___: no, but you can via pm with sequell 11:24:24 how 11:26:17 add o=% to the end of the command 11:26:37 -!- switch___ is now known as switch__ 11:27:32 lol 11:27:35 artificer has the highest win% 11:27:55 I should really write a statistical disclaimer and put it in the learndb 11:28:15 what about? 11:28:31 any and all summary statistics derived from the crawl online games dataset 11:29:12 does ar have the highest win rate because it is the strongest background? or because it's unpopular among most players except for some who think it is optimal for streaking? 11:29:19 well 11:29:29 it's over 4500 games, which is less than 1% of the sample size 11:29:43 while fighter, with a 1.4% winrate (3.04% for artificer) is more than a fifth of all games 11:29:59 so i think the answer is pretty clear 11:30:03 !lg * current job=ar x=cdist(name) 11:30:05 4548 games for * (current job=ar): cdist(name)=964 11:30:49 6296/489797 games for * (current): 143/4548x Artificer [3.14%], 147/7109x Warper [2.07%], 300/14892x Hunter [2.01%], 147/7363x Chaos Knight [2.00%], 99/5397x Arcane Marksman [1.83%], 655/38411x Berserker [1.71%], 347/21214x Wanderer [1.64%], 437/28925x Gladiator [1.51%], 1536/109531x Fighter [1.40%], 113/8239x Venom Mage [1.37%], 190/14214x Earth Elementalist [1.34%], 127/10114x Assassin [1.26%], 11:30:49 229/18565x Wizard [1.23%], 99/8223x Air Elementalist [1.20%], 378/32080x Monk [1.18%], 104/8926x Ice Elementalist [1.17%], 92/7957x Summoner [1.16%], 181/17068x Skald [1.06%], 161/15805x Conjurer [1.02%], 128/13484x Necromancer [0.95%], 221/23824x Fire Elementalist [0.93%], 145/18454x Enchanter [0.79%], 137/23440x Abyssal Knight [0.58%], 180/32014x Transmuter [0.56%] 11:31:10 switch__: the whole point of pming sequell is to not clutter crawl-dev with its replies 11:31:28 i actually truncated that post quite heavily before pasting it 11:31:32 and then accidently pasted the full version 11:31:34 :) 11:32:25 but uhh 11:32:30 transmuter = noob trap ? 12:05:15 Unstable branch on crawl.akrasiac.org updated to: 0.23-a0-473-gc8aadda (34) 12:17:47 -!- MarvinPA_ is now known as MarvinPA 12:50:27 !lg * won current x=avg(dur),var(dur) 12:50:28 Unknown function: var 12:55:47 hey all, quick server question for you. i've noticed some significant slowdown in loading a game ever since i switched over to ubuntu 18 and re-installed everything. the game itself runs fine once it loads, but sometimes i see that loading screen on the order of a minute to more. any thoughts on things i can check? 13:01:18 espais: i have heard this issue reported on some other servers in the past 13:01:31 it's likely that some server admin is familiar with it 13:04:00 fair enough 13:04:10 i haven't seen the issue myself, sorry 13:04:18 HDD space is fine, RAM usage is OK, normal things aren't really impacting 13:04:48 alrighty, thanks floraline 13:04:58 btw, did a run through of quick crawl last night, that is super fun 13:05:07 first orb get in a while for me :) 13:05:37 the way you described that the other day made it sound interesting, i'll have to check it out 13:53:41 New branch created: pull/912 (1 commit) 13https://github.com/crawl/crawl/pull/912 13:53:41 03stenella02 07https://github.com/crawl/crawl/pull/912 * 0.23-a0-459-g1d2436e: MSVC support cont. 10(3 minutes ago, 3 files, 34+ 3-) 13https://github.com/crawl/crawl/commit/1d2436efc1e9 14:04:19 03stenella02 07https://github.com/crawl/crawl/pull/912 * 0.23-a0-459-g53c7bd8: MSVC support cont. 10(13 minutes ago, 3 files, 34+ 3-) 13https://github.com/crawl/crawl/commit/53c7bd86fa93 14:04:27 (fixed a typo) 15:34:59 I have a relatively small suggestion that I'm 100% serious about: remove the returning ego, rename tomahawks to boomerangs and make them always return 15:35:25 Why? 1. Reduces inventory clutter while keeping the returning effect, allowing the gameplay of that one bailey to remain (yes boomerangs aren't axes but who cares) 15:35:33 are the boomerang made out of bamboo 15:35:36 2. Tomahawks, to my knowledge, have not been found in Australia while boomerangs have 15:36:18 -!- amalloy_ is now known as amalloy 15:36:22 3. Making them *always* return, instead of sometimes returning (it's unreliable even at 27 throwing) makes the effect more interesting since it means you can actually throw them at water/lava without risking losing them to the lava/water 15:37:23 what about things like dispersal tomahawks 15:37:41 those were actually extremely useful for e.g. early formicid, at least for me 15:39:58 Those would be renamed to dispersal boomerangs and otherwise be unchanged 15:41:44 so, to be more clear, the mulch rate is still 20%, dispersal/exploding still do their thing (explode), but common/silver/steel tomahawks now all have 100% returning ego 15:42:50 += poisoned 15:43:30 err, s/20%/5% 15:45:19 yeah 15:45:37 returning doesn't increse the damage or decrease the mulch rate or anything so it shouldn't affect balance 16:07:29 minmay: intriguing. I've been thinking about blowgun removal (turning needles into darts) and some tweaks to throwing brands 16:10:04 i'm not actually super familiar with the balance of throwing when it comes to gameplay 16:10:12 is it possible to play with throwing as your primary source of damage, on any class? 16:10:29 s/possible/reasonable 16:12:31 switch__: why don't you try 16:12:33 :) 16:13:41 well, when i'm not doing things i have a habit of letting myself get behind on, or learning crawl (the code), i'm working on greaterplayer, for which i have... 8 wins left? i think 16:13:43 so, soon 16:30:36 ebering: btw, one basic problem with the maps you sent is that they aren't a real choice; you don't get to understand what choosing a given path will lead to 16:31:42 when one of you gets a chance, that PR I just put up doesn't really need to be looked over (unless I'm misunderstanding the part with CREDITS.txt) 16:31:56 in the case of hall_of_mirrors you're just presented with an initial choice 16:32:07 but it's not really going to be clear to the player what's being decided 16:33:29 oh, that's neat. since when does github put "xyz force-pushed this branch from abcd to 1234" messages into PRs? 16:34:17 pretty recently i think 16:34:52 because i went back to one of my first PRs and it had literally 15-20 force pushes in a row that I'm pretty sure weren't there before :P 16:36:44 gammafunk: yeah, I think I had a slightly different conception of the rework from yours. 16:36:44 amalloy, same reason as last time. This file is only used by MSVC, and without it chrono and ratio break really hard. 16:37:02 working on robe of Clouds right now 16:37:18 https://github.com/crawl/crawl/pull/878/files 16:37:40 err, https://github.com/crawl/crawl/pull/878/files#diff-13a5ee50343a26462bbfe00974d4c914 16:38:32 no worries, I had plans for a similar idea, but maybe I can incorporiate the more flavorful layout 16:41:47 ho ho 16:41:51 he's updated the wiki page 17:04:48 03stenella02 07https://github.com/crawl/crawl/pull/912 * 0.23-a0-459-ga2bc8c0: MSVC support cont. 10(3 hours ago, 4 files, 36+ 4-) 13https://github.com/crawl/crawl/commit/a2bc8c08dc19 17:14:56 gammafunk: you pointed me at https://github.com/crawl/sizzell before, right? 17:15:27 yes, that's the current irc bot people use 17:17:22 does this need to run on the same system as the game? or can it get everything it needs from the public directories? i imagine it demands few resources 17:17:55 and is every server running its own bot? is this expected of an official server? 17:18:57 floraline: it does need to run on the server, yes, so that it can see crashes and rebuilds. We have a number of servers not running this bot, so do feel that you have to run it 17:19:24 Since Postsequell exists for ##crawl, we can get milestones/game messages for CKO 17:19:34 so the benefit of this bot is just crash notifications and rebuild notifications 17:19:49 the former are pretty helpful for us since otherwise we don't know when crashes happen 17:20:05 but several servers (cpo, cwz, lld) lack this already 17:20:46 er 17:20:46 s/so do feel/so don't feel/ 17:21:39 floraline: so in summary, you definitely don't have to run it; the only things we truly require are related to just availability and public meta data 17:23:31 I suppose crashes in principle could be extracted by a non-server-specific bot, they are just milestones 17:24:57 yeah that's probably true 17:36:37 03stenella02 07https://github.com/crawl/crawl/pull/912 * 0.23-a0-459-g37e5af4: MSVC support cont. 10(4 hours ago, 5 files, 35+ 5-) 13https://github.com/crawl/crawl/commit/37e5af4e0ac0 17:44:01 k, that should be the last of that. maybe if i give amalloy his clean PURE definitions back he won't scrutinize my PRs so hard 17:46:56 switch__: wait, i didn't notice this before - why is /docs/CREDITS.txt added to the gitignore? that file doesn't exist normally. does some build process put it there? 17:47:14 03stenella02 07https://github.com/crawl/crawl/pull/912 * 0.23-a0-459-g11796d7: MSVC support cont. 10(4 hours ago, 4 files, 36+ 4-) 13https://github.com/crawl/crawl/commit/11796d77756b 17:47:30 i'm not 100% on this either. it's definitely inside the makefile, and it's definitely inside the release version's docs folder 17:47:52 but it's also probably definitely not inside any development builds because nobody else added it to the gitignore 17:49:48 line 1409/1729 of the makefile definitely seems to copy it into that directory, and it's definitely in /docs (and only in /docs) when i download the release version for windows tiles 17:50:23 hmmmm. the makefile seems to put it in $datadir_fp/docs, which shouldn't be ./docs 17:51:00 ah okay, so 1409 puts it in the directory above docs 17:51:31 and 1729 moves it to docs, for some reason 17:51:39 but only for windows 17:51:41 the directory above docs is where it normally lives 17:51:53 1409 copies it from there into the destination 17:52:58 and 1729 copies it from $destination/docs to $destination, undoing the step from 1409 specifically for windows. why? i'm not sure yet 17:53:18 but it still shouldn't wind up in /docs in any of those cases 17:54:24 oh, you're right, i read it backwards somehow 17:54:35 uhh, let me double check the reason i thought this in the first place 17:54:47 yeah 17:55:00 if you download crawl off develz.org and extract the zip files, it's in docs 17:55:15 sure, but that's fine 17:55:20 (i was going through and comparing my build environment to that, since i was missing actually important files, like aptitudes) 17:55:22 the release build puts it into docs/ for whatever reason 17:55:40 but it's docs/ of the release archive, not docs/ of the dev environment 17:56:04 so it's entirely unnecessary to have it in docs, then 17:56:32 i'll remove that part of the gen file/gitignore, no problem 17:57:52 cool. PR looks good after that but i'm heading out shortly so won't merge it right away 17:57:52 fuk 17:58:01 this little discussion is why i don't believe there's such a thing as a PR that doesn't need to be looked over 17:59:13 oh, i wrote in my original statement that i wasn't sure about the credits thing 17:59:57 pushed 18:00:01 well, sure. but there was also the limits.h stuff 18:00:06 true 18:00:23 honestly now that i'm WAYYY more comfortable with MSVC I should probably go revisit some of the hacks we pushed through 18:00:42 03stenella02 {amalloy} 07* 0.23-a0-474-gee2f500: MSVC support cont. 10(4 hours ago, 4 files, 31+ 3-) 13https://github.com/crawl/crawl/commit/ee2f500f8203 18:00:53 03stenella02 07https://github.com/crawl/crawl/pull/912 * 0.23-a0-459-ge781e4a: MSVC support cont. 10(4 hours ago, 4 files, 31+ 3-) 13https://github.com/crawl/crawl/commit/e781e4a546b0 18:02:20 something went weird there 18:02:24 maybe just the commit message 18:02:46 looks correct on github 18:03:33 I think you just didn't update the commit message, and chei printed them out of order 18:03:36 but too late now 18:04:00 oh, yeah 18:04:03 i didn't update the commit message 18:04:13 so it has the credits change that didn't happen 18:05:04 and the stdint stuff 18:05:41 100% my fault, ofc, as malloy said he was leaving and i asked for a merge anyway 18:06:16 good thing i'm the only one that ever sees the results of these changes >_> 18:06:32 until i convert all of you to MSVC 18:06:41 !messages 18:06:43 No messages for ZiBuDo. 18:07:04 gammafunk: I'll be shutting the server down today ( the crawl server) so that alex can do the final back up 18:07:14 ZiBuDo: cool, thanks for the heads up 18:07:28 I can make a post about that and I guess I should go ahead and announce CKO 18:07:37 good night sweet prince 18:07:42 we don't yet have it hooked up to CAO scoring, so I'll have to mention that 18:07:44 ZiBuDo: thanks for hosting the server over the years 18:07:55 !lm . cjr alive 18:08:05 No milestones for gammafunk (cjr alive). 18:08:12 !lm . cjr alive_all 18:08:14 No milestones for gammafunk (cjr alive_all). 18:08:16 hmm 18:08:22 !kw alive_all 18:08:23 Keyword: alive_all => ktyp= type!=crash 18:08:27 !lm . cjr 18:08:29 1153. [2018-08-26 07:16:50] gammafunk the Slayer (L27 GrSk of Ru) found the Orb of Zot! (Zot:5) 18:08:32 !lm . cjr desu 18:08:33 No milestones for gammafunk (cjr desu). 18:08:38 woah 18:08:45 !lm . desu alive 18:08:47 No milestones for gammafunk (desu alive). 18:08:47 what's up 18:08:54 not sure, I think I have a desu on cjr 18:08:59 but I don't see milestones 18:09:03 doesn't matter 18:09:11 I was just curious about that game 18:09:21 doh! 18:09:23 it's nick 18:09:30 !lm omegafunk alive cjr 18:09:32 6. [2018-08-03 13:06:28] gammafunk the Convoker (L13 DESu of Sif Muna) reached level 6 of the Lair of Beasts on turn 18806. (Lair:6) 18:09:38 Webtiles server stopped. 18:09:59 oh i was wondering if that message was announced 18:10:13 yeah, guess that's another benefit of the irc bot 18:10:29 rip my pretty decent desu turncount run that I turned into a normal run but now will be lost forever 18:10:41 I couldn't win it since it would have been part of my streak and would be kind of cheating 18:10:43 gammafunk: working on irc bot .. i don't mind running it for crashes / rebuilds. i would prefer milestones be read by someone else 18:11:04 floraline: cool, we can ask Kramin to include it in Postsequell 18:11:11 great 18:11:32 !tell Kramin would you be able to add cko to postquell's announcing? Thanks 18:11:32 ebering: OK, I'll let kramin know. 18:20:06 correct me if i'm wrong: 18:20:39 Unstable branch on underhound.eu updated to: 0.23-a0-473-gc8aadda88e (34) 18:20:42 basically all monster-related functions pull from a monsterentry, which is what you get from get_monster_data(). this, and the actual monster object, aren't actually directly linked in any way 18:21:04 e.g. to get a monster's habitat, you'd get it's type, pull it's monsterentry, and check that monsterentry for that monster's habitat 18:21:18 why isn't that data just protected data inside the monster object itself? 18:22:16 I don't know the legacy reason (you could pull out git blame but I bet it'd go back to the initial revision, so "inherited from linley") 18:23:06 heh 18:23:31 i guess the monster object is big enough as it is 18:23:34 I'm not sure that's the full story 18:23:41 but these checks are soooooo convoluted for literally no reason 18:23:47 monsters get saved 18:23:49 it has the practical side effect of making save compatibility for changes to monsterentries automatic, instead of having to add corrections in the unmarshalling code 18:23:55 yes 18:24:17 ah, that makes sense 18:29:38 03ebering02 07* 0.23-a0-475-gd9f91f9: checkwhite 10(36 seconds ago, 1 file, 3+ 3-) 13https://github.com/crawl/crawl/commit/d9f91f930101 18:29:51 switch__: remember to checkwhite! 18:30:06 does github not show leading spaces? 18:30:09 amalloy: advil maybe we should re-visit alexjurkiewicz suggestion we add checkwhite to travis 18:30:16 not by default 18:30:20 i thought it did 18:30:24 won't happen again 18:32:36 03ebering02 07* 0.23-a0-476-g6e8b230: Revise the robe of Clouds 10(34 minutes ago, 11 files, 38+ 25-) 13https://github.com/crawl/crawl/commit/6e8b230bb5f2 18:32:36 03ebering02 07* 0.23-a0-477-gda178e9: Remove the storm bow's rain clouds 10(18 minutes ago, 1 file, 0+ 12-) 13https://github.com/crawl/crawl/commit/da178e91be7a 18:35:56 &rc . cjr 18:36:00 http://crawl.berotato.org/crawl/rcfiles/crawl-cjr/gammafunk.rc 18:36:08 &rc . trunk cjr 18:36:10 http://crawl.jorgrun.rocks/rcfiles/crawl-git/gammafunk.rc 18:38:10 (ignore the little string to the left) reading left to right, this is what i was screwing with yesterday https://i.imgur.com/k4DiuRp.png . pretty sure this entire top half (mons_primary_habitat) doesn't do anything, other than prevant lava-dwelling creatures from drowning in deep water 18:39:02 03ebering02 07* 0.23-a0-478-g22ffb11: Convert Death Channel/Bind Soul to fineffs 10(4 hours ago, 3 files, 84+ 30-) 13https://github.com/crawl/crawl/commit/22ffb116fb95 18:39:16 (looking at blame goes back like 5+ years) 18:40:06 maybe the performance impact is minimal, but that's a hell of a lot of calls just to return whether or not a creature's primary habitat is lava, and when is that ever going to try to drown in deep water anyway? 18:41:50 trust in the optimizer 18:44:34 (also, doing this kind of check on an invalid monster ("programming bug") will make you very confused, because invalid monsters are set to look relatively normal, and have land as their habitat). the comment above the definition for invalid monsters states that this is intentional, so people can use it as a template 18:44:35 it comes up with salamanders in snake 18:44:53 salamanders are not lava, they're lava_amphibious, and fail this check 18:45:04 ... right? 18:45:17 floraline: what rebuild schedule did you settle on, by the way? daily? 18:45:26 I'm making the wordpress post right now 18:48:26 yeah, salamanders totally drown 18:49:21 gammafunk: daily for now at 3:30 EST 18:49:22 floraline: and another question, does cko support prompting for trunk save transfers or is that done regardless 18:49:25 thanks 18:49:28 ??rebuild[2 18:49:28 rebuild[2/2]: Trunk update start times in UTC (second number during summer time): CAO: 1700; CDO: 0600/0500; CUE: 2300/2200; CBRO: 0800/0700; CXC: 0500/0400; CPO: every 15 minutes; CJR: every hour 18:49:53 i'm not opposed to moving to a more frequent schedule 18:49:57 but that will take some consideration 18:50:07 I think daily is just fine 18:50:12 i haven't changed that behavior, it appears to ask if you want to transfer to the new trunk 18:50:17 cool 18:50:26 i don't intend to force users to move forward all that often 18:50:34 if ever, depends on disk space i guess? 18:51:38 yeah, if you have trouble with disk space that's one way to reclaim it 18:51:47 johnstein has done this for cbro 18:51:56 i am running 2 GB memory and 2 vCPU right now. i am prepared to resize for two more steps.. to 4GB 2 vCPU, then to 8GB 4 vCPU 18:52:11 i'm not prepared to upgrade past that and i hope i won't need to 18:53:43 i have some disk space reclamation scripts running already to offload ttyrec and morgue files every 3 hours, and clear unused trunk versions daily 18:54:23 yeah, I don't think there's a huge benefit for a rebuild schedule that's twice a day since we could just do a manual rebuild to test things 18:54:36 i need to write a script to clear trunk builds whose binary has been removed, and i'm unsure how quickly logfiles/milestones/etc will grow 18:54:39 but an hourly rebuild was sort of nice to find bugs when we didn't expect that; we still have cpo for this, of course 18:54:43 which is every 15 minutes 18:55:12 but it's far more important that users have a good play experience and that your server doesn't get overloaded 18:55:16 i think i'd be able to run an hourly build for only one branch, like trunk 18:55:37 i can definitely do that if i end up moving to 4 CPUs 18:55:49 ah, well pretty much only talking trunk here; stable version changes very infrequently 18:56:05 well i was thinking it would also be nice to update experimentals more regularly 18:56:07 we usually only change them as we cherry-pick changes before a point release 18:56:11 but i guess that's less priority than trunk 18:56:31 i do have 0.21 and 0.22 still building daily, i wasn't sure if they will ever see changes at this point 18:56:32 yes, but for experimentals it's probably even more the case that the dev behing the experimental would do a manual rebuild 18:56:48 oh that makes sense.. well my rebuild script is not working right now so i'll have to fix that 18:57:30 yeah, it's very nice if you do have stable versions set up to pull daily; they won't see any changes 95% of days, but when there is a fix it will happen automatically, just as it does for trunk 18:58:10 sadly cwz doesn't ever even get an actual stable release; hong just installs the pre-release beta and then never updates it :( 18:58:50 but he does have to deal with applying local patches and is running his own sort of setup 19:10:43 Unstable branch on crawl.jorgrun.rocks updated to: 0.23-a0-478-g22ffb116fb (34) 19:20:42 ??rebuild[2 19:20:42 rebuild[2/2]: Trunk update start times in UTC (second number during summer time): CAO: 1700; CDO: 0600/0500; CUE: 2300/2200; CBRO: 0800/0700; CXC: 0500/0400; CPO: every 15 minutes; CJR: every hour 19:21:45 !learn edit rebuild[2] s|CJR.*|CKO: 0830/0730| 19:21:46 rebuild[2/2]: Trunk update start times in UTC (second number during summer time): CAO: 1700; CDO: 0600/0500; CUE: 2300/2200; CBRO: 0800/0700; CXC: 0500/0400; CPO: every 15 minutes; CKO: 0830/0730 19:21:49 ??rebuild 19:21:49 rebuild[1/2]: http://crawl.akrasiac.org/rebuild/ http://underhound.eu:81/rebuild/ http://crawl.berotato.org/crawl/rebuild/ http://crawl.xtahua.com/rebuild/ https://crawl.jorgrun.rocks/rebuild/ Bug gammafunk, |amethyst, or Nap.Kin for CDO. Use your powers wisely. 19:22:36 !learn edit rebuild[1] s/jorgrun.rocks/kelbi.org/ 19:22:36 rebuild[1/2]: http://crawl.akrasiac.org/rebuild/ http://underhound.eu:81/rebuild/ http://crawl.berotato.org/crawl/rebuild/ http://crawl.xtahua.com/rebuild/ https://crawl.kelbi.org/rebuild/ Bug gammafunk, |amethyst, or Nap.Kin for CDO. Use your powers wisely. 19:23:06 !learn edit rebuild[1] s/gammafunk/gammafunk, advil/ 19:23:07 rebuild[1/2]: http://crawl.akrasiac.org/rebuild/ http://underhound.eu:81/rebuild/ http://crawl.berotato.org/crawl/rebuild/ http://crawl.xtahua.com/rebuild/ https://crawl.kelbi.org/rebuild/ Bug gammafunk, advil, |amethyst, or Nap.Kin for CDO. Use your powers wisely. 20:09:56 floraline: Made a post about CKO: http://crawl.develz.org/wordpress/new-crawl-server-cko-and-server-cjr-taken-offline 20:14:49 ??plan 20:14:49 plan[1/2]: https://github.com/crawl/crawl/wiki/0.23-Development-and-Release-Plan 20:14:59 The build passed. (master - 22ffb11 #10645 : Edgar A. Bering IV): https://travis-ci.org/crawl/crawl/builds/460528325 20:25:06 gammafunk: awesome! thank you for helping me get everything set up 20:25:25 np! 20:25:25 i'll have the irc bot printing crash/rebuilds in here eventually, and i still need to fix rebuild script. otherwise everything should be good to go 20:26:02 sounds good; hopefully we can get CAO updated soonish, but that might take a week or so 20:31:11 It's Official 20:34:53 Shatter is still appearing on player ghosts, is this intentional? 20:38:25 It lets you break into stone/metal vaults etc. without investing the xp to cast shatter yourself, especially now that they can use stairs 20:39:47 that's true, but this is also true of monsters with the dig spell 20:39:58 in terms of luring them around to dig into areas where that might be helpful 20:54:49 alexjurkiewicz: you got Zibudo's message about CJR webtiles going offline so you could make a final backup? 20:55:32 gammafunk: dig doesn't affect stone or metal though, which minmay was careful to qualify 20:55:48 amalloy: yes, I know, but I'm saying it's the same principle 20:56:15 there are even vaults that have diggable rock blocking access 20:56:30 so you can lure any of the digging monsters over to those if you don't have dig or sacced artifice or whatnot 20:56:49 but even beyond that you could have situations where digging set up something favorable for you 20:57:28 I don't feel particularly strongly that player ghosts need to have shatter, to be clear, but it's not quite the case that we forbid monsters to have terrain changing stuff 20:57:43 i think it's a bit different because getting shatter on a ghost isn't something you can count on (i.e. it's RNG) 20:58:42 issue has nothing to due with RNG; the issue is what we permit monsters to do 21:00:31 but that's exactly my point, are ghosts even really traditional monsters? 21:01:17 i don't have an opinion, i just mean to say you can probably treat this as a special case 21:02:11 I don't know what that statement means, but yes player ghosts are monsters (you fight them for xp and loot) and it's not anything about RNG 21:02:36 floraline: I've updated the wordpress and other stuff about CKO as well as sequell commands 21:03:19 floraline: but there may be other things that are out of date, so you can just let us know about those as you find them 21:03:28 also updated the html website with the new server json info 21:03:59 I had to use New York City for latitude/longitude, but I doubt it matters if the server is hosted elsewhere in the state 21:04:06 gammafunk: https://github.com/gammafunk/webtiles needs CKO!! 21:04:07 since there's no other server close enough for that to matter 21:04:17 ebering: I have that locally actually 21:04:28 but...how do you know about that repo? 21:04:50 you told me about it in advance of the tournament for deploying rcs 21:04:55 ah, yep 21:07:01 -!- amalloy is now known as amalloy_ 21:25:21 gammafunk: thanks, i'll keep an eye out for anything that needs updating. the datacenter is in fact in NYC, so that's perfect 21:29:09 -!- amalloy_ is now known as amalloy 21:32:05 gammafunk: on /wordpress/howto, crawl.kelbi.orb should be org, and update should be updated 21:33:45 shatter on pghosts should probably be removed 21:33:53 the logic on removal from pan lords an jorgrun applies 21:34:03 %git 9c1766aa3124b70 21:34:03 07MarvinPA02 * 0.22-a0-25-g9c1766a: Remove Shatter from pan lord spell lists 10(11 months ago, 1 file, 0+ 1-) 13https://github.com/crawl/crawl/commit/9c1766aa3124 21:37:27 03ebering02 07* 0.23-a0-479-g35efe26: Nerf player-to-monster confusion 10(45 minutes ago, 2 files, 17+ 41-) 13https://github.com/crawl/crawl/commit/35efe2662d19 21:37:40 gammafunk: ok doing this now 21:37:40 ruining crawl since forever 21:38:18 I have morgues, ttyrecs and rcfiles from CJR backed up. If people think anything else should be preserved speak now or forever hold your peace 21:38:32 Also, if anyone would like access to the S3 bucket let me know 21:39:12 logfiles and milestones 21:39:21 I think those were grabbed elsewhere but 21:39:38 not since the server was officially shut down 21:40:24 alexjurkiewicz: did your backup include the .ts files used by footv? 21:40:52 yes, those are written alongside morgues 21:40:53 alexjurkiewicz: also it'd be good to include .lst files in that backup if you can 21:41:09 ts and lst are both in morgues dir so those are safe 21:41:16 and I guess your rcfile backup includes the .macro and .persist stuff? 21:41:30 ok, sounds like anything in the relevant dirs, which works fine 21:41:48 those are in rcfile dirs 21:41:51 ebering: backing up logfiles/milestones now 21:42:00 hrm 21:42:00 confirming macro/persist files are backed up 21:42:06 is cjr still reachable from console? 21:42:39 yeah, it is 21:43:00 !tell Zibudo Can you also disable console crawl access? console games are currently still playable 21:43:01 gammafunk: OK, I'll let zibudo know. 21:43:14 I should have asked about that when he stopped webtiles 21:43:37 either way shouldn't be too big a deal, just means if someone played a console game it wouldn't be seen 21:54:24 final backup complete 21:54:43 so is |amethyst keen on storing this in an online disk somewhere? 21:54:57 alexjurkiewicz: are there permanent public urls for this data? 21:55:23 currently it's in a private S3 bucket, but I can open it to the world. However S3 doesn't support directory listings 21:55:29 ah, right 21:55:43 oh, but for the logfile/milestone files, there are fixed urls? 21:55:52 and likewise for morgue files 21:56:06 not in terms of any listing, but just permanent fixed urls 21:56:11 I haven't heard back from snark yet 21:56:25 i see that ttyrec/morgue files appear to generate using UTC, is it OK that cko is otherwise on local time? 21:56:50 yes it's fine for it to use its local timezone; the data are always written in UTC afaik 21:57:01 ok 21:57:07 data meaning logfile/milestone date fields 21:57:34 possibly the timestamps in the morgue/ttyrec filenames are also written in utc 21:58:15 yeah cdo is on CET 21:59:32 ebering: I think you broke butterflies 21:59:46 advil: oh no 22:00:16 oh dear I did 22:00:20 !zap ebering 22:00:20 they are permanently enchanted as confused but don't register as confused in mons_confused, so now they don't even get confused 2/3 turns 22:00:20 (∩ ͡° ͜ʖ ͡°)⊃━☆゚. o ・ 。゚ ebering 22:00:22 heh 22:01:15 that's why the mon-behv code checked the enchantment directly 22:01:17 I think I broke every monster that's permanently confused because if its type 22:01:21 yeah 22:01:32 wasn't sure what others there are 22:01:37 twisters? 22:01:41 vortices and twisters 22:01:50 heh that'd suck a bit 22:01:57 (literally!) 22:02:02 sounds like a firestorm buff 22:02:26 which obviously was ebering's true intent all along 22:02:28 true, as long as they have a target 22:02:34 otherwise they'll just head to you 22:02:45 interesting 22:02:48 you can tell your butterflies to attack things too 22:02:53 (they don't do much) 22:02:56 fearsome 22:02:59 so I assume you'd have to to 'ta' after the firstorm 22:03:03 otherwise they'd run over to you 22:03:18 yeah, I'd guess they won't have a target to start 22:04:13 gammafunk: yes I can enable fixed permanent URLs for everything if you like? 22:04:30 alexjurkiewicz: ok, that sounds good; I can update the sequell config after you do that 22:04:36 and give me the relevant urls 22:04:57 well, to be clear, I need a fixed base url for morgue files and then fixed urls for logfile/milestone files 22:05:14 and I guess cjr ttyrecs will have to stay broken for a while 22:05:20 yup 22:06:19 alexjurkiewicz: how long does trunk take to build on cpo? 22:06:27 from scratch? ages 22:06:40 does it not build from scratch often? 22:06:47 gammafunk: https://s3-us-west-2.amazonaws.com/crawl.jorgrun.rocks/_crawl-master_crawl-{git,0.19-22}_saves_{logfile,milestones} 22:06:55 excellent, ty 22:07:27 floraline: no, it keeps a repo directory and runs `git fetch ; make ` every 15mins. But if someone modifies a base header it's basically a full rebuild 22:07:40 floraline: specifically probably 5-20mins for a full build 22:07:54 floraline: CPU is Intel(R) Xeon(R) CPU E5-2686 v4 @ 2.30GHz and there's one of em 22:08:18 15-20 minutes is my time frame, it builds from scratch every time and i don't know why 22:08:24 using dgl update-trunk 22:08:37 gammafunk: morgues are at https://s3-us-west-2.amazonaws.com/crawl.jorgrun.rocks/morgue/01000101/morgue-01000101-20180109-035213.txt and ttyrecs/rc-files are in their own base directory 22:09:32 i'm assuming it is the makefile which allows you to incrementally build, and you aren't using ccache or anything? 22:09:32 floraline: sorry mate, I don't use dgl scripts. Maybe you could install ccache but that might be a deliberate design decision (since most dgl servers rebuild nightly they probably don't notice/care) 22:09:36 yes i suspect it is an intentional decision on dgl's part 22:10:04 right. You can see this by compiling locally, then `touch player.cc ; make` will only recompile player.o, since only that .o's source files have a newer mtime than the .o 22:10:06 right, we had a discussion about that some time back 22:10:27 but if you install ccache it will solve that problem I think 22:10:41 Unstable branch on crawl.jorgrun.rocks updated to: 0.23-a0-479-g35efe2662d (34) 22:10:51 well i've installed and set up ccache but it's not being used 22:11:00 so i got frustrated and started thinking why it wasn't incrementally building in the first place 22:11:44 maybe you don't have the env variables set up properly (re: ccache) 22:12:01 it's in my path, and i ran the symlink script that comes with it 22:12:16 the instructions don't say anything else is needed so i'm at a loss 22:12:18 well, the env set by the rebuild script might be different 22:12:27 03ebering02 07* 0.23-a0-480-g1869896: Don't unconfuse intrisically confused monsters 10(3 minutes ago, 1 file, 7+ 0-) 13https://github.com/crawl/crawl/commit/1869896531fd 22:12:40 I assume the dgl rebuild runs from cron, you might need to add it to your crontab file 22:12:40 it does work on my local system 22:12:40 than the env you're testing as crawl-dev 22:12:54 so much for my killer butterflies 22:13:05 ^ 22:15:30 ??dgl 22:15:31 dgl[1/5]: https://github.com/crawl/dgamelaunch 22:17:07 ok, so it seems you run `dgl update-trunk` when you want to build trunk 22:17:29 that's right. if i'm doing it manually it's `sudo dgl update-trunk` otherwise it is in root's crontab 22:17:33 floraline: in the crontab file, on the line before your dgl update-trunk entry, add PATH= 22:17:58 like, whatever `echo $PATH` shows as your user when ccache is working 22:17:58 well 22:17:58 don't do that 22:18:00 just add the ccache bits 22:18:13 yeah if you can identify them that's better 22:18:14 in fact clobbering your entire path sounds like a bad idea 22:18:29 i have export PATH="/usr/lib/ccache:$PATH" in root's bashrc 22:18:38 that's not going to matter for cron 22:18:59 does cron use the env in crontab? 22:19:05 well you can do 22:19:22 PATH="/usr/lib/ccache:$PATH" dgl update-trunk 22:19:27 as the actual command 22:19:53 bash supports this syntax to set vars for the given command 22:20:11 otherwise you could possibly set it in the crontab like alex is saying, just not sure if that's used 22:21:12 there's also the rebuild script, so I'm not sure if there's a good central way 22:23:19 yeah if you can identify them that's better 22:23:21 oops 22:23:55 i see. i've prepended the command with that and the cache is now filling up 22:24:09 🙌 22:24:52 thanks for the help :) i'm going to stick that in crontab and try hourly trunk builds 22:26:48 although more frequent trunk builds means more binaries lying around 22:27:29 yeah, I believe cjr did not have prompting for trunk updates? 22:27:35 actually I'm not sure, it probably did 22:27:54 that is an option; force trunk updates 22:28:36 it's what cpo does already 22:30:27 also, ccache means setting aside a chunk of disk space just for the cache 22:32:20 i couldn't care less if my save was tranferred without my consent but i guess that might matter to some people 22:33:19 you can set the ccache cache size to about 1gb i think 22:33:54 it does look like it is going to come out to under 1gb for trunk 22:33:54 but i wonder, with a higher cache size will it cache multiple branches at once? 22:34:05 then i would not lose my entire cache every time i switched branches 22:35:00 the cache is LRU, so basically if you are building two branches and you have 2gb storage, both will be fully cached 22:35:18 if you build two branches and have 1gb storage there's probably no benefit from ccache :| 22:35:40 revised estimate: 700mb per branch 22:36:12 lol that is looking like a very accurate estimate so far 22:36:31 yeah that's how i was hoping it worked. i have it set to 5GB right now 22:36:52 but also if two branches are very similar, like maybe trunk and known-traps, then they would share much of the same cache files right? 22:36:59 so that wouldn't really take up the space of two whole branches, necessarily 22:37:25 even a single character change in any source file will mean ccache doesn't work. So I would say each branch is 100% unique 22:37:42 really? surely known-traps doesn't touch every single file 22:37:53 it touches header files that are #include'ed by almost every file 22:38:06 oh, yeah ok. :( 22:38:46 well, that's ok. i think i'm comfortable trading disk space for shorter compile times 22:39:12 i should be able to get away with minimal disk space since ttyrec are archived offsite 22:39:56 alexjurkiewicz: how quickly do the logfiles/milestone/scores files grow? i don't have a solution for those yet 22:42:38 score file remains negligible forever, since nobody wins :D 22:42:58 there is no "solution" per se except more disk space, since a whole lot of online play infrastructure expects those files to be permanent append only logs 22:43:44 that's what i thought 22:44:04 or rather that's what i was afraid of 22:44:35 for every 100k games, expect 400mb of logfile/milestone data 22:44:42 !lg * cpo 22:44:44 160051. DangYo the Firebug (L1 OpWn), slain by a bat on D:1 (minmay_arrival_semitransparent) on 2018-11-28 03:42:36, with 0 points after 48 turns and 0:00:29. 22:44:48 !lg * cjr 22:44:50 221828. wingeater17 the Cleaver (L7 HOFi of Yredelemnul), blasted by an orc wizard (puff of flame) on D:5 on 2018-11-27 23:06:55, with 723 points after 4726 turns and 0:08:37. 22:45:29 I was quite worried about those files but it's not a huge issue TBH. Nothing compared to ttyrecs 22:45:49 haha yeah, frankly i don't know what i'd do without object storage 22:47:41 trunk gave me 768MB in cache 23:10:42 Unstable branch on crawl.jorgrun.rocks updated to: 0.23-a0-480-g1869896531 (34)