00:50:41 alexjurk1: awesome, thanks for the work on that; I'll try to look at that tonight 01:01:12 Shift-x, shift-e stopped jumping to closest exclusion. 13https://crawl.develz.org/mantis/view.php?id=12287 by Yermak 01:20:44 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-3260-gdd6b58b38f 03:42:28 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-3260-gdd6b58b38f 03:55:33 Fork (bcadrencrawl) on crawl.kelbi.org updated to: 0.22.1-2805-g6ef1cdeb50 07:36:30 "Roots grabs something" reports offscreen neutral shambling mangrove. 13https://crawl.develz.org/mantis/view.php?id=12288 by Yermak 08:54:18 BerryKnight (L20 HuFi) ASSERT(!targets.empty()) in 'mon-cast.cc' at line 2385 failed. (No push space from (73,34)) (Vaults:2) 09:14:05 advil: oh nice, looks like you added basic t stats to the home page 10:00:28 -!- amalloy is now known as amalloy_ 10:06:42 -!- amalloy_ is now known as amalloy 10:22:32 -!- amalloy is now known as amalloy_ 11:19:00 yeah, wanted to check the winrate, and it turned out there was already code for this 11:20:20 t0.24's raw winrate (calculated in the same way) was 2.84%, so we're just about exactly on track 11:22:07 haven't seen a "no push space" crash in a while! 11:22:07 http://crawl.akrasiac.org/rawdata/BerryKnight/crash-BerryKnight-20200620-125418.txt 11:25:49 hm, looks like sealing should just fail in that case, but maybe it's getting it wrong because there is a free square that is part of the same door 12:22:39 man, seqdb is not tuned for limited memory 12:24:38 1946 postgres 20 0 193180 3804 3664 S 0.0 0.4 0:44.91 postgres 12:25:04 you're gonna have to be a bit more wordy 12:25:06 ... at least it's 1/3 the size of mysqld, but between them that's 4/5 of the memory :-/ 12:25:16 ah 12:25:28 the oom I was hitting is in seqdb itself on initial inload, not postgres 12:27:15 Oh! I appreciate this is a stupid variant question, but we run our own Sequell (as you know) and it works fine _except_ it only notices new games if it's restarted. The thing that might be odd about it is all the files it uses are local (because scoresd is already downloading them so why not). I don't suppose this is a known problem? 12:28:03 I'm not sure, I've never interacted with sequell before today 12:28:08 I know that cao scoring handles local files differently 12:28:33 and since sequell doesn't run on a play server, it might not have code to handle that? 12:28:53 ok, increasing the droplet to 2GB seems to have done the trick 12:29:26 if I wanted a quick solution tbh I'd expose my data on a web server and treat them as non-local 12:32:54 Yeah, or I might set it to download one file and see what happens; I just thought I'd ask on the offchance you said "oh, yes, you have to set option XYZ" 12:34:57 there might be an option for that? I can't remember the details of what cao does for local logfiles but it is exactly to handle them changing 12:35:17 but there might just not be an option for that for sequell since I don't know that it has ever been needed 12:35:24 Mmm. I mean, you put them in, it uses inotify to monitor them, all fine except nothing happens. 12:36:38 I assume you do have the local things set in sources? 12:36:50 though I don't know what that really does, since I notice that cao has it set 12:37:31 Yeah. No matter; it's our problem not yours. 12:43:07 basically what I do know is that wget handles incremental file downloads/restarts very well, and it needs something else for local files 13:41:35 !lg * 13:41:36 58716. me2cool8 the Caller (L2 FeSu), slain by a gnoll (a +0 spear) on D:2 on 2020-06-20 17:41:13, with 15 points after 662 turns and 0:01:45. 13:42:06 please don't yet connect any other bots to NotSequell, I want to assess basic load issues first (it's running in the smallest do droplet that would support its memory needs for now) 13:45:54 but feel free to use it in this channel or ##crawl-sequell 13:46:01 I guess PM should work too 14:04:53 elliptic the only thing I imported was learndb 14:05:16 so everything else is what was in the repo, which won't include nick mappings 14:05:26 (I assume) 14:06:14 !learn add NotSequell A temporary Sequell bot with only 0.25 milestones. Probably missing all sorts of other data! 14:06:14 NotSequell[1/1]: A temporary Sequell bot with only 0.25 milestones. Probably missing all sorts of other data! 14:06:52 !time 14:06:52 ugh, not even !time 14:07:12 if someone knows how to import all the command and keywrod stuff / where to get it from, I can do that 14:07:49 gong 14:07:49 $(weighted-rand 40:GONNNNG BOUMMMMG PTOANNNG PANG SHROANNG)! 14:07:53 lol 14:11:38 (no idea if that means something is misconfigured or if weighted-rand is missing a definition) 14:20:25 advil: yeah, I was just seeing what was imported so far 14:21:10 I got a bit confused when I saw that behaviors (like the gong thing) were around, but then I remembered that those are implemented somewhere in learndb for some reason (I think) 14:21:10 $(weighted-rand 40:GONNNNG BOUMMMMG PTOANNNG PANG SHROANNG)! 14:21:20 ahh 14:21:24 yeah, I was wondering that too 14:21:35 it would be nice to get commands, functions, and nicks 14:21:43 luckily enough cdo backs up learn.db in a crontab, so I could just pull that directly 14:22:23 commands/functions are usually in https://loom.shalott.org/userdef.html (though not in a great format for importing), but that seems to contain just a few things (presumably for testing Sequell) 14:22:48 I believe gammafunk said he had a backup of that page though 14:23:36 for nicks, iirc those were also available somewhere but I don't remember the URL 14:24:58 learndb claims they were at some point in https://loom.shalott.org/nickmap.txt 14:25:16 hm, looks like shallott is not generally in the wayback machine 14:25:29 but that's dead now, so unless someone has a backup for that then we are out of luck 14:26:19 ??behavior 14:26:19 I don't have a page labeled behavior in my learndb. 14:26:23 ??beh 14:26:23 I don't have a page labeled beh in my learndb. Did you mean: be, bh, bmh, eh, geh, veh, :beh:, bees, ehe. 14:26:28 ??:beh: 14:26:28 :beh:[1/25]: NotSequell: help ::: help[1/5]: !help (I need somebody) !help (not just anybody) Try https://github.com/crawl/sequell/blob/master/docs/listgame.md or https://github.com/crawl/sequell/blob/master/docs/learndb.md 14:26:33 (I'd tend to guess that greens.nark didn't actually delete all the old commands/functions/nicks and just moved them somewhere while testing/upgrading sequell) 14:26:45 yeah, I'm sure they are backed up somewhere 14:28:01 !fn fact (n) $(if (> $n 1) (* $n (fact (- $n 1))) 1) 14:28:02 Defined function: !fn fact (n) $(if (> $n 1) (* $n (fact (- $n 1))) 1) 14:28:15 .echo $(fact 5) 14:28:15 120 14:28:50 crucial 14:29:48 I'm pretty certain that the .db files themselves are exposed somewhere on shallot, so I could then at least get the testing version (instead of an empty db), but I don't know where 14:29:54 !lg * t xl>19 s=cikiller 14:29:55 1918 games for * (t xl>19): 1041x, 52x an orb of fire, 49x an Orb Guardian, 32x a titan, 32x an ancient lich, 21x a frost giant, 18x a pandemonium lord, 16x a Brimstone Fiend, 16x a greater mummy, 16x the Royal Jelly, 16x a lich, 15x a yaktaur captain, 15x an acid blob, 15x a tengu reaver, 14x a deep elf annihilator, 14x a stone giant, 13x a caustic shrike, 12x a vault warden, 12x a Hell Sentinel,... 14:31:24 !lg * t s=src x=sum(dur) o=sum(dur) 14:31:25 35665 games for * (t): 8437x cko [181d+20:02:30], 6638x cao [167d+10:53:57], 8198x cbro [149d+10:39:26], 4501x cxc [96d+18:07:21], 4076x cue [95d+5:39:32], 1788x cwz [37d+13:04:01], 1179x cpo [23d+16:46:21], 844x lld [13d+14:14:54], 4x cdo [0:13:19] 14:31:45 poor cdo 14:31:49 heh 14:32:00 I did get it fixed in the tourney scripts at least 14:33:00 anyway learndb/query seems to be working well (aside from a few learndb entries that use userdef functions I guess) 14:33:26 !learn set NotSequell[1 A temporary Sequell bot with only 0.25 milestones. Currently missing all commands, nicks, and keywords! 14:33:27 NotSequell[1/1]: A temporary Sequell bot with only 0.25 milestones. Currently missing all commands, nicks, and keywords! 14:44:40 !learn del :beh:[25 14:44:40 Permission db::beh: denied: nick advil not authorized 14:44:49 haha I guess I should make myself an admin 14:45:20 (something is broken with that :beh:) 14:52:52 !learn del :beh:[25 14:52:53 Deleted :beh:[25/25]: .echo echochamber (colon)(colon)(colon) echochamber (colon)(colon)(colon) continue 14:53:11 ??:group:admin 14:53:11 :group:admin[1/1]: |amethyst greensnark elliptic Kramin wheals read advil 14:53:19 oops, sry for pings 14:53:21 who is read? 14:58:32 wazuzu (L13 DsCj) Crash caused by signal #6: Aborted (Lair:3) 15:00:47 !crashlog 15:01:19 :'-( 15:03:20 !learn set NotSequell[1 A temporary Sequell bot with only 0.25 milestones and learndb. Currently missing all commands, nicks, and keywords! 15:03:20 NotSequell[1/1]: A temporary Sequell bot with only 0.25 milestones and learndb. Currently missing all commands, nicks, and keywords! 15:04:27 !learn edit NotSequell s|milestones|games/milestones| 15:04:28 NotSequell[1/1]: A temporary Sequell bot with only 0.25 games/milestones and learndb. Currently missing all commands, nicks, and keywords! 15:04:53 it seems like !time should be working, as it's a builtin 15:05:13 !time 15:05:40 lol the lack of an installed python is definitely not going to help 15:07:18 !time 15:07:18 Time: Jun 20, 2020, 07:06:54 PM, UTC. The 2020 0.25 tournament ends in 8 days, 0 hour, 53 minutes and 5 seconds. 15:07:24 in fact that one was just the ubuntu doesn't install /usr/bin/python by default gotcha 15:07:53 (also, how on earth did it know about the tournament??) 15:10:38 aha the repo knows 15:14:05 looks like that fix also caused t to be defined 15:14:09 !lg t 1 15:14:10 No games for t. 15:14:14 !lg * t 1 15:14:15 1/35786. vt the Shield-Bearer (L1 DDFi), slain by Natasha on D:2 on 2020-06-12 20:01:06, with 3 points after 67 turns and 0:00:18. 15:22:57 t was already defined actually 15:23:09 (I had tested it in PM) 15:24:15 oh, I guess that makes sense, sequell might have still been up when the new time was decided 19:35:58 Rollback on disconnect when on skills menu with 0 skills being trained 13https://crawl.develz.org/mantis/view.php?id=12289 by Midn8 20:40:24 advil: ok if I set up my bots to use NotSequell for now? 20:46:58 !tell gammafunk hi 20:46:59 gammafunk: OK, I'll let gammafunk know. 20:47:03 test 20:47:03 gammafunk: You have 1 message. Use !messages to read it. 20:47:06 !messages 20:47:06 (1/1) gammafunk said (8s ago): hi 20:54:55 gammafunk maybe go ahead and point cerebot at it and see how that goes? 20:54:55 gammafunk: do I remember correctly that you said you had a backup of some sequell userdef stuff? 20:55:58 elliptic: I do, but it's just the html dump; it does contain all command, function, keyword, and behaviour defs, but you'd have to parse it somehow 20:56:02 advil: ok, will do 20:56:43 what advil used for learndb was the actual (?) sqlite3 db of learndb, copied from shalott.org by cdo 20:56:49 so I think that was an easy direct import 20:56:53 at the very least we could manually copy specific useful commands as we desire them 20:57:03 e.g. !tstats 20:57:06 no idea what form the userdef stuff is natively 20:57:12 yeah, let me pastebin it then 20:57:24 or put it somewhere suitable, at least 20:57:48 it's a sqlite3 db 20:58:16 fairly simple schema 20:58:17 ok, well, maybe we could consider setting up a backup of that on cdo for situations like this in the future 20:58:52 yeah, definitely 20:58:52 looks like my copy is from october 2019, so it will have all the core stuff but be missing recent changes 20:58:52 it's probably not worth the effort to read in everything? 20:59:44 sure; we can just get a few base things manually and let others fill in what they want? 20:59:47 https://cdn.discordapp.com/attachments/205316046230388737/724065796170711160/user_defines.zip 21:00:01 I don't suppose you have nicks backed up also? 21:00:06 hrm 21:00:08 I might 21:00:23 what was the url for that 21:00:26 it also was public 21:01:00 it was a .txt file I think 21:01:17 looks like https://loom.shalott.org/nickmap.txt 21:01:27 alas I do not seem to have a copy 21:01:44 it's ok 21:02:01 I've always been of the opinion that's elliptic's nick was too elaborate 21:02:31 guess that's another good thing to back up; I assume that's also an sqlite3 db? 21:03:49 my nick fortunately not relevant for 0.25 games :P 21:03:51 hm, not actually sure 21:04:19 !nick advil advil hyperadvil 21:04:19 Mapping advil => advil hyperadvil 21:04:19 (thanks for the userdef backup) 21:04:19 looks like just a text file 21:04:19 called nicks.map 21:04:21 !cmd !flip .echo (╯°□°)╯︵ $(map (fn (w) (apply concat (map (fn (c) (if (< $c T) (lookup $c (flipcharhash1)) (lookup $c (flipcharhash2)))) (reverse (split "" $w))))) (join " " (reverse (split " " ${*:-┻━┻})))) 21:04:22 Defined command: !flip => .echo (╯°□°)╯︵ $(map (fn (w) (apply concat (map (fn (c) (if (< $c T) (lookup $c (flipcharhash1)) (lookup $c (flipcharhash2)))) (reverse (split "" $w))))) (join " " (reverse (split " " ${*:-┻━┻})))) 21:04:28 !flip Sequell 21:04:29 (╯°□°)╯︵ $(lookup ${c} $(flipcharhash2))$(lookup ${c} $(flipcharhash2))$(lookup ${c} $(flipcharhash2))$(lookup ${c} $(flipcharhash2))$(lookup ${c} $(flipcharhash2))$(lookup ${c} $(flipcharhash2))$(lookup ${c} $(flipcharhash1)) 21:04:44 heh I guess it needs the hash 21:06:12 your nick would be relevant if it turns out we implemented meleebug 2.0 in 0.25!! 21:06:28 advil gets to the important stuff first, I like it 21:07:09 !kw splat xl>=17 21:07:10 Defined keyword: splat => xl>=17 21:07:39 !flip Sequell 21:07:40 (╯°□°)╯︵ llǝnbǝS 21:07:43 there we go 21:10:35 !cmd !hs !lg $* max=sc sc<1000000000 21:10:35 Defined command: !hs => !lg $* max=sc sc<1000000000 21:11:14 nice, NotSequell is already in ##crawl-sequell 21:11:44 yeah, it was in the default config 21:13:28 interestingly I can message myself; guess that was "prevented" by an old beh 21:13:30 not important though 21:13:51 I will try to get what t userdef stuff important as I have time 21:13:58 s/important/imported/ 21:14:14 and s/userdef/userdef stuff I can/ 21:14:18 yeah, me too 21:15:18 looks like elliptic is doing some over pm (I left the existing admin group there, which includes him) 21:15:28 although doing !cmd over pm is still getting denied for me 21:15:33 well, actually I was trying to figure out whether adding userdef stuff over pm is possible :P 21:15:51 as an admin at least some of it should work 21:16:04 in principle I think it should be possible to modify this :acl: stuff to make !cmd usable by admins in pm 21:16:26 but I think !fn over pm might not be possible regardless 21:16:34 !fn double 21:16:35 No user function 'double' () 21:17:01 probably spamming ##crawl-sequell with !cmd and !fn defs is not actually an issue 21:17:06 oh, huh, wonder why it didn't print DENY messages for you 21:17:16 it's also in ##crawl-bot if you want some place even more secluded 21:17:21 Cannot define functions in PM. 21:17:33 is the response I got 21:18:01 and looking at the ACL documentation I don't see anything about !fn permissions 21:18:14 for some things it does print an error on permissions failure 21:18:17 in the log 21:18:49 yeah, for !nick I got Permission nick:elliptic denied: PM not authorized 21:19:00 right, but nothing in the log corresponds to that 21:19:23 whereas for some reason on every cerebot query I get something like `[2020-06-21 01:18:36] DENY:proxy Cerebot/msg: unauthorized nick` even though it seems to work 21:20:06 maybe something is not set up right with !RELAY 21:20:48 I suspect for !nick or !cmd or !kw in pm it is just a matter of editing 21:20:48 ??:acl:db:* 21:20:48 :acl:db:*[1/1]: ##crawl ##crawl-dev ##crawl-sequell ##csdc ##dieselrobin 21:20:48 in some way 21:21:30 seems plausible 21:22:02 ??:acl:proxy 21:22:02 :acl:proxy[1/1]: Kramell DIESELBOT 21:22:02 there's one example in the docks about *deying* PM editing, but it uses a syntax that I see only there 21:22:47 I don't really know how the proxy bot stuff works 21:23:00 *in the docs 21:24:12 advil: oh, what 21:24:31 it definitely should not allow defs of anything through Cerebot 21:24:39 I don't think it does 21:25:17 someone did try to do some !learn add stuff via there, looks like you, but I'm also seeing those for any regular command 21:25:23 ah I see what you mean 21:25:29 and yeah that was me testing that it didn't work 21:25:39 no idea about the regular queries generating that message 21:29:47 it seems to be working, so I'm going to ignore it I guess 21:30:39 !learn set :acl:db:* @admin ##crawl ##crawl-dev ##crawl-sequell ##csdc ##dieselrobin 21:30:41 :acl:db:*[1/1]: @admin ##crawl ##crawl-dev ##crawl-sequell ##csdc ##dieselrobin 21:31:47 hm, not sure that worked 21:31:50 !nick elliptic elliptic hyperbolic 21:31:52 Mapping elliptic => elliptic hyperbolic 21:32:19 I'm not 100% convinced that the @admin thing itself is working 21:32:21 maybe I should figure out whether I'm actually authenticated with nickserv 21:32:46 looks like it 21:40:49 oh, seems I can't edit commands in ##crawl-sequell 21:40:56 !cmd nchoice .echo $(let (combo (if (is_int ${1:--1}) (nth (int ${1:--1}) $(split | (replace " => " | (!kw nchoice)))) $1)) (let (wins (!lg * t won char=$combo fmt:"$n" stub:"0")) (if (and (> (int $wins) 0) (= ${1:-x} x)) (=nchoice.new) (concat (canonicalise char $combo) ": $wins $(plural win (int $wins)) || " (or (=lastnchoice.combostatus $combo) "No $combo games in progress."))))) 21:40:56 Permission cmd:!nchoice denied: nick gammafunk not authorized 21:41:31 !learn set :acl:db:* ##crawl ##crawl-dev ##crawl-sequell 21:41:33 :acl:db:*[1/1]: ##crawl ##crawl-dev ##crawl-sequell 21:41:42 !cmd nchoice .echo $(let (combo (if (is_int ${1:--1}) (nth (int ${1:--1}) $(split | (replace " => " | (!kw nchoice)))) $1)) (let (wins (!lg * t won char=$combo fmt:"$n" stub:"0")) (if (and (> (int $wins) 0) (= ${1:-x} x)) (=nchoice.new) (concat (canonicalise char $combo) ": $wins $(plural win (int $wins)) || " (or (=lastnchoice.combostatus $combo) "No $combo games in progress."))))) 21:41:42 Defined command: !nchoice => .echo $(let (combo (if (is_int ${1:--1}) (nth (int ${1:--1}) $(split | (replace " => " | (!kw nchoice)))) $1)) (let (wins (!lg * t won char=$combo fmt:"$n" stub:"0")) (if (and (> (int $wins) 0) (= ${1:-x} x)) (=nchoice.new) (concat (canonicalise char $combo) ": $wins $(plural win (int $wins)) || " (or (=lastnchoice.combostatus $combo) "No $combo games in progress."))))... 21:41:48 thanks 21:42:46 oh, I see 21:42:57 that's unfortunate for being able to give admins PM access without giving everyone PM access, I think 21:42:58 that was just a random guess but it seems I guessed right 21:43:23 maybe the entire entry has to be @admin? 21:43:51 I think if that was the case then non-admins wouldn't be able to edit it elsewhere 21:44:03 it's checking the categories separately basically 21:44:04 I mean for it to unpack @admin 21:44:12 no, I don't think that was the issue 21:45:43 basically I was hoping that the thing I set it to would mean "access permitted if user is @admin OR you are in ##crawl OR you are in ##crawl-dev, etc 21:46:39 it is clear that @admin was working since I was able to modify something here and gammafunk wasn't 22:11:09 advil: an issue for you to look into if you have the chance 22:11:21 !echo $(re-find "^a" "a") 22:11:31 .echo $(re-find "^a" "a") 22:11:31 wrong number of arguments (given 3, expected 1..2) in $(re-find ^a a) 22:11:54 this built-in seems to be failing due to something expected by NotSequell's regexp installation 22:12:06 according to https://github.com/crawl/sequell/blob/master/docs/commandline.md it should work 22:25:18 !learn edit NotSequell[1] s/ all / most / 22:25:19 NotSequell[1/1]: A temporary Sequell bot with only 0.25 games/milestones and learndb. Currently missing most commands, nicks, and keywords! 22:51:52 nikheizen (L12 FoCj) Crash caused by signal #15: Terminated (Lair:2) 22:53:34 hm 22:54:07 I did have to force install the perl re2 package, there was an error I just couldn't figure out 22:54:07 so it could well be broken 22:55:25 elliptic implemented a workaround for !nchoice, so all is well for now 22:55:47 !learn set :acl:db:* ##crawl ##crawl-dev ##crawl-sequell ##crawl-bot 22:55:48 :acl:db:*[1/1]: ##crawl ##crawl-dev ##crawl-sequell ##crawl-bot