00:00:07 not working for me on cbro either.. 00:00:07 equip_bar=true in rc 00:01:19 ^status 00:01:20 31 Crawlers. CBRO disk usage=94% (135GB) | RAM usage=27% (4GB)| uptime/CPU= 00:01:18 up 53 days, 4:39, 2 users, load average: 0.88, 0.62, 0.51 (4 Cores) http://status.berotato.org 00:02:18 &rc gammafunk 00:02:28 http://crawl.berotato.org/crawl/rcfiles/crawl-git/gammafunk.rc 00:03:23 I'm was testing locally 00:03:38 but I'm recompiling trunk atm since I wanted to test webtiles 00:05:32 hrm, does it only work in console? 00:05:36 that could be the problem 00:06:03 oh 00:06:11 lol 00:06:20 alex why no console on CPO 00:07:47 "what's webtiles" - elliptic 00:08:10 :P 00:08:57 advil dev blog post when 00:09:27 kramin: well then where's your patch for webtiles equip bar, mr. fancy pants 00:09:54 * kramin hides 00:09:54 advil does his own dev blog?! 00:10:07 I think you do it 00:10:27 oh you mean he does a trunk update 00:10:33 historically I think only Grunt, PF, and I have ever done them 00:10:36 no I mean an announcement celebrating the "new" dev! 00:10:45 only 4 months old or something by now 00:10:52 ah, I think we did the announce on tavern/mailing list 00:10:55 but not the blog, alas 00:11:03 well 00:11:07 i mean, *everyone* gets that! 00:11:25 I think I'll do a dev blog post dunking n1k for getting his twitter hacked though 00:11:33 you hacked it! 00:11:36 nop 00:11:47 honestly, it was wheals 00:12:14 story checks out 00:12:16 yeah actually that's way more plausible 00:16:01 what is 'dgamelaunch' mean? 00:17:45 ??dgamelaunch 00:17:46 dgl[1/4]: https://github.com/crawl/dgamelaunch 00:18:03 ASCIIPhilia: you shouldn't need to use that unless you'd like to support console games (over SSH) 00:19:20 it runs when SSH users connect to give them a menu to start crawl or edit settings 00:19:29 ah i understanded. but i interested in ssh support. 00:19:42 I missed not having SSH on cwz :) 00:20:00 -!- stephen is now known as smazza 00:20:02 the full instructions for setup we use are here 00:20:02 yep 00:20:15 https://crawl.develz.org/wiki/doku.php?id=dcss:server:setting_up_dgamelaunch_and_webtiles 00:20:28 thank you XD 00:20:29 describes setting up a chroot to run webtiles+dgl 00:21:16 good 00:21:20 I'll ask here when i fail :) 00:23:19 !cmd .gfgk 00:23:19 Command: .gfgk => .gfmap br!=D|lair kmap!~~entry|enter|runelock $* 00:23:30 !cmd .gfmap 00:23:30 Command: .gfmap => !lg * (( kmap~~gammafunk|wizlab_lehudib || kmap=evilmike_wizard_prison newprison )) $* 00:31:44 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.20-a0-1050-g57248f3 (34) 00:34:59 -!- 92AAAYFBA is now known as smazza 00:47:35 -!- wheals_ is now known as wheals 02:58:05 -!- amalloy_ is now known as amalloy 03:16:02 Why do downstairs look like exactly unexplored squares on minimap in trunk? 03:25:26 Unstable branch on crawl.beRotato.org updated to: 0.20-a0-1050-g57248f3 (34) 03:58:26 -!- amalloy is now known as amalloy_ 04:18:40 Formacids cannot Wall Jump if their Last Move was Digging 13https://crawl.develz.org/mantis/view.php?id=11033 by bcadren 04:23:50 Color indicators of downstairs and player character changed to less visible 13https://crawl.develz.org/mantis/view.php?id=11034 by Leszczynek 04:28:38 03gammafunk02 07* 0.20-a0-1051-gc5baec5: Fix WebTiles minimap colors (Raine357) 10(3 minutes ago, 2 files, 4+ 3-) 13https://github.com/crawl/crawl/commit/c5baec59d198 05:10:12 Unstable branch on crawl.jorgrun.rocks updated to: 0.20-a0-1051-gc5baec5 (34) 06:21:24 -!- herself1 is now known as herself 06:58:49 incorrect tile shown for deep elf elementalist next to map 13https://crawl.develz.org/mantis/view.php?id=11035 by hollowman 07:20:43 -!- neady is now known as eady 08:55:29 wow has it really been 4 months? 08:55:30 I guess so 08:59:00 if I really need announcing on the blog it can just be folded in with something else :) 08:59:08 0.20 release changelog or something 09:03:01 kramin, gammafunk: yeah, equip_bar is console only, I believe at the moment the option will do nothing at all on webtiles 09:03:01 not sure what it will do in tiles 09:35:08 -!- Palyth_ is now known as Palyth 09:35:21 !tell elliptic actually getting some runs back on the qw genetic algorithm...my current reward is #branches/#levels/orb and its really focusing in on spriggans, kind of interesting 09:35:21 espais: OK, I'll let elliptic know. 09:35:39 !tell elliptic out of curiosity, does qw handle non-berserkers at all? 09:35:40 espais: OK, I'll let elliptic know. 09:38:14 espais: it has some makleb logic 09:38:22 I think it won a DD of makh, in fact 09:41:08 qw has won a ton of different classes, you can query it 09:41:09 !lg qw won s=god,char 09:41:10 108 games for qw (won): 50x Trog (23x GrBe, 5x MiBe, 4x DDBe, 2x NaBe, DsBe, DDMo, DDNe, DDSu, DDTm, DDVM, DDIE, DDEn, HOBe, HuBe, MfBe, DDEE, DDCK, TeBe, TrBe, VSBe), 32x Makhleb (18x DDFi, 2x DDGl, DDAE, DDAr, DDAs, DDCj, DDFE, DDAM, DDAK, DDHu, DDSk, DDWn, DDWr, DDWz), 10x Okawaru (BaFi, CeFi, DDFi, DrFi, FoFi, GhFi, MiFi, MuFi, OgFi, VpFi), 3x The Shining One (2x MiFi, GrFi), 2x Uskayaw (2x Mi... 09:41:20 well hot damn, you guys are the best 09:41:27 and my knowledge of Sequell is awful 09:41:36 i was just going by the commented combos in his rc file 09:41:47 !lg qw won s=god 09:41:48 108 games for qw (won): 50x Trog, 32x Makhleb, 10x Okawaru, 3x The Shining One, 2x Yredelemnul, 2x Cheibriados, 2x Ru, 2x Uskayaw, 2x Qazlal, , Hepliaklqana, Xom 09:41:54 !lg qw won xom 09:41:55 1. qw the Conqueror (L27 MiFi of Xom), escaped with the Orb and 3 runes on 2016-11-19 04:17:52, with 1981201 points after 52997 turns and 1:10:28. 09:41:59 geez 09:42:02 a xom win 09:42:11 what I'd really like to do at some point is train a RNN of some kind on crawl tvs 09:42:44 why an RNN over an ANN? 09:43:48 hmm, so i guess the moral of the story here is that qw is pretty damned impressive and it can probably handle any combo, more or less 09:44:38 !gamesuntilwin qw qaz 09:44:40 qw had 39 games before their first win. 09:44:52 well, it only has specific logic for some god abils. and it throws certain really valuable potions on the ground due to not having logic for using them 09:45:00 !gamesuntilwin qw chei 09:45:02 qw had 4 games before their first win. 09:45:12 !gamesuntilwin qw trog 09:45:14 qw had 161 games before their first win. 09:46:02 proof that Chei is better than trog, even for a bot 09:46:04 like if qw plays qaz currently, it just "benefits" from the passive storms, doesn't use the invocations 09:46:42 qw is impressive, don't get me wrong, it just is only certain kinds of fantastic ;) 09:56:50 !gamesuntilwin rumflump qaz 09:56:51 No wins for rumflump qaz. 09:57:08 espais: ann is a very broad category, lots of neural network architectures just aren't great at doing much of anything interesting. RNNs are _much_ more effective at complex tasks than ANNs in general 09:57:16 insulting qw when he's better than you! :O 09:57:57 see e.g. http://karpathy.github.io/2015/05/21/rnn-effectiveness/ 10:02:37 advil: i think it probably depends on the situation 10:03:04 there's a lot of movement in CS right now to throw RNNs/CNNs at everything rather than understand the problem domain 10:03:23 but thank you for not calling it deep learning...for some reason that term bugs me 10:08:21 yes, I'm in an adjacent field where we study the problem domain 10:09:15 I'm not sure that issue really applies to crawl though :) 10:10:02 I doubt any neural network architecture without recurrence of some kind would be able to get anywhere at playing crawl 10:10:47 depends though i think 10:11:00 on? 10:11:04 i'd imagine a crawlbot could get by with simple monitor/response actions 10:11:14 your basic NN just takes in sensor readings and outputs an action 10:11:35 that's really all a basic bot does right? 10:11:45 you hit 50% health and it triggers a run away neuron 10:11:52 <|amethyst> to be fair, "understanding the problem domain" involves throwing a really big really poorly understood neural network at it :) 10:12:09 that's always the CS argument though :) 10:12:18 why use a screwdriver when you can use a hammer 10:12:48 i'd bet that an elegant RNN would probably play more intelligently though 10:13:20 sorry you know what 10:13:25 i'm thinking CNNs, not RNNs 10:13:28 RNN yes i'd agree 10:13:37 you need that feedback loop to learn 10:13:39 duh 10:14:13 sorry, recently sat through a presentation on CNNs and how they should be used in any place that any other type of NN is used 11:34:25 qw is hand coded 11:34:28 QED 11:34:40 where are the CNNs that are as good as qw 11:34:52 bring them to the floor and show us what they got! 12:06:14 %git efdfc71969b016ec043879 12:06:14 07|amethyst02 * 0.17-a0-1883-gefdfc71: Fix crash on player-reflected IOOD/returning. 10(1 year, 7 months ago, 1 file, 12+ 3-) 13https://github.com/crawl/crawl/commit/efdfc71969b0 12:06:38 this reflection+injury mirror crash is menv[YOU_FAULTLESS] issues again 12:16:14 I'm sort of surprised anything works, right now if ranged damage reflects of the player and hits something, monster::hurt is called with MONS PROGRAM_BUG as the agent 12:27:40 |amethyst: maybe the player should get injury mirror damage if they reflected a ranged attack, not the attack source? (contrary to what you suggested yesterday) 12:28:00 it's a lot easier to implement, by doing something like what you suggested in the commit msg for efdfc71969b0 12:45:50 03advil02 07* 0.20-a0-1052-g14f416c: Make monster::name return something more informative for YOU_FAULTLESS 10(3 minutes ago, 1 file, 4+ 0-) 13https://github.com/crawl/crawl/commit/14f416c661c3 13:05:37 Unstable branch on crawl.akrasiac.org updated to: 0.20-a0-1052-g14f416c (34) 13:09:25 Unstable branch on crawl.jorgrun.rocks updated to: 0.20-a0-1052-g14f416c (34) 13:29:03 -!- amalloy_ is now known as amalloy 16:45:31 03MarvinPA02 07* 0.20-a0-1053-g0c1af0a: Give DDoor an independent cooldown 10(6 minutes ago, 5 files, 11+ 5-) 13https://github.com/crawl/crawl/commit/0c1af0aa458e 16:45:31 03MarvinPA02 07* 0.20-a0-1054-gd950c8b: Make Borgnjor + DDoor only cause normal paralysis 10(5 minutes ago, 2 files, 2+ 4-) 13https://github.com/crawl/crawl/commit/d950c8b884ad 16:45:31 03MarvinPA02 07* 0.20-a0-1055-g19bae6a: Give Berserk an independent cooldown 10(9 weeks ago, 12 files, 27+ 22-) 13https://github.com/crawl/crawl/commit/19bae6a9a3b1 16:45:31 03MarvinPA02 07* 0.20-a0-1056-g7622305: Give Recite an independent cooldown 10(9 weeks ago, 5 files, 12+ 5-) 13https://github.com/crawl/crawl/commit/7622305906e7 16:46:17 nice 16:47:04 huge draconian^zin buffs 17:09:21 Unstable branch on crawl.jorgrun.rocks updated to: 0.20-a0-1056-g7622305 (34) 17:47:10 not to mention ettin of trog+ru 17:47:45 finally "bloodlust is viable on ru 17:55:49 <|amethyst> I'd really rather have some kind of generic framework for cooldown timers 17:56:09 <|amethyst> not sure what it would look like though (beyond the current duration stuff) 18:15:46 Unstable branch on underhound.eu updated to: 0.20-a0-1056-g7622305 (34) 18:45:10 !tell hellmonk updated 18:45:11 alexjurkiewicz: OK, I'll let hellmonk know. 18:57:32 !gitgrep 1 [Jjerky] 18:57:32 %git HEAD^{/[Jjerky]} 18:57:32 07MarvinPA02 * 0.20-a0-1056-g7622305: Give Recite an independent cooldown 10(9 weeks ago, 5 files, 12+ 5-) 13https://github.com/crawl/crawl/commit/7622305906e7 18:57:50 wow, must have been a stealth jerky nerf! 18:57:56 oh, lol 18:57:58 good regex 18:58:04 !gitgrep 1 [Jj]erky 18:58:04 %git HEAD^{/[Jj]erky} 18:58:04 07MarvinPA02 * 0.20-a0-504-g2c65edc: Remove pizza and beef jerky 10(4 months ago, 48 files, 58+ 430-) 13https://github.com/crawl/crawl/commit/2c65edc55d88 19:22:02 oh, ?noise is marked fully as useless now? 19:35:52 -!- amalloy is now known as amalloy_ 19:36:23 did the minimap on webtiles change? 19:36:23 there used to be a white dot denoting where you were 19:36:40 and now it's purple and very hard to tell 19:36:52 because it doesn't contrast much from the default grey floor 19:38:34 ?noise is extremely useful! you use it lure elves out of their vault 19:42:33 %git c5baec59d198ef772deec5 19:42:33 07gammafunk02 * 0.20-a0-1051-gc5baec5: Fix WebTiles minimap colors (Raine357) 10(15 hours ago, 2 files, 4+ 3-) 13https://github.com/crawl/crawl/commit/c5baec59d198 19:43:19 I think this must have done something (unintended?) but I haven't looked into it 19:44:03 !bug 11034 19:44:04 https://crawl.develz.org/mantis/view.php?id=11034 19:46:44 Crash in Vaults:2: ASSERT(idx < TILE_FEAT_MAX) in 'rltiles/tiledef-dngn.cc' at line 86 failed. 13https://crawl.develz.org/mantis/view.php?id=11036 by Qwert225 19:46:47 or maybe that commit fixed it and it hasn't rolled out yet? 19:47:38 yeah, I think that's it, looks like gammafunk did it directly in response to that bug 19:47:53 so, it'll update tonight at the latest for all servers 19:49:32 A bunch of good changes the last couple days: transporters and cooldown cleanup. Too bad about Monks, but it definitely didn't work as well as it seemed it would on paper. 19:49:32 Brannock: You have 2 messages. Use !messages to read them. 19:55:48 mm 20:00:48 0.20 is shaping up to be pretty action packed! 20:00:59 It seems there's a couple changes to Gnoll in the pipeline, too 20:04:22 alexjurkiewicz: I forgot to follow up, have you used a mouse on webtiles in retina any time lately? 20:05:00 I should have fixed that cursor offset thing but who knows, and I don't have easy access to a retina display myself 20:05:07 i'll try now! 20:05:11 it works!! 20:05:39 (although I almost bought a new macbook pro today, and then got demotivated adding all the extra dongles to the cart) 20:05:42 awesome, thanks! 20:06:25 unfortunately osx screenshot hides the cursor so I can't prove this unless I take a potato photo, but it's great 20:07:13 well, I believe you 20:08:41 thanks for the fix, i think it's huge for newbies 20:09:32 ??rebuild 20:09:32 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 |amethyst or Nap.Kin for CDO. Use your powers wisely. 20:09:56 yeah, it's one of those things the I think flew under the radar for a while because so few devs use a configuration that triggers the bug 20:10:22 plus mouse in webtiles is still experimental (though I think with this fix doesn't need to be) 20:10:44 advil: the tile crash? I'm not sure my fix addressed that 20:11:00 my fix was about the minimap 20:11:28 yeah, I was responding to a question about that 20:11:50 I think advil is asking about a mouse position fix committed a couple days ago that they couldn't find testers for? 20:11:53 I was just confused about the timing (and that crash report came in the middle) 20:11:55 although following on yours 20:12:00 right, that was separate too :-) 20:12:20 oh 20:12:36 yeah that was good stealth Cheibriados 20:12:46 I'm rebuilding cbro to fix my minimap 20:13:27 s/my minimap/the minmaps of all our dear players on cbro/ 20:14:30 heh 20:15:37 hmm, we should remember to put the mouse on webtiles in the 0.20 changelog (IIRC it was in 0.19a for a while and then disabled for release) 20:16:05 I guess I could just add it to the current changelog retroactively 20:16:06 Crawl has tiles? # sorry devs 20:16:31 many devs don't play with tiles 20:16:35 no worries on that front 20:16:52 yeah, there are a few missing changelog things 20:17:00 I missed the GSC base delay change 20:17:28 wow, no shout outs from the funkmeister to summoners 20:18:15 I said mages 20:18:18 did summoners change? 20:18:34 I assume he's referencing the opening of my update post 20:18:43 oh 20:19:00 obv. CanOfWorms is just another sword-and-board melee toon 20:19:09 that's the only type of character I play! 20:19:22 why mess with a winning strat? 20:19:25 right? 20:25:31 !lg * race=gnome s=v 20:25:32 3547 games for * (race=gnome): 967x 0.4.5, 735x 0.4.4, 584x 0.2.7, 501x 0.3.4, 195x 0.1.7, 134x 0.3.3, 75x 0.5.0, 68x 0.3.0, 61x 0.4.3, 59x 0.3.2, 41x 0.4.0, 40x 0.4.1, 30x 0.2.4, 26x 0.2.5, 17x 0.3.1, 14x 0.4.2 20:26:29 03advil02 07* 0.20-a0-1057-g9a88637: Convert YOU_FAULTLESS back to you for injury mirror 10(3 minutes ago, 3 files, 29+ 3-) 13https://github.com/crawl/crawl/commit/9a88637c54ed 20:28:11 !lm !bot br.enter=tomb newtomb / lg:br=tomb 20:28:25 !lm !bot br.enter=tomb current !newtomb / lg:br=tomb 20:28:37 19/72 milestones for bot (br.enter=tomb newtomb): N=19/72 (26.39%) 20:28:41 443/3854 milestones for bot (br.enter=tomb current !newtomb): N=443/3854 (11.49%) 20:28:55 deadly 20:29:06 well it will come down fast 20:29:13 sad 20:29:33 people will learn to adapt to it; the strats I've seen by the tvs I've watched have been appalling 20:30:12 !wizlabdeaths * recent 20:30:15 WizLab deaths for * (recent): 59/1013x Cigotuvi's Fleshworks [5.82%], 62/1073x Lehudib's Moon Base [5.78%], 51/1026x Iskenderun's Mystic Tower [4.97%], 47/1024x Doroklohe's Tomb [4.59%], 42/1038x Wucad Mu's Monastery [4.05%], 28/1124x The Chambers of the Cloud Mage [2.49%], 24/1124x The Roulette of Golubria [2.14%], 18/1059x Zonguldrok's Shrine [1.70%], 15/1030x The Hall of the Hellbinder [1.46%] 20:30:21 interestingly people seem to be dying in fleshworks more 20:30:26 due to the new cig's monsters 20:30:34 03advil02 07* 0.20-a0-1058-g0491fde: Retroactively add webtiles mouse to the changelog 10(78 seconds ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/0491fdee9de3 20:30:52 !lg * recent kmap~~cigotuvi s=cikiller% 20:30:53 53 games for * (recent kmap~~cigotuvi): 39x Cigotuvi's Monster (73.58%), 7x a very ugly thing (13.21%), 4x a large abomination (7.55%), 3x a shapeshifter (5.66%) 20:31:03 when did the CM change? 20:31:04 like, almost completely due to that monster 20:31:08 some time ago 20:31:28 unknown monster: "cogotuvi's monster" 20:31:28 %0.17?cogotuvi's monster 20:31:33 unknown monster: "cogotuvis monster" 20:31:33 %0.17?cogotuvis monster 20:31:35 er 20:31:39 unknown monster: "cigotuvi's monster" 20:31:39 %0.17?cigotuvi's monster 20:31:42 unknown monster: "cigotuvis monster" 20:31:42 %0.17?cigotuvis monster 20:31:44 ug 20:31:46 lol 20:32:57 hrm 20:32:58 weird 20:33:00 Cigotuvi's Monster (13X) | Spd: 10 | HD: 23 | HP: 200 | AC/EV: 5/5 | Dam: 22, 17, 13, 903(constrict) | 05demonic, 10doors, amphibious, see invisible, 07vault | Res: 06magic(160), 05fire, 02cold, 10elec, 03poison, 12drown, 04rot, 13neg+++, 13torm | Vul: 08holy++ | XP: 4221 | Sz: Giant | Int: normal. 20:33:00 %0.16?cigotuvi's monster 20:33:05 Cigotuvi's Monster (13H) | Spd: 10 | HD: 27 | HP: 122-169 | AC/EV: 13/5 | Dam: 40, 2507(trample), 25(reach), 2503(constrict) | 07vault | Res: 06magic(108), 02cold, 03poison | Chunks: 06mutagenic | XP: 3765 | Sz: Medium | Int: animal. 20:33:05 %0.18?cigotuvi's monster 20:33:09 unknown monster: "cigotuvi's monster" 20:33:09 %0.17?cigotuvi's monster 20:33:21 <|amethyst> %git b0fe834 20:33:21 07regret-index02 * 0.18-a0-1606-gb0fe834: Buff wizlabs Cigotuvi & Cloud Mage, tweak wizlab Wucad Mu 10(1 year, 1 month ago, 2 files, 36+ 39-) 13https://github.com/crawl/crawl/commit/b0fe83464918 20:33:52 it does have quite a nasty array of attacks, I guess 20:33:57 for scrolls of fear, it uses mass_enchantment(ENCH_FEAR, 1000); and learndb says it casts Cause Fear at roughly power 134...anyone know how 1000 convert into spellpower 134? or was there some change along the way I missed? 20:36:59 or does it actually mean that the spell caps at effectiveness at spellpower 134, even though it goes to 200; thus, scroll of fear just maxes out the spellpower with 1000 as a placeholder? 20:38:41 hmm 20:38:50 that seems suspicious 20:39:10 <|amethyst> didn't enchantment power use to be spellpower * 6 20:39:14 the 200 cap for regular cause fear is enforced in _ench_power which is never called for that case 20:41:02 I think I see why in mass_enchantment: pow *=3; pow /=2; pow = min(pow, 200); 20:42:18 so it's referring to 134 being enough to hit the 200 cap in mass enchantment for the cause fear spell 20:42:27 ah, I see 20:42:45 so the 1000 number for fear is likely either a placeholder or not changed from if enchantment power used to be it's own thing 20:48:07 heh that number is pre-dcss 20:48:16 so much for trying to figure out why 20:48:31 I didn't check the degree to which power / ench caps might have changed out from under it, they could have 20:49:34 Unstable branch on crawl.beRotato.org updated to: 0.20-a0-1058-g0491fde (34) 21:09:23 Unstable branch on crawl.jorgrun.rocks updated to: 0.20-a0-1058-g0491fde (34) 22:15:14 alright, any feedback on these Gnoll changes before I make a pull request for them? https://github.com/crawl/crawl/compare/master...Floodkiller:gnoll_balance 22:15:33 hellmonk: no doggo in hellcrawl? sad! 22:15:33 alexjurkiewicz: You have 1 message. Use !messages to read it. 22:15:38 !messages 22:15:38 (1/1) hellmonk said (33m 49s ago): thank you! 22:15:47 havent gotten around 2 it yet 22:15:55 this is mainly an attempt at re-balancing the current system, as well as giving them something to make them less bland 22:16:42 i like the bark effect 22:17:15 if they still aren't good with the current system, I would recommend holding them back from 0.20 to get an alternative system/approach made and tested 22:17:30 current system meaning with those changes 22:17:46 I like shifting some of their power with aptitudes into an ability, as well as making the curve harsher 22:17:54 maybe I should steal that bark mut as a generic goodmut 22:18:02 Psychic Scream 22:18:06 let people bark as an ogre 22:18:50 looking at that new XP table I think it'll help solve the issue of people trying to ignore the species's constraints by focusing on a minimum few skills 22:19:05 because that curve becomes really whoa 22:19:30 yeah. How much XP do you need to reach level 14 in a skill now, compared to current gnoll and compared to 0 apt? 22:19:50 because to me that's the sort of important cut off, weapon mindelay with most classes 22:19:59 14? 6000 baseline, old gnoll: 10723, new gnoll: 15076 22:20:26 eggs pee 22:20:41 Dungeon Crawl Stone Soup v3.12-hellcrawl-2-g3d782e9 22:20:55 is this the version with fewer weapons? 22:20:58 no 22:21:05 I have not pushed anything related to that yet 22:21:15 Floodkiller, do you have that gnoll xp comparison spreadsheet still? 22:21:16 oh, sorry, I thought your message was about having pushed that 22:21:19 Brannock: thanks 22:21:23 https://docs.google.com/spreadsheets/d/1UpAgbzL8OENPac_ubosSXxroNSN8llOtZWkuAUtyXX8/edit#gid=0 22:22:32 current gnoll is column AO, new gnoll from those changes is column Y 22:22:39 15076 is still doable if you're very dedicated, which I think is fine (gnoll shouldn't lock you out *totally*), but it begins to heavily suggest "level something else" 22:23:00 hellmonk: weird hellcrawl bug. you know that early trog altar vault with a caged moth of wrath? I can't autoexplore in that room after killing everything else now. I press o, nothing happens 22:23:11 huh 22:23:18 but when I press a movement direction the game says "_Key pressed, stopping explore" 22:23:25 so it's not terminating the autoexplore? 22:23:29 interesting 22:23:30 neat bug 22:23:40 you can spec me if you want to see it in action rn 22:23:53 ok, so 50% more xp needed for level 14 with your new scaling 22:23:57 seems good 22:24:01 wait. now it works! 22:24:12 maybe i'm crazy 22:24:34 you always seem to get these interesting bugs 22:24:42 alexjurkiewicz, did you still get the bugs with the Xom friendship altar activating twice? 22:24:44 do i? 22:24:47 when you write bad code the bugs become much more interesting 22:25:01 no, the next time I saw it I had no issues 22:25:18 my rc file is so boring though, and I only press o and tab 22:26:08 btw I love this comment: // Demonspawn probably want to use weapons close to the "natural" // demon weapons - demon blades, demon whips, and demon tridents. // So pick from a selection of good weapons from those classes // with about a 1/4 chance in each category of having the demon // weapon. // XXX: this is so ridiculously overengineered 22:36:56 Floodkiller, yeah I like the apt change, should make it a lot more punishing to train skills above 12 or so 22:46:36 not sure the bark ability is really necessary 22:46:47 they are already strong early game 22:51:20 many of the complaints about gnolls were about how bland they were, with the shifting aptitudes not being interesting enough to stand on its own 22:52:38 I don't really understand that complaint :P 22:53:10 I like the aptitude gimmick 22:53:24 Is making the doggo apt lock even harder really the answer? 22:53:43 yes imo 22:54:16 needs to actually be hard to win the game without hybridising with gnolls 22:55:07 at least 2 three win games went with a Trog-esque skill build, without Trog; another good half of the games hyper focused a couple skills and barely spread out unless required 22:55:12 three rune* 22:59:23 needs to actually be hard to win the game without hybridising with gnolls 22:59:46 how are you going to do that with skill restrictions alone though, when stuff like quick blades and shadow creatures exist 23:00:27 I guess I'll just end up repeating myself on this 23:00:50 if shadow creatures is giving you a dispropotionate amount of power for your skill investment maybe that is an issue with shadow creatures 23:02:27 My point is that there are a ton of ways to excel in a specialty without having much in the way of skill levels: good short blades, demon weapons, some spells, several gods 23:03:36 If the goal is for "hybridizing" to be better than not hybridizing, that goal has already been met - in fact, it's met by the majority of species in the game 23:05:50 But if you want specializing with gnolls to be way worse than doing it with other races I think you need more than an apt gimmick because specialization isn't that strongly based on skills 23:07:07 you mean it's based more on the equipment you find? 23:07:54 yeah, and god choice 23:08:54 there are multiple gods that let you specialize in using allies to kill things, and be REALLY good at it, without any skills at double-digit levels 23:11:00 but even without gods, weapons don't have skill requirements and spells have wizardry 23:13:08 I think a lot of the complaints about blandness, incidentally, stem from a perception of highly specialized characters as exciting, and more optimally built characters as boring 23:13:37 I was exciting once 23:13:40 what if you compare e.g. Hu and DE to Hu and Gn 23:13:59 level 9 spells and exec axes have Crawl Meme Status whereas inexperienced players aren't too excited about that kiku character with a spear of pain, even though it probably does more damage than the exec axe 23:14:08 can you argument not also be applied to say DE are basically the same as humans? 23:14:23 probably more so than gnolls 23:14:52 I don't recall ever arguing that gnolls are basically the same as humans 23:15:03 well 23:15:41 DE are not as strong as Hu, so you have to be specific as to what you mean 23:16:15 I have argued that the optimal skill training for them is basically the same, but the apt gimmick means that they get a ton of frontloaded power on their skills, not to mention they have fangs which is more impactful than gargoyle's entire mutation list 23:16:37 seriously fangs are amazingly good 23:16:44 they have fangs? 23:16:46 heh 23:16:56 I've not played enough Vp to know 23:17:39 but in terms of "same as Hu", that was talking about the hybridization aspect specifically 23:17:49 yeah 23:18:00 which is not good for the Gn design if it's the central premise 23:18:55 it does try to nudge you more strongly into doing that, which some people are ok with saying "eh, good enough"; I think others don't like how you have to keep looking at your skills to see apt changes 23:19:11 like internally I guess you still want to do that even if you're not Gn but that species makes it really explicit 23:19:54 ??fangs 23:19:55 sharp teeth[1/2]: Grants an auxiliary unarmed biting attack, increasing in damage with the level of the mutation. Vampires start with 3 levels of sharp teeth. 23:20:08 ??fangs[2 23:20:08 bite[1/1]: An {auxiliary unarmed} attack you can get if you have sharp teeth or fangs. Damage is (fang mutation level * 2) + ((strength-10) / 5, randomly rounded). If you have less than 10 str, it's just (fang mutation level * 2). For VS, it instead is fangs level * 2 + XL * 1/3 23:21:00 when it says damage, is that the die size of damage, or actually the damage? 23:21:21 everything is rolled right 23:21:28 I'm rolling right now 23:22:04 it's die size yeah 23:22:24 !stats gnfi 23:22:25 Starting stats for GnFi: Str 15 Int 8 Dex 13. Stat gain: Ba sid/5 23:22:30 with respect to comparing to humans, the main reason to say gnolls are a bad species I think would be that they aren't sufficiently different from existing ones 23:22:34 also it's random2 rather than a "die" so it actually rolls 0 through 6 23:23:00 if you just want to trade one boring species for another then you could cut halfling 23:23:02 but still, if you start a GnFi you get an extra 0 through 6 damage on 40% of your melee attacks 23:23:05 remove the tolkien minority 23:23:07 that's pretty amazing 23:23:09 oh are doggos fangs 3? 23:23:13 thought it was just 1 23:23:31 I think it was only fangs 1 that brannock added 23:23:35 oh huh it is 23:23:42 0 through 2 damage is still amazing! 23:24:03 free infuse? 23:24:04 lol minmay looks like you could have used crawl showing you damage numbers 23:24:07 a bit less 23:24:26 too bad you couldn't convince the devs to add it 23:24:32 FR: remove dev team hatred towards showing damage numbers 23:24:32 (it applies fighting and slay too of course, also gets the str multiplier despite already having str-based base damage) 23:24:39 gammafunk: jokes on you ive never played GnFi 23:30:21 actually seems odd that Vp would have fangs 3, doggos fangs 1 23:30:46 I guess our Vp look like that one in the movie of Salem's Lot 23:31:12 nosferatu teeth 23:31:24 gammafunk: but man can you really blame me for not being able to tell the difference between like 0-35 extra damage and 0-39 extra damage 23:31:47 er, more like 0-35 and 0-43, sorry 23:32:04 yep, just like I blamed elliptic for not detecting meleebug despite multiple wins in affected versions 23:32:09 you guys are supposed to be the pros 23:34:01 I didn't play during meleebug, for all you know I would have noticed it on the first hit! 23:34:02 you should add a species that does double damage in melee 23:34:11