00:10:52 -!- blabber has joined ##crawl-dev 00:11:52 -!- eith has quit [Ping timeout: 272 seconds] 00:46:57 -!- blabber has quit [Quit: leaving] 01:16:33 -!- Textmode has quit [Quit: Over the ages, man has been certain of a great many things. some of these things were even true.] 01:21:43 -!- ortoslon has joined ##crawl-dev 02:15:51 -!- felirx has quit [Ping timeout: 260 seconds] 02:22:46 -!- felirx has joined ##crawl-dev 02:27:39 -!- felirx has quit [Ping timeout: 248 seconds] 02:42:53 -!- felirx has joined ##crawl-dev 02:47:08 -!- casmith789 has quit [Remote host closed the connection] 02:47:28 -!- casmith789 has joined ##crawl-dev 02:49:39 -!- Amonchakad has joined ##crawl-dev 02:49:55 -!- by has joined ##crawl-dev 02:49:58 Is the draconian "growing" xl changed in trunk? 02:50:30 Because I reached xl 7 and I didn't change 02:53:40 i see nothing in git 02:53:41 hey rob, ama 02:53:44 ... Amonchakad. 02:53:52 hello 02:53:55 hi 03:08:42 Draconian not becoming mature at xl 7 (https://crawl.develz.org/mantis/view.php?id=1699) by Amonchakad 03:08:45 it's an older bug, if you choose your draconian through tab-select, it won't have a color 03:08:53 oh 03:09:07 if you dig through the BRs you'll find it 03:09:41 I just looked at the recent ones 03:10:34 -!- ortoslon has quit [Quit: bye] 03:12:43 ogaz: if it's an older bug, then why hasn't it been fixed? 03:12:50 I don't know 03:20:54 And, there's no easy way to restore one such bugged save, right? 03:22:12 -!- by has quit [Quit: leaving] 03:23:46 Issues with the new tiles game selection menu (https://crawl.develz.org/mantis/view.php?id=1700) by Amonchakad 03:32:58 -!- syllogism has joined ##crawl-dev 03:44:39 apparently there is an option to set a pizza flavor. 03:45:09 in what way? 03:46:23 There was. 03:46:29 I'm pretty sure it got deleted. 03:46:36 -!- ogaz has quit [Ping timeout: 240 seconds] 03:49:38 pizza is linley's direct rip of NETHACKOPTIONS=fruit:potato 03:49:52 which added an item to the game with a specific, user chosen name 03:50:08 nethack even went so far as for the item to keep its name in bones 03:50:16 I completely understand that he prefered pizza over slime molds 03:50:29 (except that you couldn't have more than 255 differently-named fruits in one save) 03:53:12 -!- ortoslon has joined ##crawl-dev 03:55:07 -!- Madtrixr has joined ##crawl-dev 04:07:27 -!- ortoslon has quit [Read error: Connection reset by peer] 04:16:42 -!- Niccus has quit [Quit: out] 04:25:40 -!- ortoslon has joined ##crawl-dev 05:31:30 -!- Mu_ has joined ##crawl-dev 05:36:03 -!- syllogism has quit [Ping timeout: 260 seconds] 05:43:26 -!- syllogism has joined ##crawl-dev 05:43:26 -!- Ashenzari has quit [Ping timeout: 258 seconds] 05:44:07 -!- Ashenzari has joined ##crawl-dev 05:44:07 -!- The topic of ##crawl-dev is: DCSS development channel! | Logs: http://tozt.net/crawl | People with +v have commit access. | Please keep general crawl-related chit-chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Ashenzari source: http://github.com/doy/crawlbot | Long stuff to pastebin.ca please. 05:50:04 -!- ais523 has joined ##crawl-dev 06:23:23 if a br is duplicate of something, one should tag it as duplicte of and close it right? 06:59:26 Yeah. 06:59:32 Remember to link it. 07:01:54 erm.. 07:02:11 Wizards have _spells_ to mutate a player? 07:02:29 polymorph other? 07:02:35 @??wizard 07:02:35 wizard (06@) | Speed: 10 | HD: 10 | Health: 20-60 | AC/EV: 0/13 | Damage: 6 | Flags: spellcaster | Res: 06magic(53) | Chunks: 07contaminated | XP: 708 | Sp: b.venom (3d15), mystic blast (3d15), invisibility, confuse, slow, teleport self / magic dart (3d5), paralyse, haste, b.lightning (3d17), confuse, teleport self / paralyse, crystal spear (3d28), blink, b.fire (3d18), b.cold (3d18), minor healing / paralyse, th 07:03:03 The wizard points at you and mumbles some strange words. 07:03:04 Strange energies course through your body. 07:03:04 A pair of horns grows on your head! 07:03:06 - 07:03:10 The wizard casts a spell at you. 07:03:10 Strange energies course through your body. 07:03:10 You feel stupid. 07:03:11 - 07:03:11 hmm doesn't look like it from that 07:03:23 how sucky.. 07:05:48 Is there any way to close and add duplicate status without generating two mails? 07:06:59 "change status to" "resolved" - it'll query for a duplicate ID 07:07:25 technically it's not resolved though codewise :( 07:07:46 choose a proper "Resolution" then 07:07:59 "duplicate" is a Resolution too 07:08:43 I really need to stop playing crawl.. these stupid mutations make me so angry. 07:10:02 time to code more stupid mutations then :p 07:10:35 well, you guys can reach me via query o/ 07:10:42 good luck 07:10:46 -!- Napkin has left ##crawl-dev 07:11:40 @whereis Napkin 07:11:41 Napkin the Hoplite (L13 MDBe), a worshipper of Trog, quit on D:15 on 2010-06-07 after 33376 turns. 07:11:45 !log Napkin 07:11:48 496. Napkin, XL13 MDBe, T:33376: http://crawl.develz.org/morgues/trunk/Napkin/morgue-Napkin-20100607-120847.txt 07:13:03 er... there's just a single harmless bad mutation 07:13:22 Hm, that log is weird. 07:13:27 Conserve : . (no shield) 07:13:30 and one (horns) that could be bad if you have an über randart helmet 07:13:33 Res.Corr. : . (helmet restricted) 07:14:19 What do shields have to do with conservation, and helmets with rescorr? 07:14:55 it's a table 07:15:23 Ah. 07:15:25 they just are on the same line. Like, there's rN in the same line as rCorr 07:15:33 Right, yeah. 07:15:54 A bit confusing, but I follow you. 07:15:57 I must be tired. 07:17:35 !tell Napkin Uhm, why would you quit due to "bad mutations" if you have a very good set with just one negative one that's basically harmless (-1 INT)? 07:17:35 KiloByte: OK, I'll let Napkin know. 07:17:52 KiloByte: Napkin just likes to ragequit every now and then, don't mind him :) 07:18:52 yeah, except that for a ragequit you need something to be enraged at 07:19:11 Not really :) 07:19:21 Well, the object of your rage doesn't have to be Crawl :P 07:19:31 * due rages. 07:19:38 -int can be fixed with a potion 07:19:54 but horns seem undesirable for this kind of character. 07:20:33 although he could probably wear an MR hat. :-) 07:21:12 he didn't even have a helmet in the first place, so at the time it was a strictly beneficial one 07:22:00 Lots of players really hate horns 07:22:11 horns are stylish. 07:22:12 I think the hate is totally overdone and silly, but they do 07:22:28 It's like people who don't use weapons of draining because of the xp loss :P 07:22:43 only issue i have with horns/antennas is the fact that all the good hats seem to generate as helmets when you get those! 07:23:12 they also suck exp. into UC, which might be something you don't want. 07:23:21 for a caster type, for example. 07:23:40 xale: If you're worried about that level of skill optimisation, you really shouldn't be playing Crawl :) 07:24:04 stabbing also sucks exp from fighting if you try to train on your own summons! 07:24:07 You just take a break until you feel better! 07:24:12 xale: you could turn the skill off 07:24:26 greensnark: I was going to say something like that. 07:24:31 ais523: You know people who complain that turned-off skills still train, right? :) 07:24:50 At that point I make soothing noises and shoot them with tranquilisers 07:24:57 lol 07:25:11 greensnark: can I complain that you can't turn a skill off before you reach level 1 in it? 07:25:20 it's stupid having to grind a skill just to be able to turn it off 07:25:23 ais523: Yeah, that's another classic complaint :) 07:25:36 and a sort of oxymoronic one 07:25:43 * greensnark says "There, there" and shoots ais523 with tranquilisers. 07:25:53 :)) 07:25:57 * due goes to bed. 07:26:08 Night bro. 07:26:12 (although, as an interface pain it's maybe worth considering, if not as a gameplay problem) 08:19:43 -!- ortoslon has quit [Quit: bye] 08:33:29 I took some artefact gloves of Nikola, they had rElec on them but he wasn't resisting elec. Was it just in his inventory and he wasn't wearing them or something? 08:34:01 Monsters have no glove equipment slot 08:34:24 Could be added easily though 08:34:53 they had MUT on too, does that do anything to enemies I wonder? 08:34:59 No 08:35:15 Lot of asymmetry in Crawl 08:39:21 that would be fun, changes their form every now and then 08:45:39 They should accumulate mutagenic radiation and then rush up to the player and give them a big hug 08:46:16 lol 08:47:11 You feel very strange. You feel a big rush of blood as a monstrous heritage opens up in you! 08:50:57 -!- ortoslon has joined ##crawl-dev 08:56:20 you call that "a hug", huh. 08:57:23 03kilobyte * r1dd4d098ea6b 10/crawl-ref/source/exclude.cc: Fix an incorrect comment. 08:57:25 03kilobyte * ra274bf62ed56 10/crawl-ref/source/makefile: Remove -fno-strict-aliasing, our code does not rely on such tricks. 08:57:27 03kilobyte * r2fd5618609e1 10/crawl-ref/source/xom.cc: Fix illegal initialization (bad in C++03 and C++0x). 08:57:29 03kilobyte * rb2525ca80704 10/crawl-ref/source/dat/des/variable/float.des: Remove my SourceForge name. 08:57:34 03kilobyte * r61007a94bd72 10/crawl-ref/source/dat/des/variable/entry.des: Update shiori_entry_whats_in_store_for_you for new player-specie monsters. 09:05:02 !tell due I have a problem with a vault (branch entry_desc_problem) failing with "MonPropsMarker should never be saved" on game start but not on &L. You once fixed such an issue before -- but I don't remember what the fix was. 09:05:02 KiloByte: OK, I'll let due know. 09:13:06 -!- ortoslon has quit [Quit: bye] 09:15:40 03greensnark * r49bbc248a8ac 10/crawl-ref/source/ (terrain.cc terrain.h travel.cc): Tweak slime wall neighbour cache and precompute slime wall neighbours only once when updating travel cache. 09:22:20 -!- ortoslon has joined ##crawl-dev 09:25:41 -!- Cryp71c has joined ##crawl-dev 09:27:11 -!- sorear has quit [Read error: Connection reset by peer] 09:31:22 -!- by has joined ##crawl-dev 09:31:34 !lm * type=crash 09:31:34 8. [2010-06-04] Helsbecter the Thaumaturge (L8 MuWz) ASSERT(!crawl state.doing prev cmd again || (crawl state.is repeating cmd() && crawl state.repeat cmd == CMD PREV CMD AGAIN)) in 'main.cc' at line 3887 failed on turn 10869. (D:7) 09:31:36 -!- eith has joined ##crawl-dev 09:32:58 -!- sorear has joined ##crawl-dev 09:33:59 crawl probably should put the game type in the save name 09:35:35 greensnark: going to ball lightning shoals:5 if you want to watch 09:35:56 st_: I don't watch shoals visits with xl>16 unless it's a splat :P 09:38:30 packed save handling looks a little scary... 09:38:56 we need more crashes :P 09:39:28 syllogism: looking to abuse them? or just for the entertainment value? 09:39:37 entertainment 09:39:42 03greensnark * re4a312aaf547 10/crawl-ref/source/travel.cc: Cache is_travelsafe_square results for all map cells before updating interlevel travel cache. 09:40:07 good thing is now that we have crash milestones, we can track down crash abusers in the tournament :) 09:43:29 mmm reproducible crash + potions of mutation/acquirement :p 09:44:39 03kilobyte * r19ee085b01a8 10/crawl-ref/source/abl-show.cc: Remove the block for Lugonu's self-banishment in Pan. [1172] 09:47:14 just smash ` 09:50:44 -!- xale has quit [Quit: xale] 09:50:54 that doesn't crash for me 09:51:22 apparently Cryp71c can make it crash, but wants to keep the details to himself 09:51:33 -!- xale has joined ##crawl-dev 09:52:56 -!- ortoslon has quit [Quit: bye] 09:53:30 by, lol I assure you, I have no details to keep from you. 09:54:03 ah, maybe you missed my question in your BR? 09:54:12 I can tell you, though, that when I use ` to repeat stuff, its almost ALWAYS spell castings. 09:54:26 by, I did, I don't recall having seen any such question. Let me go take a look. 09:56:07 by, "I believe I was repeat-casting a spell or wand at a creature, though I can't say what the creature was or provide a link to the crash log. 09:56:07 " 09:59:15 oh looks like it's not as crash prone as it was a while ago 10:03:41 I fixed the reproducible one (there was an issue for repeating &X) 10:09:46 backspace doesn't work on macro prompt anymore? As in it thinks I want to macro it 10:11:04 -!- xale has quit [Quit: xale] 10:11:41 yes 10:12:31 za.``N` 10:12:58 i.e., respond to the the "really target yourself prompt" twice with `, then abort, then do again 10:17:59 think I'll just trash the direction special repeat code 10:21:38 !tell dpeg Don't suppose we could clean up DS Brainstorming, eh? Remove all the already implemented/definitelly won't do mutations from that list? 10:21:38 Cryp71c: OK, I'll let dpeg know. 10:25:51 Antennae keep bugging me, shouldn't they detect creatures that are invisible to you, but within your LOS? 10:26:11 03by * r32bcd3415375 10/crawl-ref/source/ (directn.cc directn.h): Trash direction_chooser special command repeat code. 10:38:36 -!- by has quit [Quit: leaving] 10:47:35 -!- ortoslon has joined ##crawl-dev 11:02:46 -!- ortoslon has quit [Quit: bye] 11:08:41 -!- by has joined ##crawl-dev 11:10:50 what's the strange vertical dark blue line http://i46.tinypic.com/ezib2s.jpg it's somewhat difficult to see, but it's a bit right of @ 11:11:16 doesn't seem to happenw with the default console font (Raster Fonts) 11:12:43 -!- blabber has joined ##crawl-dev 11:15:41 I don't see anything 11:16:48 it's there, really thin 11:17:09 follow the unmapped corridor to right of @ 11:19:51 I can see it if I crank the brightness up, but it's probably a terminal rendering issue 11:20:46 Also: if you can see that normally, your screen is way too bright :P 11:20:50 :P 11:20:57 it's easier to spot ingame :P 11:21:27 it's probably nvidia digial vibrance that's making it brighter 11:23:53 what video card do you recommend for crawl console? 11:24:27 s3 trio 11:24:31 go! 11:25:46 by: Didn't you fix #591? Can we close it? 11:26:31 no, but feel free to close 11:26:47 You didn't fix it? :P 11:26:56 I thought monsters now chop down plants and stuff to get at the player 11:28:13 they do, but not if they find a "good move" that doesn't require chopping down plants, as in that situation 11:28:25 Aha 11:29:40 might try the suggested fix (only consider moving closer to the player) 11:30:00 First we apparently need a nice way to place a monster on a precise point 11:30:15 Wasn't there a wizmode command that allowed you to use 'x' and place a monster exactly at the cursor 11:30:22 m 11:30:54 x, target monster, m 11:31:14 Ah, nice 11:36:57 03by * rc163fd603c8b 10/crawl-ref/source/main.cc: Abort instead of asserting for recursive command re-do. 11:37:13 maybe we really should trash the repeat code 11:37:28 Wouldn't bother me :) 11:37:44 I got a few crashes, but couldn't really figure out how :P 11:38:32 I thought this must be old code, but it isn't 11:40:58 -!- ortoslon has joined ##crawl-dev 11:49:51 maybe it can be rewritten for the client side once the client/server split is done 11:57:21 -!- Poor_Yurik has joined ##crawl-dev 11:59:49 KiloByte: yay, I can reproduce something close to your starvation endless loop in 0.5 now 12:01:57 I think elliptic has a pretty good point in ##crawl :) 12:05:11 being skeptical about new DS, or that duplicating effects is fine? 12:09:01 !lg kilobyte char=mdfi place=swamp:4 -tv:<3 12:09:02 1. KiloByte, XL16 MDFi, T:48837 requested for FooTV. 12:09:24 !lg kilobyte char=mdfi place=swamp:4 -tv:cancel 12:09:24 1. KiloByte, XL16 MDFi, T:48837 cancel requested for FooTV. 12:09:27 !lg kilobyte char=mdfi place=swamp:4 -tv:<3 12:09:27 1. KiloByte, XL16 MDFi, T:48837 requested for FooTV. 12:09:40 !lg kilobyte char=mdfi place=swamp:4 -tv:<3 12:09:40 1. KiloByte, XL16 MDFi, T:48837 requested for FooTV. 12:11:26 by: Why the DS redesign was such a pressing matter :) 12:11:43 What's kilobyte up to in the swamp anyway 12:12:33 03Cryp71c * rd6a1b5e301d3 10/crawl-ref/source/fight.cc: Change claw aux dmg to scale with claw level 12:14:03 greensnark, as I told him..it wasn't a pressing or urgent matter, but a race which is constituted by regurgitated effects from everywhere else in the game is a terribly lazy design. 12:14:19 Not that their old design was bad or that the designers made poor choices or had bad direciton. 12:14:47 But when a more unique approach is available for something as large as an entire playable race, the unique option allows for more varied and entertaining play. 12:15:11 I agree that DS could be improved 12:15:32 I'm not convinced about the timelines here, though 12:15:40 Particularly whether they'll be in good shape for 0.7 12:15:55 greensnark, that is a legitimate concern, certainly. Especially with 0.7 hosting the tournament. 12:17:26 -!- morik has joined ##crawl-dev 12:18:46 -!- Textmode has joined ##crawl-dev 12:20:33 morning all 12:21:04 -!- ais523 has quit [Remote host closed the connection] 12:21:38 morning all 12:22:28 * Textmode is apprently in a loop of saying 'morning all' 12:22:35 hrm, can't watch kilobyte's ttyrec because of unicode failure 12:24:32 Unicode failure? 12:25:07 display is messed up, and no combination of LANG and LC_* variables seems to fix it 12:26:05 Hmm, there's some hardcoded utf-8 stripping in footv, might be messing it up 12:26:12 !lm KiloByte mdfi swamp:4 -tv:<3 12:26:13 No milestones for KiloByte (mdfi swamp:4). 12:26:22 !lg KiloByte mdfi swamp:4 -tv:<3 12:26:23 1. KiloByte, XL16 MDFi, T:48837 requested for FooTV. 12:26:23 no, not footv (that's fine), I downloaded the ttyrec 12:26:44 probably need some hardcoded utf-8 stripping locally 12:27:00 What terminal are you using? Terminal.app? 12:27:07 urxvt 12:27:15 Oh, that should be fine then 12:27:15 Terminal.app isn't better, though 12:27:32 Yeah, Terminal.app is a broken mess on 10.5 and below 12:27:35 10.6 is a little better 12:28:07 !ttyrec KiloByte mdfi swamp:4 12:28:08 1. KiloByte, XL16 MDFi, T:48837: Can't locate remote ttyrecs 12:28:12 Brrh 12:28:15 Got a link for the ttyrec? :P 12:28:30 http://crawl.develz.org/ttyrecs/KiloByte/2010-01-27.13:26:43.ttyrec.bz2 12:28:39 that might not be the one, but it has the problem 12:29:13 though I expect something to be wrong with my setup rather than ipbt or the ttyrec 12:30:38 12:30:40 sorry 12:31:41 tried ttyplay instead of ipbt? 12:32:03 no 12:33:56 ttyplay actually works in Terminal.app, but not urxvt 12:35:43 need to go for a bit now -- thanks for your help! 12:56:14 03Cryp71c * rfa4d29f457d4 10/crawl-ref/source/mutation.cc: Slight nerf to demonic guardian mutation 12:58:40 try s/inst->cfg.utf8_override = FALSE/inst->cfg.utf8_override = TRUE/ or s/strcpy(inst->cfg.line_codepage, "")/strcpy(inst->cfg.line_codepage, "UTF-8")/ in the ipbt source 12:59:04 looks like ipbt might be never turning utf-8 support on in its internal putty code 13:04:01 -!- Luca__ has quit [Remote host closed the connection] 13:04:18 -!- Luca__ has joined ##crawl-dev 13:47:50 Luca__: thanks; the line_codepage = "UTF-8" helps a lot (no more multiple characters per glyph, just a wrong mapping) 13:51:55 is breaking save compatibility for map_knowledge refactoring acceptable, or is that to be avoided? 13:52:01 (i.e. raising TAG_MAJOR_VERSION) 13:52:29 Feel free to break compatibility if necessary 13:52:49 Save compatibility must be retained in stable branches; master is open season 13:53:01 That said, if it's easy to retain compatibility, it's worth doing 13:54:01 03greensnark * r156635a1e972 10/crawl-ref/source/ (5 files in 4 dirs): [591] Test case to fail if a monster behind plants jumps back and forth between a set of locations when trying to move to the player. 13:57:59 Luca__: yeah, definitely acceptable 13:58:05 -!- ortoslon has quit [Read error: Connection reset by peer] 14:00:43 -!- Niccus has joined ##crawl-dev 14:25:53 -!- blabber has quit [Quit: leaving] 14:26:30 while you're at breaking saves, someone should add gametypes to savefiles 14:28:08 03Cryp71c * r035a96aadae3 10/crawl-ref/source/mon-stuff.cc: Allow powered by death to be triggered by pets 14:43:22 -!- Luca__ has quit [Remote host closed the connection] 14:43:37 -!- Luca__ has joined ##crawl-dev 14:48:24 in case anyone's interested, ipbt has seen some development at http://github.com/jmazon/ipbt.git 14:48:43 the changes sound promising re unicode, though they didn't help for me 14:57:12 okay, the last commit doesn't sound so promising: "Partially fix bug; rebrand it as a feature." :-D 14:58:20 well, yes, but it does also say something about unicode terminals, and it uses ncursesw 15:01:09 looking at http://github.com/jmazon/ipbt/commit/a04f94503aa65c055842c4fac8c843e5bca9ddd8 it seems you need to have your LC_* set up correctly 15:01:43 but as you already noticed there are a lot of variables that might break :( 15:02:49 KiloByte's ttyrecs all have 'c' for the trees and 'K' for the water in swamp 15:04:21 nice, clang compiles crawl after some trivial patches 15:04:32 apparently correctly 15:04:48 How fast is it compared to g++? 15:04:57 Compile-time, I mean 15:05:02 Luca__: do want! 15:05:19 haven't measured 15:05:35 hey, moment, "trivial patches" must involve changing the Makefiles, negating the "trivial"! 15:06:03 Luca__: have you tried running crawl through the static analyser? 15:06:07 no 15:06:29 the only real problem is that you must use -MD instead of -Wp,-MD for dependencies 15:06:57 -!- Amonchakad has quit [Quit: When two people dream the same dream, it ceases to be an illusion. KVIrc 3.4.2 Shiny http://www.kvirc.net] 15:16:23 by: I get 'K' for the water and 'H' for the trees. No 'c's around :) 15:17:43 -!- Fangorn has quit [Quit: Fangorn] 15:17:44 ah, actually I get 'H' for shallow water and 'K' for deep water 15:18:34 KiloByte's current rc defines shallow water as 0x2248, deep water as 0x224B 15:18:53 trees are at the default 0x2263 15:19:50 so those are just the low bytes 15:22:05 patches at http://repo.or.cz/w/crawl/crawl-nettiles.git/shortlog/refs/heads/for-crawl 15:22:11 some fixes for clang 15:22:19 plus marshalling of variable sized ints 15:22:45 (to be used by the refactoring patches for the client/server split) 15:22:56 if I look into the ttyrec, I see K" K=0x48 "=0x22. An endian issue? 15:24:18 argh, strike that, I was being stupid 15:24:26 http://repo.or.cz/w/crawl/crawl-nettiles.git/shortlog/refs/heads/clientserver has the currently done part of the client/server refactoring 15:24:34 (not guaranteed to work at all for now) 15:26:53 wouldn't it be a good idea to store SP_BASE_DRACONIAN in you.species, and the future draconian color somewhere else? 15:27:18 or perhaps even roll it once you mature (and save immediately) 15:28:26 yes, that sounds sensible; I think they used to actually have different aptitudes 15:28:41 (before XL 7) 15:29:19 wasn't the plan to completely hide that? 15:29:25 hence, use the base aptitudes instead 15:29:38 yes, that just explains why it is how it is currently 15:29:44 the code seems to do that btw 15:29:53 end of species_apts 15:30:33 yes; but they used to have different apts, so it used to be necessary to have you.species decided at game start 15:31:37 the late rolling doesn't work btw because you can be level drained 15:32:05 hmm... should a draconian really revert to immature form on level draining? 15:33:33 I'll look at merging the clientserver branch -- it'd be nice if someone else could check the compile and marshalling stuff? 15:35:12 the compile and marshalling stuff should be quite easy to review 15:35:38 the clientserver branch instead is very intrusive, lightly tested, and changes the game behavior 15:36:09 hmm, should we wait for 0.7 to be branched do you think? 15:36:20 no idea 15:36:55 when should 0.7 be released? 15:39:45 as soon as possible, I guess; there's some desire to release it "in time before the tournament" 15:57:10 Luca__: the LOS part of the level map retains the normal colours in clientserver, while it should mostly be greyed out 15:57:41 you mean that parts in the los rectangle but not visible are not grey? 15:58:10 visible things in LOS are not gray on the X map 15:59:07 Luca__: what about not storing the future colour anywhere, and not rolling it until you actually reach XL 7? 15:59:08 there were also some other weirdnesses with stair colours, but I want to make sure we didn't break them in master recently first 15:59:17 KiloByte: you can get level drained 15:59:39 so you would eventually need to store it, thus it's probably simpler to store it from the start 15:59:55 by: it's quite likely clientserver broke it 16:00:07 is the X map thing a bug or a feature? 16:00:28 everything gray? feature, at least I'm used to it and it looks weird otherwise 16:00:49 btw, what's the reason of a fully gray map? 16:00:49 but the colour transform could just be done in the X map code 16:00:54 wouldn't a fully colored map be better? 16:01:12 yes, sure 16:01:15 Luca__: nothing else removes any abilities on level drain 16:01:55 Luca__: gained mutations, gained stats, gained hp/mp (minus four), anything else, stays 16:01:59 hmm right, then it should work 16:02:08 only problem, the game must be saved before the player knows the color 16:02:19 otherwise you can cause the server to crash after you gain xl 7 16:02:39 (after rolling the color, before telling the player) 16:04:38 by the way, are these draconian thoughts related to the bug reports? 16:05:52 no 16:06:03 they are related to looking at how to serialize struct player 16:06:08 ok 16:07:25 choosing at XL 7 sounds fine 16:15:51 greensnark: the missing travel cache updates means exclusions on the current level aren't saved unless you leave via stairs 16:16:35 greensnark: does a level info update at the start of files.cc:_save_level sound good? 16:18:45 -!- Enne has joined ##crawl-dev 16:22:01 by: coloring issues should now be fixed 16:22:51 any other issue? 16:23:32 btw, the map can be made fully colored with a single-character source change 16:23:54 get_cell_glyph(env.map_knowledge(c), Options.clean_map, -1) -> get_cell_glyph(env.map_knowledge(c), Options.clean_map, 1) 16:29:55 I didn't run into anything else yet, but I'm currently trying to fix some travel cache bugs first and getting detoured 16:31:47 re the travel cache, doesn't is_travelsafe_square check for blocking monsters in a weird and potentially broken way? 16:32:28 i.e. it seems to do the calculations on the real monster, but only if you saw it and checks that the map knowledge monster type is the same as the real one 16:33:48 that may well be -- it's only recently that it's even trying to work on map_knowledge instead of env.grid 16:35:37 map_knowledge almost surely doesn't have enough information in the current form though 16:35:53 the clientserver version instead should let you do anything you should be able to do 16:40:13 btw, would allowing the player to recognize that two monster sightings are the same be considered a bad thing? 16:40:30 (in practice, removing the rememember monster if it comes in view again) 16:42:30 https://crawl.develz.org/wiki/doku.php?id=dcss:usability_project:report , Problem 6.7 16:44:57 crawl does assume you can distinguish monsters in some sense 16:47:53 there was a drac color leak: gray ones gained a stat boost at level 4 (,7,10,+3), while other colours get them only at 8 (,12,+4). 16:48:42 not giving other colours the stat at level 4 looks like an omission, and they're one of weakest races, so I'll just give the stat at every level % 4 16:51:10 so 6.7 is something that should be fixed also in the non-edge-of-los case? 16:51:36 why does it assume you can distinguish monsters? 16:51:56 is there any way to distinguish them right now? 16:53:09 for example, you only get a "comes into view" message once per monster 16:54:08 right 16:55:35 that may be the only thing however 16:56:45 indirectly, you remember things about equipment, so if you see an orc with a mace of draining, you know it's the one that hit you before 16:58:09 so perhaps we should give a monster id to the client 16:58:17 and clear previous sightings on new ones 16:59:04 probably it doesn't make much sense to expose a monster number in the client though 16:59:35 maybe a randomized description could work there, like the ones for pan lords 16:59:49 but probably no one cares anyway, and it's a lot of work to write that for all monsters 17:00:17 -!- Cryp71c has quit [Quit: Leaving] 17:09:13 also, targeting remembers the previously targeted monster by id, similarly you.pet_target 17:10:38 Actually, sending a monster ID to the client would actually make client-server code easier. 17:11:25 we might not want these ids to be assigned sequentially, though 17:12:55 I think they should be assigned the first time they are sent to the user 17:13:10 so they are sequential, but in the player timeline, not the internal one 17:14:37 or equivalently 17:14:45 the first time the monster is seen 17:17:50 btw, rolling the draconian color after character creation doesn't seem a great idea after trying 17:18:07 the problem is that if you don't save at the right moment, the player could be encouraged to crash the server 17:18:36 and if you do, you risk creating an inconsistent savefile since level_change is called from several places 17:25:36 Luca__: it's a tremendous code simplification, already >100 lines removed. And I don't think we should put too much effort into removing any possible gains from crashes. 17:25:52 hmm are implementing that too? 17:25:55 you* 17:26:12 since crashes will always be exploitable if we don't save before every single message 17:26:24 btw I don't see how you save 100 lines by rolling late 17:26:34 yeah, and the apt display looks so much better now 17:26:37 you can save them by you.species = SP_BASE_DRACONIAN, but that's separate 17:27:01 I'm doing that too... 17:27:22 -!- by has quit [Quit: sleep] 17:27:53 mine saves 52 lines currently 17:27:59 but I didn't merge gray and the others 17:33:03 took a break for a few days from tile art, back to work on it now :) 17:34:06 * Enne cheers. 17:37:06 KiloByte: how about if (you.species == SP_PALE_DRACONIAN && you.experience_level > 5) 17:37:21 pale dracs seem to get steam resistance at level 6 17:37:28 how about moving that to level 7? 17:37:56 (or more specifically, removing the check, and setting you.species = SP_BASE_DRACONIAN before xl 7) 17:39:19 not only pale, also mottled draconians 17:46:11 of course 17:46:41 lemme push what I've changed so far 17:47:07 03kilobyte * rbefb9a82fc5c 10/crawl-ref/source/ (12 files in 2 dirs): Don't use hacks for hiding draconian colour, change the species at level 7. 17:47:09 03kilobyte * re9f9e3f36741 10/crawl-ref/source/ (20 files): Simplify species_name() now that draconian hacks are not needed. 17:49:15 black draconians got rElec++, too, since they got both the mutation and a hard-coded check 17:50:52 KiloByte: your code won't run grey draconian specific code when they get to level 7 17:51:04 because you assign the color in the case for non-grey draconians 17:51:32 so it seems on first sight at least 17:52:46 yup, probably has no effect though 17:53:19 ah, you do a weird hack where you check for gray in the non-gray case 17:53:29 there's only one specific thing to do, the stat gain 17:53:31 yeah 17:53:32 KiloByte: how about checking for SP_BASE_DRACONIAN and assigning before the switch? 17:53:52 the level 7 gain is a special case anyway 17:54:17 they get bonuses at 4, 7, 9, 12, [+3] 17:54:45 with irregular gaps 17:56:14 btw, my patch is at http://pastebin.com/Z8WjGH55 17:56:28 very similar, except I add you.hidden_species and still roll at start 17:57:15 btw, shouldn't you squash you.species in tags.cc when loading for compatibility? 17:57:40 I don't see what's the reason to wait with granting the resistance until level 14/18. What about just giving rF+ to reds the moment they mature? 17:58:50 you.hidden_species may be a good idea, yeah 17:59:57 but I'm not sure about random mutations and acquirement, it may be better to force a save during big random things, though 18:00:22 is the current savegame code callable at arbitrary points though? 18:00:43 problem is you need to call between assigning and sending the messages out 18:00:57 wait, how much memory does crawl use on average? 18:01:05 and currently there seem to be mashed together 18:01:29 how about changing the code so that all messages are buffered until the end of the turn, and saving then? 18:01:54 right now I think you can e.g. block at arbitrary points on --more--, right? 18:02:22 that patch breaks save compat 18:03:14 mine? 18:03:15 --more-- will allow SIGHUP and thus proper save 18:03:21 or yours? 18:03:47 marshalling needs to be protected by a version check 18:04:09 hmm what you are referring to exactly? 18:04:39 in my patch, I just marshal you.hidden_species instead of you.species 18:04:46 the savegame format does not change 18:05:10 and if max_level > 7, I set you.species = SP_BASE_DRACONIAN on loading 18:05:19 ah doh, indeed. 18:05:24 yours doesn't seem to do that, which I this is a problem 18:05:36 i.e. old immature draconians will get their color revealed 18:05:42 think* 18:06:59 let me create a yours2mine diff 18:07:07 aka current master to mine 18:10:13 KiloByte: http://pastebin.com/fBBTWH1j 18:10:15 wait, how much memory does crawl use on average? 18:11:15 hmm, that's wrong for non-draconians 18:11:37 100something for tiles iirc 18:14:19 KiloByte: fixed in http://pastebin.com/crGbuB6e 18:16:01 03kilobyte * r4dfb8e7ee213 10/crawl-ref/source/player.cc: Salvage games with buggy dracs in progress. 18:16:03 03kilobyte * r0150219d1967 10/crawl-ref/source/tags.cc: Hide immature draconians for old saved games. 18:16:04 03kilobyte * r3f3a3ecef2c4 10/crawl-ref/source/output.cc: Drop a couple of unneeded xl checks. 18:21:27 KiloByte: only rolling at game start is different now, see http://pastebin.com/Ehn8YPki 18:21:36 (btw, the previous patch was missing an hunk) 18:23:40 correction, see http://pastebin.com/kXSunKHX 18:23:58 not sure about this, rolling at start prevents savescumming and deliberate crashing 18:24:13 but makes things slightly more complicated 18:25:13 at any rate, this is a great improvement either way 18:25:14 deliberate crashing can be fixed by removing the ` command 18:25:32 savescumming... is not an issue on servers 18:26:02 sure, hopefully :) 18:26:39 of course perhaps the "proper" solution is to save a salt and use a cryptographic hash to compute stuff like the draconian color from it 18:27:04 oh please, now you're just getting carried away 18:27:19 been there done that, it's not nearly as awesome as it sounds 18:27:20 for instance you could compute random stat increases as (hash(saved_salt + "_stat_" + xl) % 3) 18:27:47 as well as use that to regenerate dungeon level procedurally in theory 18:28:45 are you aware that dungeon levels are mutable 18:29:58 in theory, you could store a difference 18:30:41 but no, it's not worth doing it for crawl dungeon levels 18:30:43 *facepalm* 18:31:15 although it could have the interesting application of making the game give the same output for a given seed and input sequence 18:31:30 and give similar output in case of minor modification 18:31:49 I beleive you're looking for #rgrd on QuakeNet 18:32:00 which could be useful to automatically test if a commit not supposed to change things actually does that or not 18:32:16 Jeff Lait is there playing every trick in the book to make roguelike saves fit in a 32KiB SRAM window on the original GBA 18:35:17 btw, something that could be interesting is support for unlimited persistent levels 18:35:31 none of the major roguelikes has this as far as I know 18:36:43 also infinitely large levels 18:39:46 I guess you don't count ZAngband as "major" 18:39:56 also it's two dozen subvariants 18:41:08 Also, have you ever seen the pre-0.4 randart engine? 18:43:03 Basically, it hashed the item's enchantment (now you know why they aren't enchantable), type, and "brand", used it to seed a PRNG, then rolled the randart's properties, every time the item was used 18:43:41 0.4 moved to a somewhat more straightforward (if memory wasteful) approach of attaching a hash table to every item and storing mappings like "RAP_TELEPORT" => 1 18:45:59 mimics did the same 18:46:10 does zangband have infinite persistant levels? 18:46:19 or just a procedural wilderness? 18:46:41 including randart generation, all until pre-0.6 18:49:48 Luca__: the wilderness is an infinite persistant level 18:50:25 I think the only thing currently done with a hash is colouring trees 18:50:33 saved or hashed? 18:51:14 hashed, as Luca__ was proposing 18:51:20 -!- Luca__ has quit [Remote host closed the connection] 18:51:36 -!- Luca__ has joined ##crawl-dev 18:51:39 hashed, as Luca__ was proposing 18:54:17 well, obviously, unless the game is drastically changed (e.g. much larger levels, persistent pandemonium), it doesn't make much sense to dynamically generate it 18:55:04 right now, it's about 120 70*70 levels, so at 8 byte per cell, it is only 5 MB per game 18:55:45 (8 byte is a vague approximation) 18:55:54 and you only have one save per game 18:57:52 more like 20MB, but yeah 18:57:57 -!- syllogism has quit [] 18:58:15 saving all vaults used and all randomness involved would be nasty 18:58:28 plus, there's digging, corruption, and stuff 18:59:11 for the record, saves are 200-200kB 19:00:30 200-220 19:00:55 KiloByte: before or after we zip them? 19:02:22 raw unpacked 19:02:53 they compress damn well 19:07:35 a sample save: .tar 11458560, .tar.gz 1267855, .tar.xz 854196, .tar.bz2 839197 19:10:13 all of those numbers look quite a bit larger than 200-220 kB 19:10:17 what am I missing 19:10:48 point of information: rax' server has about 3000 active games. on a 150GB disk 19:11:21 (which, intriguingly, puts the available space per save at about the same as the available RAM per game instance) 19:12:56 these numbers are for an entire game, not for a single level (the save from bug 1493, to be exact) 19:13:59 of course, most games will have only D:1-3 generated 19:22:19 -!- paxed has quit [Read error: Operation timed out] 19:24:16 does that 150gb disk also host all the (much larger) ttyrecs, or just saves? 19:25:33 -!- paxed has joined ##crawl-dev 19:27:25 -!- Mu_ has quit [Quit: Defecator, may everything turn out okay so that you can leave this place.] 19:32:16 Luca__: Also hosts ttyrecs. 19:32:22 I actually need to make it bigger before the tournament. 19:32:33 I've upgraded the disks but not yet the partition. 19:32:41 TOURNAMENT YAY. 19:32:42 due: You have 2 messages. Use !messages to read them. 19:32:45 Oh crap. 19:32:46 !messages 19:32:46 (1/2) turnerjer said (2w 4d 22h 29m 8s ago): that never happened.. it is a fragrant slanderosity. 19:32:48 !messages 19:32:49 (1/1) KiloByte said (10h 27m 45s ago): I have a problem with a vault (branch entry_desc_problem) failing with "MonPropsMarker should never be saved" on game start but not on &L. You once fixed such an issue before -- but I don't remember what the fix was. 19:33:02 Hm. 19:33:20 KiloByte: That *is* strange. I think it was Matthew who actually fixed it, but I'll try to have a look. 19:34:01 /dev/sda3 148G 137G 3.7G 98% /chroot 19:34:05 That's getting kinda dangerous. 19:34:10 Eek. 19:34:14 I should do that this weekend. 19:35:00 That might be a good idea. 19:35:05 How is the rax? :) 19:45:17 raaaaaaaax :O 19:46:00 137 + 37 = 148 *blink* 19:46:18 *3.7 19:47:51 extsomething with a reserved part? 19:48:09 Maybe just fragmentation, I don't know. 19:48:25 That ought to be counted under used space 19:48:33 that made sense in 1970 when disks were 20MB big, but today, reserving several GB is pure waste 19:48:51 Then again, I will never trust df again after that time it told me I had -37 PiB of free space on /dev/hda 19:49:04 rax: you planing to clear out files, or add more diskspace? 19:49:32 17:32 <@rax> I've upgraded the disks but not yet the partition. 19:50:00 ah 19:50:14 * Textmode scrolls up 19:57:10 I'm sort of obsessive about keeping the data. 19:57:15 I'm considering backing it up offsite. 19:57:30 The ttyrec archives, all indexed and playable via Henzell, make me very happy. 19:58:40 so we can watch the mummy farming bots of old... 19:59:13 Mummy farming compresses *very* well. 20:46:56 -!- Fangorn has joined ##crawl-dev 20:51:53 03dolorous * r9c53ec38642d 10/crawl-ref/source/dat/database/monspeak.txt: Remove obsolete Donald complaint wrt Lugonu's self-banishment. 20:52:19 -!- violetj_ has joined ##crawl-dev 20:52:45 -!- pineappl1 has joined ##crawl-dev 20:52:52 -!- RjY_ has joined ##crawl-dev 20:53:18 -!- jarpiain_ has joined ##crawl-dev 20:53:25 -!- violetj has quit [Ping timeout: 240 seconds] 20:53:25 -!- Xiberia has quit [Ping timeout: 240 seconds] 20:53:25 -!- RjY has quit [Ping timeout: 240 seconds] 20:53:25 -!- pineapple has quit [Ping timeout: 240 seconds] 20:53:26 -!- lorimer has quit [Ping timeout: 240 seconds] 20:53:26 -!- jarpiain has quit [Ping timeout: 240 seconds] 20:53:29 -!- lorimer has joined ##crawl-dev 20:55:36 -!- Xiberia has joined ##crawl-dev 21:12:58 -!- eith has quit [Ping timeout: 252 seconds] 21:13:52 -!- Luca__ has quit [Remote host closed the connection] 21:14:12 -!- Luca__ has joined ##crawl-dev 21:21:20 -!- Enne has quit [Quit: Enne] 21:35:17 -!- Luca__ has quit [Remote host closed the connection] 21:35:34 -!- Luca__ has joined ##crawl-dev 21:50:41 -!- Luca__ has quit [Remote host closed the connection] 21:50:57 -!- Luca__ has joined ##crawl-dev 22:00:10 Shops are confusing if easy_confirm = none (https://crawl.develz.org/mantis/view.php?id=1701) by rlbond86 22:16:09 bah, I had figured this out a while back, but I've forgotten again, whats the editor thats used on CAO? 22:16:17 ??editor 22:16:17 I don't have a page labeled editor in my learndb. 22:24:12 BirdoPrey (L5 VpAs) ASSERT(item.base_type == OBJ_CORPSES && item.sub_type == CORPSE_BODY) in 'misc.cc' at line 155 failed. (D:4) 22:36:38 -!- ogaz has joined ##crawl-dev 22:42:21 -!- xale has joined ##crawl-dev 23:08:37 -!- neunon has quit [Quit: Leaving] 23:09:16 -!- RjY_ has quit [Quit: Changing server] 23:09:18 -!- RjY has joined ##crawl-dev 23:09:19 -!- Zaba has quit [Ping timeout: 260 seconds] 23:15:23 -!- Zaba has joined ##crawl-dev 23:31:21 03dolorous * rd6eca3ee69d8 10/crawl-ref/docs/changelog.txt: Update the 0.7 changelog again. 23:40:50 what happens right now on CAO/CDO if the connection is dropped when a player is looking at something like an acquirement prompt? 23:41:34 I think there's some code to handle level changes, but what about other similar situations?