00:02:05 03elliptic * r0cc5d388db87 10/crawl-ref/source/ (6 files): Make ponderous items have a larger effect on characters with limited armour slots. 00:02:15 03elliptic * rd7543a2c5c49 10/crawl-ref/source/ (godpassive.cc godpassive.h mutation.cc): Let Chei know when your number of armour slots changes (due to mutation). 00:03:08 Unstable branch on crawl.develz.org updated to: 0.9-a1-687-g5b1bdd3 (32) 00:06:20 oh cool, good timing 00:08:36 oh no 00:08:45 now i have to decide on what key to assign shroud of golubria to! 00:08:54 every other buff has its own key :( 00:09:50 -!- hoody has joined ##crawl-dev 00:12:51 I guess the chei changes just failed to make it in that update 00:13:26 just as well given that I think it might be slightly bugged still with mutations 00:14:08 marvinpa: you should implement the idea tossed around here the other day of making spell letters configurable in your rc 00:14:20 i should 00:14:27 that would be awesome 00:14:48 i'd win three times as many crawl games without needing to spend time every game reconfiguring letters! :P 00:16:10 -!- ortoslon has joined ##crawl-dev 00:18:17 Windows builds of master branch on crawl.develz.org updated to: 0.9-a1-689-gd7543a2 00:19:20 I guess my chei buff made it into the windows builds 00:32:04 03elliptic * r99256ef0fb37 10/crawl-ref/source/mutation.cc: Move a check for change in number of slots to the correct location. 00:36:08 -!- Twilight-1 has quit [Read error: Connection reset by peer] 00:38:40 moin 00:38:57 -!- Twilight-1 has joined ##crawl-dev 00:48:14 skills.cc: 50 you.species = SP_MINOTAUR; 00:48:23 what, why, I don't get it 00:52:08 I guess exp_needed depends on your race 00:52:50 and maybe skill_cost_needed isn't supposed to depend on your race 00:53:05 still looks silly :) 00:59:19 it does 01:11:56 -!- ophanim has quit [Read error: Connection reset by peer] 02:01:52 -!- ortoslon has quit [Quit: bye] 02:49:46 -!- monqy has quit [Quit: hello] 02:53:01 -!- hoody has quit [Remote host closed the connection] 03:03:33 -!- hoody has joined ##crawl-dev 03:09:08 elliptic: about the reaching+hostile rat abuse: I fixed it but I have second thoughts if I should commit that. It would cause you to hit your allies... 03:09:38 kilobyte: the idea was only to hit hostiles in the way 03:09:58 your allies and you can work together to avoid hitting them ever 03:10:10 (but you still might fail to reach the target) 03:11:29 basically I see reaching through a hostile kobold (who is attacking you) to hit something as different from reaching over the shoulder of your orc ally to hit something 03:25:17 -!- ortoslon has joined ##crawl-dev 03:27:38 -!- galehar has joined ##crawl-dev 03:29:36 -!- syllogism has joined ##crawl-dev 03:30:05 03kilobyte * r5601322e8ec3 10/crawl-ref/source/evoke.cc: Make reaching past a monster avoid friendlies (elliptic) 03:30:15 03kilobyte * r3aa79263c3a0 10/crawl-ref/source/evoke.cc: Stop the reaching + hostile rat abuse. 03:30:16 03kilobyte * r35113e786bdc 10/crawl-ref/source/evoke.cc: Restore teleporting of mimics after a reaching attack. 03:41:03 !tell dpeg if you don't give us a formula for the Singing Sword, you risk it having misleading behaviour in 0.9 03:41:03 kilobyte: OK, I'll let dpeg know. 03:41:08 -!- hoody has quit [Remote host closed the connection] 04:07:06 We can also let the ogre behind the rat step on it. If there's a big difference in threat level (ash's tier?), the strong monster will cut through the punny one to get to you. 04:09:44 galehar: Should work on size not threat imo. 04:09:57 Or maybe a combination. 04:11:42 -!- elliptic has quit [Ping timeout: 240 seconds] 04:23:58 the fiend impales the midge en route to your location! 04:42:29 -!- ortoslon has quit [Read error: Connection reset by peer] 05:59:13 -!- galehar_ has joined ##crawl-dev 06:01:21 -!- edlothiol has joined ##crawl-dev 06:04:27 03galehar * r8b8caab12b50 10/crawl-ref/source/skills.cc: Add a comment. 06:21:58 -!- evilmike has quit [] 06:34:50 -!- casmith789 has quit [Read error: Connection reset by peer] 06:49:50 -!- galehar_ has quit [Remote host closed the connection] 07:21:48 uh oh, killing a zombie/skeleton in one blow gives you most of xp the monster would give when alive, killing it in small blows gives almost nothing 07:22:07 what counts is the maxhp at the time of death 07:22:56 kilobyte: lol 07:23:08 kilobyte: Is that true for all monsters? 07:23:20 Like the summoning goblin things? 07:23:27 They should give inf xp. I hate them. 07:23:57 summons never give xp 07:24:13 I mean the actual summoners. purple g's 07:29:32 -!- st_ has joined ##crawl-dev 07:31:41 boggart? 07:33:11 Yeah... 07:33:32 They have pretty low HP, do they give very little XP because of that? 07:39:22 @?? boggart 07:39:23 boggart (06g) | Speed: 12 | HD: 2 | Health: 6-16 | AC/EV: 0/12 | Damage: 5 | Flags: spellcaster, see invisible | Res: 06magic(18) | XP: 21 | Sp: confuse, slow, invisibility, blink, shadow creatures. 07:40:13 I don't think XP is always directly related to HP. kilobyte was talking about a problem specific to undeads. 07:42:25 kilobyte: that reminds me, rotted monsters are displayed as unharmed, because maxhp is reduced. maybe they give less xp too. 07:43:39 same for draining, but there, we do want reduced XP. However, monster should probably be displayed as hurt. 07:44:35 for undeads, we want full xp but hide wounded status. 07:44:55 for rotted, we want full xp and show it as wounded 07:52:14 monsters can't cure rotting with healing too, but that's probably irrelevant. 07:56:20 so, we need to give a new property to monsters. total_hp? decreased only by draining. Used for XP and showing wounded status. drained monsters will still be displayed as unharmed. Maybe it's ok. 08:40:13 -!- galehar has quit [Ping timeout: 252 seconds] 08:40:59 -!- blackpenguin has quit [Ping timeout: 258 seconds] 08:44:16 -!- galehar has joined ##crawl-dev 08:46:49 -!- elliptic has joined ##crawl-dev 08:53:15 -!- Whitewater has joined ##crawl-dev 08:55:41 -!- blackpenguin has joined ##crawl-dev 09:03:35 about D:1 balance, instead of boosting player HP, it has been suggested to reduce hobgoblin and kobold damage by 1 (maybe snake too). 09:06:57 boosting player HP makes more sense because it deals with other problems as well 09:07:01 like exploding darts 09:07:16 MojiPittan (L8 VpEn) (D:7) 09:07:17 or early gnolls 09:07:51 or monsters inexplicably having +9 clubs or whatever 09:07:58 I'd rather boost player HP also 09:08:11 and snakes aren't relevant for d:1 balance because they don't generate there :) 09:08:36 snakes do, quite a lot 09:09:20 no they don't 09:09:20 !lg * place=D:1 killer=snake 09:09:27 !lg * place=D:1 killer=giant newt 09:09:29 not by normal generation 09:09:29 2440. laan the Magician (L1 MuWz), slain by a snake on D:1 on 2011-06-24, with 40 points after 376 turns and 0:01:12. 09:09:32 3137. phyphor the Magician (L2 DEWz), slain by a giant newt on D:1 on 2011-06-24, with 28 points after 192 turns and 0:00:54. 09:09:48 !lg * place=d:1 killer=snake cv>=0.8 09:09:48 550. laan the Magician (L1 MuWz), slain by a snake on D:1 on 2011-06-24, with 40 points after 376 turns and 0:01:12. 09:09:58 !lg * place=d:1 killer=goblin cv>=0.8 09:09:59 4548. Iru the Skirmisher (L1 DSAs), succumbed to a goblin's poison on D:1 (fountainhead) on 2011-06-24, with 52 points after 356 turns and 0:03:36. 09:10:13 they can generate on stairs or through OOD generation only 09:10:32 !lg * place=d:1 cv>=0.8 / killer=snake 09:10:34 64061 games for * (place=d:1 cv>=0.8): N=550/64061 (0.86%) 09:10:40 Eronarn: gnolls don't show up in packs on D:1 09:10:46 but they account for less than 1% of D:1 kills 09:10:56 so don't worry about them 09:11:27 okay I should remove quits there I guess 09:11:40 !lg * ktyp!=quitting ktyp!=leaving place=d:1 killer=snake cv>=0.8 09:11:40 550. laan the Magician (L1 MuWz), slain by a snake on D:1 on 2011-06-24, with 40 points after 376 turns and 0:01:12. 09:11:53 !lg * ktyp!=quitting ktyp!=leaving place=d:1 cv>=0.8 / killer=snake 09:11:55 49482 games for * (ktyp!=quitting ktyp!=leaving place=d:1 cv>=0.8): N=550/49482 (1.11%) 09:12:03 still not significant 09:13:26 !lg * ktyp!=quitting ktyp!=leaving place=d:1 cv>=0.8 turn<1000 / killer=snake 09:13:28 44352 games for * (ktyp!=quitting ktyp!=leaving place=d:1 cv>=0.8 turn<1000): N=265/44352 (0.60%) 09:13:56 and mainly people who have been on the level for a long time 09:15:17 !lg * ktyp!=quitting ktyp!=leaving place=d:1 cv>=0.8 turn<1000 s=ckiller% 09:15:18 44352 games for * (ktyp!=quitting ktyp!=leaving place=d:1 cv>=0.8 turn<1000): 11860x a hobgoblin (26.74%), 10946x a kobold (24.68%), 4211x a goblin (9.49%), 2989x a jackal (6.74%), 2929x a giant gecko (6.60%), 1636x an ooze (3.69%), 1631x a giant cockroach (3.68%), 1456x a gnoll (3.28%), 1435x a megabat (3.24%), 1424x a rat (3.21%), 620x a giant newt (1.40%), 392x a small snake (0.88%), 315x a wor... 09:15:59 if we only tweak a few monsters, how do we decide when to stop on that list? 09:16:36 when we reach ancient liches, they are a valid spawn on D:1 if you stick there long enough 09:18:01 right, so let's increase HP instead :) 09:31:20 how? Do we increase the base hp given to backgrounds? To all backgrounds? By the same amount? Currently, it ranges from 8 to 15. 09:34:48 I'd say either add a fixed number to the base hps (somewhere in the 2-5 range) or multiply them by a fixed number (somewhere in the 1.2 to 1.5 range) 09:35:14 not sure how large an effect we want to try out 09:37:15 multiplying means a larger boost to the strong backgrounds. adding would be nicer to the weaker ones, but reduce differentiation. 09:46:50 galehar: it'd only reduce early differentiation, and they're not that differentiated in HP early on 09:47:44 difference between 8 and 15 is quite big I'd say. 09:48:26 but if we give say +2 to everyone, difference between 10 and 17 is greatly reduced, but still significant. 09:48:48 -!- edlothiol has quit [Ping timeout: 255 seconds] 09:50:21 difference between 8 and 15 is quite big but you also get the difference between a wizard and a fighter or berserker there... so people usually don't compare the HP 09:51:36 anyway I'd vote for trying out a small additive boost... +2 or +3 or +4 09:52:07 I vote for +2 or +3 09:52:26 * kilobyte for +4 or +5 09:53:20 !random2(5) 09:53:25 or something more drastic but with a nerf of the regen boost 09:53:27 doesn't work? 09:53:37 start scumming 09:53:46 no randomization, yeah 09:54:08 no, I was joking. Just trying to ask a random number from the bots. 09:54:19 oh okay :) 09:54:22 !rng 2 3 4 5 09:54:23 The RNG chooses: 3. 09:54:32 here we go! 09:55:31 -!- hoody has joined ##crawl-dev 09:56:55 kilobyte: if SAVE_DIR_PATH is defined (like it is for windows installer), user can't redefine the directory either from option or command line. Comment says it's for security reasons (initfile.cc:2402). 09:57:18 Does it make any sense? It's up to the OS to make sure the user doesn't write were he shouldn't. 09:58:58 it _did_ make sense when all multiuser installs were setgid 10:00:08 but it's not the case anymore, right? 10:00:17 so we can fix it. 10:00:46 exactly 10:02:01 if you're talking about that bug report, the issue about directory separator being \ or / is no longer relevant. Some wise guys at Microsoft snuck in compatibility with the rest of the world in early NT days. 10:04:45 no, it's about the fact that save_dir option and dir command line option are ignored for windows installer. Not sure if there's a mantis about it, but I've heard complains (and more importantly, it annoys me too!) 10:07:00 ok, one of recent bug reports mentioned both of these 10:10:26 re 10:13:29 -!- hoody has quit [Remote host closed the connection] 10:30:47 there are so many open bugs in crawl, it's kind of depressive 10:31:43 just think of all the microsoft-people 10:48:27 -!- galehar has quit [Quit: Page closed] 11:05:19 -!- Whitewater has quit [Quit: Nettalk6 - www.ntalk.de] 11:19:41 -!- edlothiol has joined ##crawl-dev 11:22:31 -!- hoody has joined ##crawl-dev 11:22:50 -!- hoody has quit [Remote host closed the connection] 11:34:05 jeanjacques (L11 HuHe) ASSERT(smc) in 'mon-util.cc' at line 1060 failed. (Lair:1) 11:39:14 -!- edlothiol has left ##crawl-dev 11:42:51 -!- Wensley has joined ##crawl-dev 11:52:25 -!- upsy has joined ##crawl-dev 12:05:11 -!- petete has quit [Remote host closed the connection] 12:15:09 -!- edlothiol has joined ##crawl-dev 12:35:47 -!- Twilight-1 has quit [Read error: Connection reset by peer] 12:36:08 -!- Twilight-1 has joined ##crawl-dev 13:07:25 Zombified creature left a corpse (https://crawl.develz.org/mantis/view.php?id=4167) by nubinia 13:16:12 -!- Zaba has quit [Ping timeout: 276 seconds] 13:19:10 -!- Zaba has joined ##crawl-dev 13:23:01 !tell Cryp71c are you going to finish the fight rewrite for 0.9? If so, you'd have to merge it in in a week... I wonder that because there are bugs in trunk you have fixed, and on the other hand bugs in your rewrite, it'd be good to know which ones to fix. 13:23:01 kilobyte: OK, I'll let Cryp71c know. 13:35:06 does any of you have a clue which way tornadoes rotate? Mostly, whether the center or the outer parts spin faster. 13:36:25 (if this tidbit would be helpful, a vast majority of tornadoes on Earth rotate counter-clockwise) 13:43:11 well really their rotation is dependent on latitude, it's just that most tornadoes occur on a single continent in the northern hemisphere. not really sure where the most powerful winds would be, really, most real-world tornadoes aren't as clear-cut as crawl's :P 13:44:08 (wondering if the squarelos maintainers converted tornadoes into squares) 13:44:27 squarelos predates tornado 13:44:42 right, but I thought eronarn or someone was keeping it up to date 13:44:49 not really 13:45:20 speaking of up-to-date, it would be nice if !apt worked on octopodes, unless more changes to their apts are expected 13:45:52 !apt is a part of Henzell, managed by greensnark who is gone. 13:45:53 Could not understand "is" 13:46:32 Wensley: i made it compile, and fixed some bugs, but there are some other bugs i have not fixed 13:46:38 and i haven't maintained it since 13:46:56 Eronarn: let it die :p 13:47:11 really eronarn just put it on erocrawl to attract players to test his new races :P 13:48:21 I'd say the other way, by a good margin: new races are something that can attract a good number of people 13:48:31 03kilobyte * rffffd972c6c6 10/crawl-ref/source/ (6 files): Get rid of remaining potion curses. 13:48:42 03kilobyte * r9d5d59771bfc 10/crawl-ref/source/makefile: Make "package-source" modular. 13:48:42 03kilobyte * rdc727d600bc9 10/crawl-ref/source/ouch.cc: Get rid of some data mining spam. 13:48:42 03kilobyte * rfcd1cdd875ed 10/crawl-ref/source/ (4 files in 4 dirs): A vault showcasing tornado rotation. The only way in is to let yourself get carried away. 13:49:37 kilobyte: squarelos is great, and should be supported. i simply have not felt like coding lately 13:49:58 kilobyte: squarelos attracts people too 13:50:05 elliptic: half true: that's why i did it at the time, but i do actually like squarelos, too 13:50:20 kilobyte: you may not realize this, but players like having more natural gameplay :P 13:50:25 eronarn: yeah, I know 13:50:27 doh, conservation of angular momentum, I'm an idiot 13:51:02 elliptic: yeah, that's a big argument against squarelos for most people 13:51:07 no 13:51:21 I don't know where you are getting "most people" from there 13:51:50 -!- ZorbaTHut has quit [Ping timeout: 240 seconds] 13:51:57 if you have any support for such statements, please give it 13:51:59 I mean, people who consider the game to be a simulation rather than a board game 13:52:32 this includes those who are more used to non-roguelikes 13:53:23 but then, other roguelikes that have limited LOS do it in a square as well. And non-roguelikes with LOS too, like Civilisation and the likes. 13:54:10 bts 631547 13:54:13 ECHANNEL, sorry 14:00:06 man, I really hope bmh has the new abyss ready in time for 0.9 14:00:15 it is my most anticipated new feature 14:03:13 yeah, and we need to come up with a schedule, badly 14:04:33 basing on greensnark's analysis he posted around 0.7 release, if the tournament is to start on Aug 1st, the forking must happen no later than end of June 14:05:38 and the actual release before July 24th. 14:05:52 and then work begins on version not1.0 14:06:20 what if crawl's version number just asymptotically approached 1 with each major release 14:07:02 with new CDO regular builds of both trunk and the beta branch, the schedule is somewhat relaxed so we can play for July 24th not a week before (as there _always_ is a slip) 14:07:46 elliptic: there was an idea of making the tourney last three weeks rather than the whole month, do you know anything more? 14:08:29 I remember after the last tourney a lot of people were in favor of making it two weeks, with a few who preferred it be three weeks 14:08:54 kilobyte: everyone seemed to like the length of the last tourney 14:09:00 notably I can't remember anyone who wanted to see another full-month tourney, although that may have just been post-tourney fatigue 14:09:06 3 weekends, so 16 days 14:10:14 considering the number of open contruction sets, this means the second half of August :p 14:11:03 Aug 13-28 or something like that sounds reasonable enough to me 14:11:16 though of course it doesn't have to start and end on weekends exactly 14:12:44 could start on the 13th and go to the 31, and then we could think of something special to happen during the final 3 days, like enforce random combos only :P 14:13:23 the thing is that some people have actual jobs so having something special that is only during the middle of a week isn't great 14:15:31 well if you want three weekends exactly, starting and ending on weekends is the only way to ensure the shortest tournament, unless the tournament only takes place on weekends or something 14:16:05 oh, and something something sprint something something 14:16:12 03kilobyte * rc48dbd625e06 10/crawl-ref/source/colour.cc: Fix tornado rotation display going the other way as monster rotation. 14:16:31 or ZotDef 14:17:10 I guess there could be a sprint tourney in the last three days of august, yeah 14:17:14 which is still a disaster :( 14:17:17 zotdef still is as bugged as ever, yes 14:17:52 people were playing it the other day but that was just because trunk was bugged to give 400M xp :P 14:18:00 hahaha 14:18:07 :) 14:18:13 (and also limitless zot points) 14:19:27 I remember someone had an idea for a defense game in the vestibule of hell, where you play as geryon and have to keep monsters from entering the hell portals 14:19:34 -!- b0rsuk has joined ##crawl-dev 14:20:16 elliptic: if Tornado could rotate items, how abusable would that be? (with a slight nerf to balance this partial emulation of apportation) 14:20:45 unlikely it would be any more abusable than tornado already is :P 14:20:48 <-- thinks it would be cool if tornado would send items spinning away from the center 14:20:54 how would you handle items falling in water? 14:21:01 it currently does rotate monsters, which hasn't been tested much yet 14:21:28 I would love getting messages like "Eurasian Badger, a mighty KoBe of Trog, has just killed your OgFi ghost.". While playing online, of course. 14:21:43 And the opposite messages - when my ghost kills someone. 14:21:57 I guess that's more of a dgamelaunch thing than Crawl thing, though. 14:22:08 b0rsuk: you do get those messages, if you're in ##crawl :) 14:22:30 Wensley: I don't like monitoring ##crawl just for that :-/ 14:22:37 monsters get temp levitation until the duration expires so there's only one chance per cast... so items would have to be levitated as well 14:22:52 well where would you expect the messages to be :P 14:22:54 I wonder if something needs to be done to prevent item destruction 14:22:57 b0rsuk: does your irc client alert you when your name is mentioned? shouldn't need to constantly monitor 14:22:59 An inconvenient question: if monsters start flying in a Tornado, why are missiles unaffected ? 14:23:16 right, that reminds me, tornado should also give dmsl :P 14:23:20 b0rsuk: an idea is to skew targetting randomly 14:23:41 kilobyte: Won't it be too powerful ? Isn't there Deflect for that ? 14:23:46 this was meant for the caster only 14:23:48 kilobyte: it's probably okay if a tornado has the potential of dropping items in water or lava, just so long as it doesn't destroy runes 14:24:28 after all other powerful spells can destroy items (shatter and potions, say) 14:24:34 runes are light, they could float 14:24:38 right, Shoals and Lom Lobon 14:24:49 and swamp 14:24:50 and cocytus 14:24:52 and Swamp 14:24:56 Runes could float just because they're magical. 14:24:57 and Cocytus 14:25:19 runes could be unaffected by the winds because they're magical 14:25:23 Runes could float even if not in a tornado. 14:25:46 A barnacled rune shouldn't be affected by winds in any case. 14:26:08 b0rsuk: this could explain special behaviour of the Orb vs apportation 14:26:11 Gold is heavier than lead, so golden shouldn't either. 14:26:13 attached to the ground? 14:26:21 bone is light... 14:26:35 Particularly kenku bone. 14:26:44 as long as we're talking realism of crawl, shatter should destroy the bone and obsidian runes :P 14:27:32 If were' talking realism in Crawl, kenku shouldn't have a penis. 14:27:52 * b0rsuk wins. 14:28:36 uhm, swans do, don't they? (I've read a lengthly article about the artwork depicting Leda once) 14:29:25 so other birds probably do too 14:29:40 As a general rule, flying birds don't have them. They don't need any extra weight. That's what our high school biology teacher told us. She made us repeat that very often. 14:29:57 Or at least they have very small ones. 14:30:16 Just like they have pneumatic bones to make them as light as possible. 14:30:43 -!- dpeg has joined ##crawl-dev 14:31:30 !seen elliptic 14:31:31 dpeg: You have 1 message. Use !messages to read it. 14:31:31 I last saw elliptic at Fri Jun 24 19:30:50 2011 UTC (41s ago) saying sair: yeah, thats more efficient on ##crawl. 14:31:37 !messages 14:31:38 (1/1) kilobyte said (10h 50m 34s ago): if you don't give us a formula for the Singing Sword, you risk it having misleading behaviour in 0.9 14:31:54 will do, kilobyte 14:31:56 Does Strength affect hungering rate in combat ? Because I think it should, that way Strength would have utility for Ogres and Draconians. 14:31:59 dpeg: <3 14:32:56 dpeg: hi 14:33:02 dpeg: there's an important question: what with the Tourney? Should it last the whole month? Or three weeks? Or three weekends (16 days) like elliptic proposes? Or something even different? 14:34:15 Obvious - 27 days. 14:35:04 By the way, I'm 27 years old. 14:35:22 let's not tell dpeg what we were talking about; I looked it up and it turns out only waterfowl do, so basing this on the knowledge about Leda hit right the only exception :p 14:36:11 b0rsuk: and it will get worse and worse from that point 14:36:45 kilobyte: now I wanna know :) 14:37:03 re tournament: follow elliptic's lead, I'd say. Shorter is a good idea in any case. 14:38:22 elliptic: was great to see you take care of Che's weak spot (my fault, of course). I used the non-linear formula for stat boost on purpose, but your solution should work as well. Players will still be encouraged to go slow, right -- and perhaps it is not as trivial when to stop now. 14:38:32 elliptic proposes Aug 13-28 14:38:41 sure, sounds good 14:39:05 seafood can don a hat and be instantly maxxed 14:39:29 yes, but it's only ponderous 1 -- not good piety wise, is it? 14:39:38 plus they have to have a hat 14:39:53 ponderous 5, slowdown is proportional too 14:40:04 dpeg: yeah, I understood there were reasons for the non-linearity... but just using that formula looked rather strange on slot-limited races, who would go 2/6/15 14:40:17 kilobyte: you still need ***** to get max stat bonus 14:40:25 so not instant :) 14:40:28 elliptic: well yeah, of course 14:40:48 octopodes could use some special-casing though 14:41:36 and if you find a hat of int on D:1 and the temple next to D:4 upstairs, walking delay 20 will let the first orc priest smite you to death before you can reach him 14:41:44 just tested this :p 14:42:09 also if an octopode of chei mutates horns 3, they are really out of luck :P 14:44:01 haha 14:44:15 yes, I wouldn't worry about Oc too much right now 14:44:32 if need be, we can rule that Chei cannot accept them for whatever silly reason 14:44:36 right, I guess they aren't going in 0.9? 14:44:49 btw, many thanks for covering my ass with the baileys -- another flake of mine 14:44:55 -!- dpeg is now known as flakepeg 14:45:01 elliptic: I wouldn't mind either way 14:45:11 baileys? oh, you mean MarvinPA adjusting their loot? 14:45:15 yes 14:45:32 MarvinPA: Thanks! This is very much appreciated. 14:46:07 I played some and find them reasonably interesting. Cannot imagine they're overpowered, apart from the gamble for insane rings (which is much more interesting than the startup scums you sometimes see) 14:46:47 so, somebody wants to make a korean DCSS. Are we gonna see south-korean DCSS e-sports events, soon? https://gitorious.org/~blmarket/crawl/crawl-korean 14:48:01 bhaak: ooh, interesting. Other people wanted German and French versions too. 14:48:26 too bad this approach won't work due to grammar 14:49:31 bhaak: hana dul set! 14:49:56 Baduk, Taek won do, Crawl -- a natural progression to me 14:50:13 -!- ZorbaTHut has joined ##crawl-dev 14:50:36 flakepeg: numerals are easy! 14:50:37 Seafood walking around casually in nothing but a hat sounds perverse. 14:52:00 b0rsuk: you mean, more than using the graphic concept of japanese hentai tentacle monsters? 14:53:21 octopode of ashenzari was meant to happen 14:53:22 !tell galehar Currently magic skills and Inv/Evo are free to train at skill_cost_level 1... the reason for this seems to be the cost reduction happening in lines 400-410 of skills.cc. I think that cost reduction should be removed anyway, right? 14:53:22 elliptic: OK, I'll let galehar know. 14:53:36 kilobyte: yeah, the gettext approach can't really work well. i wonder what they're gonna try when/if they hit the limits of gettext 14:53:53 Wensley: <3 14:56:01 If there's no decent support for localized versions of crawl, there may be a fork. Then koreans won't be playing DCSS, they will be playing something that forked X years ago. 14:56:26 do you think they're aware of the limitations of their approach? should someone inform them before they waste a lot of time? 14:57:36 flakepeg: are there differences in the rules between baduk and go. i think, so far, i only know differences between counting/scoring 14:57:37 dpeg: about Skalds: there seem to be no decision and no more new arguments. What would you say: yes, no, futher discussion? 14:58:06 b0rsuk: you can use nethack's solution. don't develop dcss further ;-D 14:58:15 b0rsuk: there was talk about having it in 0.10... there are half-baked proposals that could actually work 14:58:41 bhaak: ah, so nethack-de and jhack can stay current :p 14:58:49 yes :) 14:59:25 -!- Pacra has quit [Ping timeout: 250 seconds] 15:03:30 Wensley: i don't know if they know about the previous talk about it. but they will notice fast that you can't expect gettext to work for every message in crawl so they probably won't waste a lot of time. 15:03:33 Does Crawl really use such a big amount of dynamic text ? There are combat messages, spells... 15:03:46 !tell edlothiol I just realized how low on time we are. Could you point me to a version that would be best for merging? I see there's no way we can release without it... 15:03:46 kilobyte: OK, I'll let edlothiol know. 15:04:00 No doubt a correct solution would end up producing non-human-readable code 15:04:08 b0rsuk: IIRC more than 1000 out of more than 3500 messages 15:04:28 (counting only mpr() and mprf() with a simplicistic script) 15:05:23 Maybe go with it to Google Summer Of Code ? Or similar ? 15:06:27 do they sponsor games at all? 15:07:14 yes, i remember wesnoth being part of it 15:09:31 b0rsuk: if you want a general solution, that probably is not possible. the problem is closely related to machine translation and for that computational linguistics proved it to be impossible to translate from a source language to an intermediate language and from there to the target language. 15:10:27 fortunately Crawl basically never has text that stretches more than a single message 15:10:30 but ... you don't need a solution that can represent every human language ... and you only havve in crawl only a subset of english 15:10:45 the only exception I can think of are Zin recites 15:11:20 and you can already dump every sentence that has no dynamic content 15:12:03 so the worst case would be messages that need custom code for a given language... which is already included in the last writeup 15:13:24 the general idea, with parts by me and jpeg, is to have "language functions" callable from translated strings 15:13:53 the biggest drawback is that this would increase the complexity of the code for the plain (==English) version 15:14:06 like, English has "s" appended to verbs if the noun is in 3rd person 15:14:36 but as a programmer I'd say somebody should try it for one language and when a new language comes along you can generalize 15:14:41 bhaak: it could even simplify it by removing all the complexity of DESC_THE_NOCAP and its ilk 15:15:16 bhaak: of course, there'd be a net increase in complexity, that's sure :( 15:15:42 although from experience, I'm sure there will be edge cases in certain language that you can't generalize. 15:16:17 but if you get a separate branch for every language and the branches share ~99% of the code ... I'd say this would still be a success 15:16:54 bhaak: exactly, yeah. In that case the translated message would consist of nothing but a call to a language function specific to that single message 15:18:05 Swap skeletons for lindwurms and blink frogs (https://crawl.develz.org/mantis/view.php?id=4168) by XuaXua 15:18:13 i've seen others proposing that approach and I think it could work quite well, with the drawback that the messages are a bit obfuscated and not as clear in the as they are now 15:18:18 there is a quite interesting problem elsewhere: god genders 15:18:35 hehe, nethack already has that problem in vanilla 15:18:50 English can avoid this issue by carefully wording messages, but there is utterly no way to do that in Polish 15:19:46 their (extreme hacky) solution, prepend female gods with a '_' and have code to handle that accordingly, e.g. "_Venus" 15:21:16 no sentient creatures can use neuter grammatical gender... the only exception is "child", and in general, referring to someone in neuter is a rare but severe insult (not implying being ball-less or impotent, but being not a human at all. Like, extreme cases of racism, etc) 15:21:51 no robot god planned? :) 15:22:00 bhaak: that's no different from human uniques, you already need a database to know that Agnes is a she and Ruper a he 15:23:34 "no sentient creatures can use neuter grammatical gender" -> only for names or also for other nouns? I mean that girl in "Lisa is a girl" could also be of different gender than "Lisa"? 15:24:08 because in german, that's possible and "the girl" == "das maedchen" is neuter 15:26:19 back 15:26:22 -!- flakepeg is now known as dpeg 15:26:55 also nouns; I can't recall any other case than "child" 15:27:04 bhaak: Korean go has not really different rules, although Ing (some rich Korean go) tried to promote his own rule set... to no avail, thank Trog. Chinese rules are different with scoring. 15:28:00 kilobyte: I replied on the forum. I am fine with either Crusader -> Skald and Warchanter. I will write a tentative background description later tonight (which will work tonight). That mail should also push the issue once more. 15:28:54 inanimate items have a quite randomly assigned gender. Like, "table" is masculine and "chair" is feminine. Getting that wrong is quite a glaring error, and can make it hard to understand a sentence (as the wrong pronoun would refer to something last mentioned with that gender, possibly several sentences before). 15:32:07 ah, so the gender of nouns is more closely coupled to the natural sex in Polish than it is in German. 15:35:06 it depends on the word, but generally, yes. For jobs for example, there are many that have two gender forms, but many use a fixed gender as well (usually but not always masculine) 15:35:54 like, "politician" which uses male gender even if the actual person is female 15:36:20 the default pronoun for an unknown person for example is male too 15:36:58 i think the last one is the same for all european languages 15:37:14 might even be an indoeuropean thing 15:38:14 kilobyte: is "god" or equivalent for any religion also universally masculine? 15:38:19 (in polish) 15:38:54 well, it seems like it's not the case for English... like, you say: "someone entered the door with a stick in THEIR hand" (not "his") 15:39:43 actually, I think a lot of strict grammarians would frown on using "their" rather than "his", although the sentence is perfectly understandable 15:40:13 singular they is controversial 15:40:27 Wensley: the word "god" has a masculine grammatical gender, but it doesn't affect proper names 15:40:31 or you could word the sentence "someone entered the door, stick in hand" and avoid pronouncs entirely 15:41:34 we could have a function that assigns gender randomly every time a god message is generated, like xom's altars. pray over ash's altar once and it's a "he", next time it might be a "she" 15:42:31 honestly, this might work in english as well, trying to avoid pronouns in all cases can be a little limiting :) 15:42:46 or we could adopt "ey" and the like 15:42:55 Wensley: but you have the same problem in english. "god" is masculine, goddess is feminine and if you want some gender-neutral you must use another word like "deity" 15:43:31 bhaak: any gender-specified word could be randomized such 15:43:32 that is not really limiting, try to avoid the character 'e' :-) 15:44:10 who books have been written without the letter e, I think avoiding all pronouns might be just as hard :P 15:44:15 s/who/whole 15:46:02 * kilobyte notices an awkward attempt to avoid e in "whole" :p 15:46:10 haha :) 15:50:42 Wensley: you did have some problems with catoblepases, any luck? 15:51:21 still trying to figure out why the cloud coloration function has no effect on the color of my petrifying clouds, but I'll definitely get it in before the fork, barring catastrophe :) 15:53:19 if you'd waste too much time, me or someone else among official devs can do that as well. I'm not implying anything about skills since it's not about general ability but merely familiarity with this specific code base -- but I have a feeling you try to do something that would be obvious to some. 15:54:10 -!- monqy has joined ##crawl-dev 15:55:12 of course, it depends whether you want to gain understanding of this code or just have this feature implemented 15:56:19 03elliptic * ra223f83c2acd 10/crawl-ref/source/religion.cc: Make Jiyva a little less likely to remove mutations. 15:59:10 I don't want to interrupt the grammatical discourse, but I'd like a quick poll on "Warchanter" vs "Skald". (No other terms allowed at this stage.) 15:59:11 kilobyte: I'll keep toying, because I'd like the answer to be as obvious to me as it is to everyone else :) but really the cloud color isn't too important, I'm just peeved that the function is inoperative. and yeah, I'd like to contribute more in the future, so understanding is key 15:59:42 one for Skald here 15:59:53 (if it counts) 16:00:11 reason being it sounds less artificial/awkward (in my opinion) 16:00:13 same here 16:00:37 and galehar seemed to prefer it too 16:00:43 I think Skald is really cool, and as a bonus there's no prior common mythological association so we can do whatever we want with it 16:00:46 skald 16:05:45 Wensley: As long as they sing, we're free to rock yeah. 16:06:15 skald 16:06:24 ok 16:06:39 warchanter 16:07:01 skald sounds nothing like the other roles 16:07:04 I like Warchanter a little better, but not enough to argue :) 16:07:06 "warchanter" wouldn't be one word, surely. or at least the name of the book would have to be changed 16:07:28 elliptic: seeing your Jiyva commit, what do you think about making all distilled potions temporary (like blood is for Vampires)? 16:07:43 i don't actually like warchanter, but i actively dislike skald (though on the other hand i *loathe* bard) 16:08:19 yes, bard is not an option -- apart from it being D&D it gives a really wrong impression on the background 16:08:31 skald is a D&D class too 16:08:36 dpeg: that would be pretty annoying to Tm and St, to make all distilled potions temporary 16:08:36 damn :) 16:08:37 wut 16:08:45 I have never heard of a skald d&d class 16:08:50 provide your source, eronarn 16:08:51 elliptic: this would fit with "what you produce is finite" and avoid the !Mut scumming 16:08:59 Wensley: it is a bard ... variant or prc, i forget. in either 3.0 or 3.5 16:09:09 elliptic: blod potions last quite a while 16:09:09 i could go find it if you really cared but i doubt you do 16:09:16 I will find it myself 16:10:09