00:00:01 basically just a variant of trj, kind of weird 00:00:11 yeah the speed and the gimmicks are notable differences 00:00:17 but it's funny how close their stats are otherwise 00:00:47 when he eats a wall it should turn the wall into an acid blob (durably summoned) 00:01:08 he shouldn't eat walls, he should have monster passwall 00:01:13 boom, emergent 00:01:15 lol 00:01:33 in fact retheme him into a slime stabber 00:01:35 I'm imagining dissolution passwallstabbing someone's deep elf while they try to rest 00:01:42 one-hit-KOs you 00:01:51 yeah how sick would that bee 00:01:54 *be 00:02:06 play some tome oozemancer and get some ideas from that? 00:02:13 no I haven't played tome 00:02:23 I dont play tome, I just know theres some class called an oozemancer 00:02:27 just have minmays mode saved in my little org file of roguelike notes 00:02:37 I am playing infra arcana now 00:02:57 ok, what if dissolution summoned caustic shrikes 00:02:57 recommended if you like difficult roguelikes btw 00:02:57 and rode on a caustic shrike 00:03:12 I think you could only put someone like sojobo on a shrike 00:03:27 would a bird really ride another bird into battle 00:03:35 definitely 00:03:38 seems inbirdmane 00:03:55 what if tiamat rode a shrike 00:04:13 what if all of the classed draconians were replaced with draconian riders that rode shrikes 00:04:29 if the pan lords all had mounts 00:04:29 like if cerebov rode this fire dragon 00:04:31 and you go gloorx on this bone dragon right 00:04:34 fire dragon is way too lame for cerebov 00:04:47 needs to ride a THE GIAGGOSTUONO 00:04:49 ok whatever a brimstone dragon, but like there's this shot 00:04:58 of them all riding their mounts 00:05:03 what would lom ride 00:05:05 need a fire simulacrum class of monsters and he could ride the fire-simulacrum version of it 00:05:07 then it pans to each one 00:05:08 no the THE GIAGGOSTUONO should ride cerebov 00:05:20 the last one would be mnoleg 00:05:29 sorry wheals_, cerebov is canon stronger than a THE GIAGGOSTUONO 00:05:33 and for the comedy shot the mount would be riding him 00:05:50 mnoleg just has a couple of shoulder-mounted wretched stars 00:06:02 I think it being upside down is better 00:06:06 not even cool enough for a proper mount 00:06:15 he's so wacky and weird the mount rides mnolege 00:06:23 yeah I guess 00:06:26 since in his tile he stands upside down 00:06:34 what if instead of a bone dragon gloorx rode two curse toes 00:06:37 ok maybe that's just what happens in the DCSS cartoon 00:06:45 but why not 10 curse toes 00:06:46 kids would like it 00:07:04 you got to give your cartoon adaptations amusing personalities like that 00:07:04 but back to my question, what would lom ride 00:07:21 they were all going to ride dragons 00:07:25 storm dragon 00:07:29 lom would ride a storm dragon 00:07:31 b o r i n g 00:07:32 see? it makes sense 00:07:46 does mnoleg upside-down ride a mutation dragon? 00:07:57 upside-down ride a quicksilver dragon maybe 00:08:04 again, only the cartoon 00:08:11 you can handle other media 00:08:13 ok 00:08:31 just asking for when I print up the first edition of the official dungeon crawl stone soup trading card game 00:08:36 wheals can handle thee twitter account 00:08:59 hellmonk, has anyone ever worked on something like a DCSS card game? 00:09:06 no idea 00:09:41 there was bcadren's "Dungeon of Nemelex Xobeh" 00:09:41 *Dungeons 00:10:01 his tavern meme project where he gave a release date of a specific day in the post 00:10:05 I should whip up a nandeck script that creates cards in Tabletop Simulator format 00:10:08 which was like 8 months in the future 00:10:09 did that actually exist or did it exist in the same sense that bcadren's other stuff existed 00:10:18 yeah more the latter 00:10:22 lol 00:10:24 classic 00:10:43 it was just funny because in this big long effort post he gives an exact "day of release" 00:10:52 I guess that made it feel more like a real project 00:11:13 I will contribute to the DCSS TCG if anyone has actual ideas for cool ass cardgame mechanics 00:11:31 that are implementable w/o serious effort on my part 00:11:42 hellmonk, I'll screw around with nandecking some cards, just to have something to play with in TTS 00:11:46 I never really got into any of the card games, but it'd be neat if people drew cool DCSS art for them 00:12:04 not right away, mind you. but I'll toss it on the heap 00:12:25 make a dissolution card if you want hellmonk on the project 00:12:28 yeah I imagine we'd start with just the sprite work and then hire an artist once we have dethroned hearthstone as the number 1 tcg 00:12:33 step 1, make cards. step 2, figure out a game 00:12:56 ok so each player has a food clock 00:13:10 and the goal is to starve them by making them fight monsters 00:13:15 how are you going to have the idea of "LOS" 00:13:22 riddle me that, so-called game designers 00:13:24 gotta bring back item destruction in the form of discarding cards 00:13:27 los is just not a core mechanic 00:13:32 unlike food 00:13:49 dcss's window-LOS is like the most central feature of the thing 00:13:57 it's why linley made crawl! (sort of) 00:14:05 ...gods? 00:14:12 those came a lot later 00:14:15 cards could have a food value. so you could discard cards to pay food costs, which should just be for special moves 00:14:21 lol 00:14:22 well they came in v. 2 00:14:27 1.x did not have gods 00:14:37 but yeah he was super inspired by all those classic RPGs and games on like C64 00:14:41 that had a viewport area 00:14:54 compare this to nethack which displays the entire level map all at once 00:14:59 but it's a lot less personal that way 00:15:37 not that those games don't also have a LOS, but the concept and the visuals of it were central to him for what DCSS was going to be about 00:15:52 and it is one of the coolest features of DCSS, how los obstruction works 00:15:59 all the tactical goodness etc 00:16:14 ty for serious responding to my irony 00:16:20 so yeah card game must have LOS or you don't get my endorsment 00:16:27 pretty sure he did not have the :SS part in mind 00:16:28 ok, LOS is in 00:16:31 :p 00:16:46 dungeon crawl: sight soup 00:18:03 sorry, I have to say LDC not DCSS otherwise geekosaur will be made at me 00:18:13 *mad 00:18:42 neh, we already determined you're the top crab :p 00:18:51 line of sight's dungeon crawl 00:19:08 who is we geeokosaur, the other irc dinosaurs? 00:19:11 do you all have a channel 00:19:35 in 5 years the tcg will be ahead of dcss development and you'll have to pull your new monsters from our cardgame 00:20:06 will the cards be scannable so we can do that easilly? 00:20:40 scannable? 00:20:49 you gotta get the cards into the scanner's LOS first 00:21:13 how are we going to get the data from the cards, johnstein? 00:21:42 I'm going to pull the graphics from craw source for prototyping 00:21:51 could share resources 00:22:07 oh, sorry, you're trying to have a non-joke conversation 00:22:35 heh 00:22:56 I take card game dev super serious 00:23:35 well as long as I can become a dcss card game greaterplayer, I'll be happy 00:23:36 i remember seeing a demo or a trailer or something for a game on the playstation, where it used a camera accessory to scan bar codes on physical cards you had to collect so that you could use them in game 00:24:13 I feel like that may also have been done before then on gameboy, or something 00:24:20 probably 00:29:49 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.21-a0-210-gbfa5647 (34) 00:37:38 advil: basically amiibo 00:37:41 *ama 00:37:42 lloy 00:41:06 llamalloy 00:56:31 interesting, my client doesn't treat llamalloy as a highlight 01:56:18 Windows builds of master branch on crawl.develz.org updated to: 0.21-a0-210-gbfa5647 02:03:56 -!- mumra1 is now known as mumra 02:40:34 !apt mp 02:40:35 MP: Dg: 2!, DE: 2!, Fe: 1, VS: 1, Te: 1, Sp: 1, Fo: 0, Ba: 0, Ko: 0, Ds: 0, Mf: 0, Vp: 0, Hu: 0, HO: 0, Ce: 0, HE: N/A, Dr: 0, Gr: 0, Ha: 0, Na: 0, Op: 0, Og: 0, Gn: 0, Mu: 0, DD: 0, Gh: -1*, Tr: -1*, Mi: -1* 02:50:39 !lg * de won x=avg(int) current urune=3 02:50:40 49 games for * (de won current urune=3): avg(int)=36.69 02:54:08 Monster database of master branch on crawl.develz.org updated to: 0.21-a0-210-gbfa5647 02:56:14 |amethyst: are you around to help with some random logic? 02:57:40 |amethyst: I'm writing a PR to hellcrawl to change how max MP is calculated, so it doesn't use spellcasting any more. I was thinking about using xl + int*(xl/27). But how can I do this in a stable rounded manner? 03:01:34 ah, I guess I can multiply by 1000 and then round at the end 03:01:55 the easiest thing to do is to just write (int*xl)/27 03:02:27 i don't think multiplying everything by 1000 and dividing at the end does you any good, because the number you divide by will be 1000 times as large 03:03:53 also make sure to consider what will happen with stat drain. the player's int can be negative 03:04:07 but MP probably shouldn't be 03:06:48 true, i should cap int at zero 03:07:07 well it sounds like this method will introduce a lot of breakpoints then 03:07:52 i wonder if rounding rather than relying on integer flooring would help 03:08:47 you'd have the same number of breakpoints, shifted by 0.5 03:22:44 Unstable branch on crawl.beRotato.org updated to: 0.21-a0-210-gbfa5647 (34) 03:57:18 I wonder if I should try some C++ bootcamp refresher course, or just keep futzing along. I'm able to make forward progress, but I'm having to relearn a lot of things (and learning a lot of patterns and shorthands I've never seen before). 03:57:51 thank trog for an IDE that I can use to explore the code to suss things out 05:21:28 -!- amalloy is now known as amalloy_ 06:09:05 New branch created: pull/601 (5 commits) 13https://github.com/crawl/crawl/pull/601 06:09:05 03johnstein02 07https://github.com/crawl/crawl/pull/601 * 0.21-a0-204-g68557be: Split out some logic into a new function 10(3 hours ago, 2 files, 15+ 9-) 13https://github.com/crawl/crawl/commit/68557be58f43 06:09:05 03johnstein02 07https://github.com/crawl/crawl/pull/601 * 0.21-a0-205-g07e99b5: Track follower transiting start time 10(2 hours ago, 2 files, 2+ 0-) 13https://github.com/crawl/crawl/commit/07e99b55d895 06:09:05 03johnstein02 07https://github.com/crawl/crawl/pull/601 * 0.21-a0-206-g1a4f2e2: Use a different approach for generalising monster_update 10(71 minutes ago, 2 files, 41+ 36-) 13https://github.com/crawl/crawl/commit/1a4f2e295521 06:09:05 03johnstein02 07https://github.com/crawl/crawl/pull/601 * 0.21-a0-207-g34e52d3: Update lost monsters after placing them 10(69 minutes ago, 1 file, 10+ 1-) 13https://github.com/crawl/crawl/commit/34e52d343643 06:09:05 03johnstein02 07https://github.com/crawl/crawl/pull/601 * 0.21-a0-208-g2590b96: add some debug messages 10(14 minutes ago, 2 files, 5+ 0-) 13https://github.com/crawl/crawl/commit/2590b962dfe1 08:40:03 E'x'amine targets cycle should embrace monsters behind glass. 13https://crawl.develz.org/mantis/view.php?id=11191 by Yermak 08:46:52 !kw xpinfo vlong>=0.21-a0-210-gbfa5647 08:46:53 Defined keyword: xpinfo => vlong>=0.21-a0-210-gbfa5647 08:47:01 !lg * xpinfo won 08:47:02 2. alcopop the Sorcerer (L26 SpEn of Ashenzari), escaped with the Orb and 3 runes on 2017-08-21 12:03:50, with 1459806 points after 91229 turns and 3:48:14. 08:47:07 !lg * xpinfo won -log 08:47:08 2. alcopop, XL26 SpEn, T:91229: https://crawl.project357.org/morgue/alcopop/morgue-alcopop-20170821-120350.txt 08:47:58 !lg * xpinfo won -log -2 08:47:59 1/2. MackTheKnife, XL27 BaEE, T:97108: http://crawl.xtahua.com/crawl/morgue/MackTheKnife/morgue-MackTheKnife-20170821-053747.txt 08:49:47 !lm * begin xpinfo 08:49:48 179. [2017-08-21 12:47:27] Kidiri the Firebug (L1 DEFE) began the quest for the Orb on turn 0. (D:1) 08:50:03 !lm * begin xpinfo !won max=xl 08:50:05 150. [2017-08-20 19:10:33] raiza the Chopper (L1 MiBe of Trog) began the quest for the Orb on turn 0. (D:1) 08:50:16 !lm * begin xpinfo !won max=lg:xl 08:50:29 150. [2017-08-21 08:46:32] MackTheKnife the Digger (L1 GrWn) began the quest for the Orb on turn 0. (D:1) 08:50:43 &dump mackthknife 08:50:44 No milestones for mackthknife. 08:50:53 &dump macktheknife 08:50:54 http://CRAWL.XTAHUA.COM/crawl/morgue/MackTheKnife/MackTheKnife.txt 08:51:07 !lm * begin xpinfo !won max=lg:xl -2 08:51:09 149/150. [2017-08-21 09:45:23] p0werm0de the Poker (L1 DgWn) began the quest for the Orb on turn 0. (D:1) 08:51:19 &dump o0werm0de 08:51:20 No milestones for o0werm0de. 08:51:24 &dump p0werm0de 08:51:25 http://CRAWL.XTAHUA.COM/crawl/morgue/p0werm0de/p0werm0de.txt 09:41:32 it the github website down for anyone else? 09:42:07 was working in the last 10 minutes but it is definitely taking some time reloading right this moment 09:42:22 https://status.github.com/ reports "elevated error rates" 09:42:49 yeah I can't seem to pull either 09:48:22 "We are investigating reports of connectivity issues to GitHub.com and related services." 10:53:24 Not all unrands listed in notes 13https://crawl.develz.org/mantis/view.php?id=11192 by argonaut 12:30:58 This is the wording for nightstalker 1: 12:31:00 You are slightly more attuned to the shadows. 12:31:03 More than what? 12:34:25 normal, which will vary based on various things (for example Dith worshippers) 12:34:52 maybe Yred too 12:41:15 I wouldn't say being Dith or Yred worshiper is a norm. 12:42:36 <|amethyst> Wouldn't "attuned to the shadows" mean you can see *further* in the darkness? 12:42:36 |amethyst: You have 1 message. Use !messages to read it. 12:43:37 <|amethyst> johnstein: you probably noticed the crashes already 12:43:47 <|amethyst> johnstein: in the abyss tests 12:43:56 <|amethyst> johnstein: https://travis-ci.org/crawl/crawl/jobs/266742375 for example 12:44:10 I don't think so, but in any case wording is bad imo. 12:44:19 Yermak, "normal" is relative to what your character is without nightstalker, is the point 12:44:50 Normally character isn't attuned to the shadows. 12:44:51 actually don't even need "relative to" there 12:44:55 soigh 12:45:03 'more' is excess here 12:45:07 ok, yes. we pretend Yred/Dith don;t exist and talk only about the most common case 12:45:15 then the message is wriong for cases you have determined don't matter 12:45:54 and then a Dith worshipper asks if nightstalker 1 will *reduce* their attunement, because of your "correct" wording 12:46:08 Dude, umbra is absolutely another stuff. 12:48:01 <|amethyst> Ignoring the "more" or lack thereof, "attuned to the shadows" doesn't say what's going on 12:49:48 <|amethyst> for example, geekosaur is bringing up Yred and Dith, which has little to do with the Nightstalker mutation 12:50:08 yeh, I'm not sure what the cases are here 12:50:18 but I am reasonbably sure nightstalker is not the *only* source 12:50:34 <|amethyst> it's the only non-item non-spell source 12:51:01 so it should pretend the other sources don't exist. 12:51:11 seems like what both of you are saying 12:51:25 Nithstalker mutation is the source of LoS reduction, whereas umbra (Yred & Dith) is the source of an aura, which doesn't affect LoS radius. 12:51:31 <|amethyst> I am saying the description should try more to convey the mechanics and less to be flavourful 12:52:18 <|amethyst> "The vision radius of you and monsters is reduced (-1 LOS)" or such 12:52:40 <|amethyst> or "slightly"/"somewhat"/"greatly" reduced 12:53:40 <|amethyst> though there is also a secondary effect on stealth, so I guess something like "(-1 LOS, Stlth+)" 12:54:10 I'd keep some flavour: Your attunement to shadows slightly reduces your line of sight. 12:54:20 <|amethyst> unfortunately Thin Skeletal Structure already uses Stealth/Stealth+/Stealth++ despite it providing only half a pip per level 12:54:49 <|amethyst> Yermak: it should make clear that the effect is symmetric, otherwise it sounds like all downside 12:54:58 agreed 12:55:42 <|amethyst> "Your attunement to shadows slightly reduces your vision and visibility (LOS-, Stealth+)" but that's too long 13:04:52 |amethyst: actually I didn't notice the abyss crashes yet. I suppose I should be more vigilant 13:05:06 The longest description now seems to be 13:05:07 "When seriously injured, your bleeding wounds create very thick armour. (AC +" 13:05:32 plus space, two digits and bracket 13:05:44 which is 80 13:06:10 <|amethyst> yeah, 80 is a pretty hard limit there 13:11:13 Darkness around you slightly reduces visibility. (LOS-, Stealth+) 13:12:22 Barachi have "Shadows flee at your approach. (+LOS)" 13:12:37 weird. github mobile site doesn't show Travis status on the PR page 13:25:49 Unstable branch on crawl.akrasiac.org updated to: 0.21-a0-210-gbfa5647 (34) 13:28:10 hm looks like I broke FULLDEBUG 13:33:27 -!- amalloy_ is now known as amalloy 13:34:16 well travis does a fulldebug build 13:34:22 do you mean for something that's been pushed? 13:34:48 <|amethyst> johnstein: that's because we only run full tests in debug mode 13:34:48 if you had broken it, travis should have build failures 13:35:04 oh I see, didn't read scrollback 13:37:31 gammafunk: yea. I haven't read the logs but |amethyst said it was an abyss test, which I haven't tested yet 13:38:19 but I'm nearly positive the crash is because I didn't update the banish monster code with my new variable 13:38:44 since I couldn't get the shafted scenario working yet 13:45:37 the PR is a definite WiP 14:14:33 -!- mumra1 is now known as mumra 15:22:57 Fumpy (L26 GrFi) (Vaults:4) 15:37:14 huh. wonder what that crash is about. apparently pressing ?# caused some sort of memory allocation failure when trying to format/print the dump 15:37:23 &dump fumpy 15:37:25 http://crawl.akrasiac.org/rawdata/Fumpy/Fumpy.txt 15:38:20 !crash 15:38:21 17632. Fumpy, XL26 GrFi, T:68584 (milestone): http://crawl.akrasiac.org/rawdata/Fumpy/crash-Fumpy-20170821-192256.txt 15:38:36 but the dump looks fairly innocuous. no 500KB comments from tiles chat 15:39:09 looks like the crash was maybe 15:39:21 in that ?# menu (or however you display the dump) 15:39:26 oh hm 15:39:41 easy to repro for me locally on latest git 15:39:49 but i couldn't repro from last week 15:40:00 (only after updating git and recompiling did it repro) 15:40:05 can you look at a backtrace? 15:40:25 well there's one in the crash. but i can load it in gdb and see if there's anything interesting 15:41:42 it's not crashing immediately; it's in some infinite loop appending stuff to the dump and eventually crashes when it runs out of memory 15:43:00 amalloy: can you try a bisect before... 15:43:03 !kw xpinfo 15:43:03 Keyword: xpinfo => vlong>=0.21-a0-210-gbfa5647 15:43:09 207 15:43:13 so 206 15:43:33 what do you mean, before that? 15:43:41 at 206 or before 15:43:47 before my XPinfo stuff 15:43:58 since that does affect chardumps (including default ones) 15:44:17 %git bfa5647 15:44:17 07gammafunk02 * 0.21-a0-210-gbfa5647: Fix turn tracking for turns_by_place and xp_by_level dump tables 10(2 days ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/bfa5647dee80 15:44:21 %git bfa5647~4 15:44:21 07|amethyst02 * 0.21-a0-206-g7d95abb: Checkwhite. 10(2 days ago, 1 file, 8+ 8-) 13https://github.com/crawl/crawl/commit/7d95abb50f06 15:44:32 so 7d95abb is good 15:44:52 oh, hmmmm 15:44:53 &versions 15:45:05 CAO: 0.21-a0-210-gbfa5647, CBRO: 0.21-a0-210-gbfa5647, CDO: 0.21-a0-2-ge494091, CJR: 0.21-a0-210-gbfa5647, CPO: 0.21-a0-210-gbfa5647, CUE: 0.21-a0-210-gbfa5647, CWZ: 0.21-a0-150-gc243083, CXC: 0.21-a0-210-gbfa5647, LLD: 0.21-a0-142-gdd972a7a06 15:45:14 alright, looks fine 15:45:24 I swear I saw one at 207 though, which would be worrying 15:46:02 %git bfa5647~3 15:46:02 07|amethyst02 * 0.21-a0-207-g4a9efaf: Remove unused descriptions. 10(2 days ago, 16 files, 0+ 64-) 13https://github.com/crawl/crawl/commit/4a9efaf00671 15:46:07 ah ok that's fine then 15:46:17 %git bfa5647~2 15:46:17 07gammafunk02 * 0.21-a0-208-g9cb1729: An xp_by_level XP tracking table for morgues 10(2 days ago, 12 files, 352+ 36-) 13https://github.com/crawl/crawl/commit/9cb1729a330e 15:46:30 turns out it started at 208, so 206 or 207 would be fine 15:47:24 https://gist.github.com/amalloy/7c4a6e930011dbd9fbaf702e272db3cd 15:47:36 gammafunk: it's failing to word-wrap your ascii table i think 15:47:53 aha 15:48:05 hrm, ok, maybe we need to simply trim whitespace 15:48:10 learding whitespace 15:48:24 it has like 14 characters for the place name but I don't think that many are needed 15:48:40 but I also don't know what the max is 15:49:20 gammafunk: you could compute it by looping over all of the place names before you start formatting strings? 15:49:35 well that's probably not necessary 15:49:41 just put a cap on it in the format string 15:49:48 right now it's %14s I think 15:50:49 %14.14s or something? 15:52:30 so right now it's 66 + 15 characters 15:52:37 what limit does it need to be, 80? 15:53:16 if 80, we can just decrease each of those whitespace strings by one char, and change %14s to %13s 15:53:45 I'll try it, let me just reproduce the crash first 15:54:33 we should fix the string formatter so it doesn't consume all memory and then crash when there's a long string with no spaces 15:55:19 yes, was just going to say 15:55:23 but that's not my department 15:55:27 I'll clean up this table though 15:55:53 i'll see if it's easy to improve the root cause 15:56:53 <|amethyst> wordwrap_line looks like it's the problem 16:03:16 <|amethyst> hm... actually 16:03:25 <|amethyst> from that gist 16:03:27 <|amethyst> #1 0x00000000011e36a9 in wordwrap_line ( s=' ' , "+---------+---------+-------+---------+---------+-------+---------", width=0, tags=false, indent=true) at stringutil.cc:205 16:04:01 <|amethyst> but formatted_scroller::add_raw_text shouldn't even be calling wordwrap_line with a wrap_col of 0 16:04:47 <|amethyst> that call is inside if (wrap_col > 0) 16:06:03 <|amethyst> my testing shows wordwrap_line does seem to handle words longer than the line width 16:06:13 <|amethyst> but probably not if the width is zero 16:06:31 03gammafunk02 07* 0.21-a0-211-gdd1ebd0: Decrease the width of the xp_by_level morgue table 10(5 minutes ago, 1 file, 5+ 5-) 13https://github.com/crawl/crawl/commit/dd1ebd0fd3e6 16:07:01 <|amethyst> wordwrap_line should have an assert(width != 0) 16:07:10 > 0 16:07:20 <|amethyst> yeah, probably :) 16:07:40 well, i have a gdb session sitting at the top of wordwrap_line with that input string. interestingly, width is 80 16:07:45 i wonder when it becomes 0 16:08:16 <|amethyst> oh 16:08:21 ah, indeed 16:08:30 <|amethyst> I guess that's just the normal "subtract the used width" thing then 16:08:31 <|amethyst> hmm 16:08:54 still not a bad assert to add, even if it's not the problem this time 16:09:02 <|amethyst> I was testing "Thisisverylongcomparatively" 16:09:09 <|amethyst> with a width of 10 16:09:34 <|amethyst> which it split into ret = "Thisisvery" and instr = "longcomparatively", as expected 16:11:51 <|amethyst> when does the crash happen? on game over? 16:11:51 <|amethyst> I guess it doesn't show up in ?# 16:11:54 it does in ?# 16:12:00 in wizmode 16:12:32 <|amethyst> oh 16:12:43 <|amethyst> my build is a couple of versions old, that would do it 16:12:47 heh 16:12:56 yeah, crawl freezes when you do ?# in-game 16:13:13 <|amethyst> hm 16:13:17 don't pull trunk though 16:13:20 <|amethyst> I can't even start crawl, let me try make clean 16:13:21 since that has a fix 16:13:37 or maybe do I'm not sure what you're doing, but if you want to see the crash it's in 210 16:13:38 <|amethyst> yeah, I'm doing 0.21-a0-210-gbfa5647 16:14:51 <|amethyst> Value returned is $1 = ' ' 16:15:36 i think i see what's going on 16:16:10 or, hm, maybe no 16:17:11 I guess I should do rebuilds, but I have to step out for a while; if either of you push a fix or stop working on said fixes and can start rebuilds, that would be appreciated 16:17:48 and if not I'll get to it in a few hours 16:18:16 so wordwrap_line sees that it can't fit the whole thing on one line, which is true. therefore, it asks _get_indent to figure out how much of the input string was indentation vs meaningful characters 16:18:28 _get_indent says "these 15 spaces at the front are indentation" 16:19:43 then we erase all "trailing" spaces from the string, except afaict it is actually replacing leadnig spaces 16:20:20 <|amethyst> it's trailing spaces from the string that is going to be returned 16:20:25 i see 16:20:37 well the string that was going to be returned is 15 spaces 16:21:20 <|amethyst> that in itself is the problem 16:21:41 <|amethyst> we shouldn't break at leading space 16:22:22 that sounds reasonable to me 16:22:43 |amethyst: https://gist.github.com/amalloy/f2645aab15418c27759cf1d863929cf5 is some gdb information when things start to go wrong, if that is useful in fixing the code 16:25:01 i've been called away for work but i'll leave the gdb session open 16:28:38 <|amethyst> I. WITH indent=true: 1. 'Th...': instr = 'longcomparatively', res = 'Thisisvery' 2. ' Th...': instr = ' Thisisverylongcomparatively', res = '' 3. ' Th...': instr = ' Thisisverylongcomparatively', res = ' ' 16:28:42 <|amethyst> II. WITH indent=false: 1. 'Th...': instr = 'longcomparatively', res = 'Thisisvery' 2. ' Th...': instr = 'Thisisverylongcomparatively', res = '' 3. ' Th...': instr = 'Thisisverylongcomparatively', res = ' ' 16:28:46 <|amethyst> doh 16:29:24 <|amethyst> the dangerous case is when indent = true and the string starts with a space, because then the string is unmodified by wordwrap_text 16:30:23 <|amethyst> there is a minor bad case when indent = false and the string starts with a space, which results in an empty or blank line (but the space is removed so it will work on the next call) 16:31:23 <|amethyst> okay, I think this fix works reasonably 16:31:41 <|amethyst> I. WITH indent=true: 1. 'Th...': instr = 'longcomparatively', res = 'Thisisvery' 2. ' Th...': instr = ' ylongcomparatively', res = ' Thisisver' 3. ' Th...': instr = ' rylongcomparatively', res = ' Thisisve' 16:31:45 <|amethyst> II. WITH indent=false: 1. 'Th...': instr = 'longcomparatively', res = 'Thisisvery' 2. ' Th...': instr = 'ylongcomparatively', res = ' Thisisver' 3. ' Th...': instr = 'rylongcomparatively', res = ' Thisisve' 16:32:41 <|amethyst> yeah, now it just wraps stupidly, but doesn't crash 16:33:03 <|amethyst> well, I guess as non-stupidly as could be expected 16:38:51 -!- Amnesiac_ is now known as Amnesiac 16:44:01 <|amethyst> oh, I guess I need to test one more special case 16:46:50 <|amethyst> hm, okay, the special case triggers one of my new asserts (and would cause an infinite loop without it) 16:47:14 <|amethyst> namely, if the indentation itself is wider than the wrap column 16:51:32 <|amethyst> okay, fixed. This will not keep the indentation in that case, but that's kind of hard to do when the indentation itself wraps... 16:52:24 <|amethyst> %git 16:52:24 07gammafunk02 * 0.21-a0-211-gdd1ebd0: Decrease the width of the xp_by_level morgue table 10(51 minutes ago, 1 file, 5+ 5-) 13https://github.com/crawl/crawl/commit/dd1ebd0fd3e6 16:57:46 -!- Menche_ is now known as Menche 16:58:01 03|amethyst02 07* 0.21-a0-212-g1ea99b4: Don't break when wrapping indented too-long first words. 10(15 minutes ago, 1 file, 19+ 2-) 13https://github.com/crawl/crawl/commit/1ea99b43009a 16:58:45 <|amethyst> I guess I should have said "hang" rather than "break", since the latter already has a meaning in this context 17:03:30 <|amethyst> amalloy: thanks for the help debugging 17:09:14 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-212-g1ea99b4 (34) 17:14:17 <|amethyst> Napkin: re the captcha... what if we just asked the user to enter the characters in reverse order? 17:15:34 <|amethyst> Napkin: seems like that would be enough to defeat bots who aren't even smart enough to reload the page 17:27:48 03amalloy02 07* 0.21-a0-213-gae198ad: Simplify code for making mimics cackle 10(6 days ago, 1 file, 2+ 4-) 13https://github.com/crawl/crawl/commit/ae198adf042e 17:27:54 New branch created: pull/602 (1 commit) 13https://github.com/crawl/crawl/pull/602 17:27:54 03amalloy02 07https://github.com/crawl/crawl/pull/602 * 0.21-a0-213-g9d5f42f: Remove a bunch of #if 0 clauses 10(9 minutes ago, 9 files, 0+ 83-) 13https://github.com/crawl/crawl/commit/9d5f42fc0e71 17:30:38 <|amethyst> amalloy: Hm, I'd keep some of those TODO comments 17:30:47 <|amethyst> amalloy: e.g. // TODO enne - fix menus so that mouse wheeling doesn't easy exit 17:30:54 <|amethyst> oh 17:31:05 <|amethyst> I guess that one is only relevant if you set 0 == 1 17:31:35 everything i removed is at least 5 years old, and most of it was from 2010. i don't think they're adding any real value 17:43:19 The build has errored. (master - 1ea99b4 #8649 : Neil Moore): https://travis-ci.org/crawl/crawl/builds/266952313 18:02:52 Unstable branch on underhound.eu updated to: 0.21-a0-213-gae198ad (34) 18:09:09 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-213-gae198ad (34) 18:43:29 !tell gammafunk fyi the xp table looks a bit silly if you look at it before you've killed any monsters. a couple entries read "-nan" 18:43:29 amalloy: OK, I'll let gammafunk know. 18:44:30 <|amethyst> you can get nans in the kills by level table too 18:44:41 <|amethyst> if you have never had an ally, for example 18:52:59 <|amethyst> oh 18:53:34 <|amethyst> amalloy: apparently there's already a check for " nan " but it doesn't handle " -nan " 18:58:17 amalloy: yeah, it replaces invalid numbers with nan, but I think it's adding a sign to those as well 18:58:17 gammafunk: You have 1 message. Use !messages to read it. 18:58:31 I'm not sure what the long-term future of this table will be 18:58:54 if we remove monster spawns and implement a new spawning system, it either will no longer be useful or we should move parts of the kill table into it 18:59:07 or maybe just fully merge the two, taking the fields we actually care about 19:04:54 woo, new spawning system 19:05:02 <|amethyst> hm 19:05:10 <|amethyst> I don't suppose anyone has MSVC handy? 19:05:40 only msys2, sorry 19:06:02 <|amethyst> oh, godbolt 19:06:28 good new expletive 19:06:50 Lasty: I'm not sure if you have time to work on any crawl things, but we sort of determined that spawn XP is 10-15% of total XP gained on many levels, some with much higher amounts, and the data is still coming in 19:07:11 but the thought was to replace XP-granting spawns with additional monsters generated awake, possibly controlling for placing those near stairs 19:07:28 and then the question is what to do as a "forward progress/movement off stairs" clock 19:07:38 MPA had suggested maybe just summons 19:07:50 gammafunk: spawn XP is only 10-15%? Dang. 19:07:59 yeah 19:08:04 !lg * won xpinfo 19:08:05 8. kupotwopie the Slayer (L27 DsVM of Gozag), escaped with the Orb and 3 runes on 2017-08-21 22:56:00, with 1795479 points after 65150 turns and 7:43:36. 19:08:08 !lg * won xpinfo -log 19:08:09 8. kupotwopie, XL27 DsVM, T:65150: http://crawl.akrasiac.org/rawdata/kupotwopie/morgue-kupotwopie-20170821-225600.txt 19:08:11 see table at end of that 19:08:22 doh 19:08:31 not that one 19:08:31 seems to be a transfer 19:08:33 so what is "spawn" here? 19:08:44 non-orbrun spawned monster 19:08:50 <|amethyst> after levelgen 19:08:52 vs. generated monster 19:09:02 I'm a big fan of the "post-gen spawns are summons" plan, if that what you mean 19:09:34 bezotted summons!!!! 19:09:46 !lg * won xpinfo -2 -log 19:09:47 8/9. kupotwopie, XL27 DsVM, T:65150: http://crawl.akrasiac.org/rawdata/kupotwopie/morgue-kupotwopie-20170821-225600.txt 19:09:52 (not an actual good idea, ymmv) 19:10:03 <|amethyst> gammafunk: same game, someone went and won since your last !lg 19:10:09 dammit 19:10:13 <|amethyst> !lg * won xpinfo -1 -log 19:10:13 stop mocking me, players 19:10:14 9. Yermak, XL27 DsSu, T:72064: http://crawl.xtahua.com/crawl/morgue/Yermak/morgue-Yermak-20170821-230924.txt 19:10:14 !lg * won xpinfo -log 19:10:15 9. Yermak, XL27 DsSu, T:72064: http://crawl.xtahua.com/crawl/morgue/Yermak/morgue-Yermak-20170821-230924.txt 19:10:18 yermak... 19:10:23 I got yermakked again, guys 19:10:40 ok that looks legit 19:10:57 that's a crazy fast zot though 19:11:06 but yeah, you see there 10% 19:11:31 it increases with turncount on the level, as you'd expet 19:11:33 *expect 19:11:37 we're pretty confident that the logging is right, I assume 19:12:02 yeah 19:12:08 you can see a number of levels with 30% 19:12:18 but those correspond with levels that took a bit longer 19:12:34 well, not always 19:12:49 spider:1 for some reason had a lot of XP from spawns but not all that many turns 19:13:06 maybe it's just variance since these numbers of monsters are also pretty small 19:13:43 but in any case there's still probably too much XP in the game, so I don't think we're going to have any problems at all if we throw in a reasonable number of extra monsters 19:13:55 yeah, agreed 19:13:56 it will probably be a difficulty increase I guess, since more monsters active at once 19:14:18 spawns will tend to generate after many monsters are killed, so they will be more likely to be found solo 19:15:20 I'm not sure I caught that part: what prevents the spawns from being found solo in the new system? Clustering around stairs? 19:16:16 oh, maybe language was a bit unclear there; by "spawns" I mean "monsters that would have been spawns in the old system, but will now be generate_awake monsters at level gen, since we've removed XP-granting spawns" 19:16:42 elliptic had proposed replacing that XP lost from removing XP-granting spawns with more generate_awake monsters at level generation time 19:16:45 if that's clear 19:17:01 possibly with a provision to not place said monsters in LOS of stairs 19:17:33 and then we could have "spawns" basically only generate through the OOD timer system as summons; they wouldn't grant XP 19:18:28 this ostensibly wouldn't be a huge difficulty increase, although it probably would make the game a bit more difficult, simply since the level would generate with more monsters initially, and some of those monsters would always be awake 19:18:45 we could try it out in a branch if we were concerned about the numbers being right 19:19:03