00:05:41 -!- Fangorn_ has joined ##crawl-dev 00:05:54 -!- Fangorn_ has quit [Client Quit] 00:20:24 -!- valrus has quit [Remote host closed the connection] 01:27:58 I'm really against square LOS. 01:28:44 Testing's fine, I gueess. 01:45:53 -!- casmith789 has joined ##crawl-dev 02:06:30 can troves ask for food currently? 02:06:49 asking the player to clear hive might be fun 02:17:43 -!- Vandal has quit [] 02:19:25 by: no, but it can easily be added 02:26:27 -!- Pseudonut has joined ##crawl-dev 02:43:03 -!- Anym has joined ##crawl-dev 03:00:18 -!- Twinge has joined ##crawl-dev 03:04:15 -!- Twinge_ has quit [Ping timeout: 245 seconds] 03:06:06 -!- syllogism has joined ##crawl-dev 03:33:13 -!- ortoslon has joined ##crawl-dev 03:34:15 -!- JamezQ has left ##crawl-dev 03:52:06 -!- ortoslon has quit [Ping timeout: 265 seconds] 03:57:05 -!- Twinge has quit [Read error: Connection reset by peer] 03:57:26 -!- Twinge has joined ##crawl-dev 04:00:44 -!- Ekaterin has joined ##crawl-dev 04:15:39 -!- jbenedetto has left ##crawl-dev 04:32:17 -!- Pseudonut has quit [Remote host closed the connection] 04:41:41 -!- Mu_ has joined ##crawl-dev 04:56:42 !tell valrus If you're missing things like stdio.h in /usr/include on MacOS, that means you're trying to build with a missing SDK. The Crawl Makefile tries to use 10.4u SDK; you probably don't have it. 04:56:42 greensnark: OK, I'll let valrus know. 04:57:24 !tell valrus You can run "make APPLE_GCC=n" to get the makefile to just use the current SDK without trying to get clever. 04:57:24 greensnark: OK, I'll let valrus know. 04:57:45 We are persecuted by clever makefiles 04:57:55 s/clever/hacky/ 04:58:30 * kilobyte babbles incoherently about the superiority of autoconf. 04:58:50 haha i was just about to jokingly suggest using autotools 04:59:08 fully expecting to be laughed out of the room. 04:59:08 All you need is .vcproj :D 05:01:16 autoconf has one big downside: forcing people to learn m4, and a smaller one: producing multi-MB output trying to be portable to 1970s shells/compilers 05:01:39 but then, it makes it easy to do all autodetection right -- allowing overrides and the like 05:02:03 of course, if you don't bother reading its docs first, it makes it easy to do autodetection wrong, too 05:03:15 if autoconf didn't pollute the crawl code base, I wouldn't mind it 05:03:42 that is, it's entirely optional and doesn't do more than provide one or two included files that could also be generated by hand 05:03:47 "pollute" as in? You typically don't put its output under version control. 05:04:21 as in, no useless INSTALL file; it doesn't touch the makefile 05:04:32 it produces config.h and Makefile -- the first makes sense to edit by hand, but the second, not really 05:04:36 it just generates simple config.h / config.mak 05:04:41 uhm, that's the whole point 05:05:06 why does it need to produce a makefile 05:05:29 kilobyte: you can get by without really learning m4. I didn't :) 05:05:44 it produces makefiles with sane functionality but insane complexity under the hood, which you're not supposed to look at -- and edit Makefile.am/configure.ac instead 05:06:13 by: well, how are you supposed to do conditional builds well? 05:06:20 yes, just imagine that it works like magic, for a fantays roguelike developer probably not a big problem :) 05:06:38 for a good example how NOT to do conditional builds, look at our current makefile 05:07:49 an use case from just 11 minutes ago: why would you have to provide weird settings like APPLE_GCC=n for something that is perfectly detectable? 05:08:07 what do you mean with conditional builds 05:08:19 overriding such stuff should be an exception rather than a rule 05:08:31 and why can't the configure script write APPLE_GCC=n to config.mak 05:09:20 by: for example, looking for system libraries in different place, and compiling contrib only if they are not available 05:09:42 why does the makefile need to be generated for that? 05:10:02 because it would force you to generate that complexity yourself, and it's error-prone and likely to be very incomplete 05:10:37 why can't the configure script just write LIBS=... and INCLUDES=... to some config.mak file? 05:10:47 I think the makefile should be far simpler than it is 05:11:05 -!- Anym is now known as Guest61290 05:11:11 -!- Anym_ has joined ##crawl-dev 05:11:19 any project of more than trivial size and/or portability is bound to have a makefile of aggravating complexity and unmaintainability, why not use a tool to write that stuff for you :) 05:11:38 and Crawl is certainly not of a trivial size 05:12:15 especially if the tool can detect what's needed and how the stuff should look like 05:12:23 -!- Guest61290 has quit [Ping timeout: 260 seconds] 05:12:45 don't let a human do a job that a machine could do 05:12:50 make itself is not the best designed tool in the world, makefiles have a tendency to become incredibly complicated 05:13:21 let's rewrite crawl's makefiles in ant! 05:13:25 kilobyte: have you ever looked at git's makefile? it makes my head hurt 05:13:34 <+by> why can't the configure script just write LIBS=... and INCLUDES=... to some config.mak file? 05:14:07 by: that's the simple part. what about all those special cases in crawl's makefiles? those that tend to break? 05:14:31 -!- Anym_ has quit [Client Quit] 05:14:37 I asked for an example that requires the makefiles to be script-generated 05:14:50 by: well, it can. but if you let automake generate the makefile you get a load of really useful build features for free, like DESTDIR which is super helpful for building debian packages 05:15:16 RjY: quite monstrous, but actually more readable than Crawl's. It's similar and suffers from the same disease. 05:15:40 RjY: that kind of unnecessary complexity is a good argument against autoconf 05:16:09 by: and crawl's makefiles aren't unnecessary complex as they are? 05:16:22 by: how do you propose doing it better than with DESTDIR? You need to somehow redirect the installation dirs. 05:18:07 by: that way with LIBS=... and INCLUDES=... would work only if you allow OBJECTS= to be written at all 05:18:15 kilobyte: I don't care about doing that, but clearly a simple makefile can have a DESTDIR variable 05:18:26 or not... it would throw dependency tracking away 05:18:45 also, crawl's build isn't done with a simple makefile 05:18:47 and dependency tracking lets us reduce the number of full rebuilds 05:19:11 I'm not saying that crawl's makefile is simple or good 05:19:35 without knowing which object file depends on what headers, you have no choice but rebuild everything every single time 05:19:51 I don't know if autoconf would be the best available option. but I have not enough experience with other configurations options to recommend any of them 05:21:43 -!- Moriasc has joined ##crawl-dev 05:22:48 -!- greensnark has quit [Ping timeout: 264 seconds] 05:23:13 -!- greensnark has joined ##crawl-dev 05:32:29 I have the new save file containers mostly finished. 05:32:43 -!- stabwound has quit [Ping timeout: 260 seconds] 05:32:48 Questions: 1. do we want an option to produce non-compressed saves? 05:33:28 because of fragmentation, I guess that not really 05:34:03 -!- stabwound has joined ##crawl-dev 05:34:24 2. do we want to allow sub-file names of an arbitrary length? We currently use max 3, my current implementation allows 4, but this can be easily changed. 05:36:32 can 'any good_item' produce gold acquirement? 05:37:31 Mu_: it's never acquirement, but it can produce a pile of gold appropriate for a very deep depth 05:37:43 ah ok 05:37:51 i pulled 6k gold out of a bailey 05:38:00 seems a little much for a character on D:9 05:39:33 good_item ignores your depth, it produces loot as if you were on D:350 05:39:51 hah ok 05:42:10 no, actually gold is special cased 05:42:54 and acquirement indeed calls normal item creation, giving the same as "gold good_item" 05:42:55 considering poison arrow already has the irresistable damage, shouldn't venom bolt be entirely resistable? 05:43:12 I'm surprised how much damage it does to resistant stuff 05:43:56 by: why? You are struck by a poisoned bolt, even if you ignore poison, it's still a physical bolt. 05:44:38 ??bolt of fire 05:44:38 bolt of fire[1/1]: A level 6 conj/fire penetrating beam attack that can hit multiple opponents in a line for pure fire damage. Melts wax. Does six dice with a maximum of (18 + (power * 2) / 3). At power 50 it does 6d8 damage; at power 100 it does 6d14. As with any fire-based beams, produces steam if it passes over water. 05:45:00 that's "poison arrow" for me 05:45:50 even sting does a bit of physical damage 05:47:31 Mu_: I think that special casing should be thrown out, separating acquirement away 05:48:17 the thing is poison resistance counts as rP+ out of rP+++ for damage reduction purposes 05:48:52 should allow monsters multiple levels of poison resistance 05:50:09 the resistable parts of poison spells still do 1/2 damage against poison resistant monsters 05:50:11 that would make sense for solid-metal golems, a sting is not going to do much 05:53:51 -!- MarvinPA has joined ##crawl-dev 06:01:56 -!- eith has joined ##crawl-dev 06:02:38 -!- tenach has joined ##crawl-dev 06:51:53 -!- ortoslon has joined ##crawl-dev 06:52:01 -!- Vandal has joined ##crawl-dev 07:09:04 -!- elliptic has joined ##crawl-dev 07:36:44 -!- Anym has joined ##crawl-dev 07:53:56 -!- flowsnake has joined ##crawl-dev 08:37:26 Just noticed that there are about 90 lines of e-mail headers in docs/develop/git/quickstart.txt. Should I file a bug? 08:39:39 if you do, also mention the mime-encoded part 08:44:14 -!- ortoslon has quit [Disconnected by services] 08:44:14 -!- ortoslon1 has joined ##crawl-dev 08:44:18 -!- ortoslon1 is now known as ortoslon 08:56:42 Anybody here regularly building Crawl under Windows? 08:58:57 Spells which cannot miss their target still prompt when fired in an ally's direction (https://crawl.develz.org/mantis/view.php?id=2265) by MarvintheParanoidAndroid 08:59:36 Just ran into a compiler error: 08:59:40 chardump.cc: In function 'void _sdump_spells(dump_params&)': 08:59:42 chardump.cc:941: internal compiler error: Segmentation fault 09:05:54 -!- ortoslon has quit [Quit: bye] 09:14:02 Hard to play the game on a bright monitor (https://crawl.develz.org/mantis/view.php?id=2266) by QQCucumber 09:14:30 Anym: that's not our fault, "internal compiler error" means a bug in gcc 09:16:23 I figured as much. I was just wondering if there was a quick workaround or something. 09:18:33 -!- QQCucumber has joined ##crawl-dev 09:18:52 Am I able to talk about a ticket I submitted? 09:19:49 sure, the one about black background? 09:20:18 I'm afraid that bug lies solely in your monitor, it's a valid colour that is the traditional background for text 09:20:29 and a vast majority of terminals use it 09:20:48 It's not really a bug. Do you really want the game to be hard to play for someone who purchases a bright monitor? 09:21:16 I bet you can tune that monitor by adjusting its settings 09:21:24 i'm not sure i understand how a black background ends up brighter than a white background 09:21:32 white background is horrible 09:21:42 if you want it to be usable as a light colour, it should be a light yellow 09:21:44 -!- June3 has quit [Ping timeout: 252 seconds] 09:22:03 clear white is very straining for your eyes on computer screens 09:23:00 I understand I'm the minority in here because I don't really use terminals that often. 09:23:16 if you're playing the terminal build, you can often do tricks 09:23:25 by redifining the terminal colours, having black as whatever you want 09:23:38 I already figured out how to change it in the terminal build, yes. It took me a lot of time to figure out how to do though. 09:23:43 in fact, you can redefine colours to use (*shudder*) black-on-white instead of white-on-black 09:23:51 But I can't play the Tiles version, which I wanted to do because it brought me back to when I used to play Castle of the Winds 09:23:57 in any case, you should talk to enne 09:24:06 none of us here at the moment even use tiles(: 09:24:19 I debug on it since it's easier to hook to! 09:24:26 -!- ortoslon has joined ##crawl-dev 09:24:30 but then, quite a lot of games use blackish backgrounds 09:24:59 or for Doom3, blackish foregrounds as well :p 09:24:59 Using a bright anti-glare monitor for me is a necessity because I have a very large window in the room the computer is in and I enjoy getting sunlight in here. 09:25:29 I don't think anyone who played Doom 3 could see the game half the time 09:25:59 no, it just sucks on LCD, it was made for CRT 09:26:06 which do have decent blacks 09:26:18 Well, that's kind of my point. Most LCDs have poor black levels 09:26:20 instead of grays like LCD 09:26:38 and pure black on an LCD looks absolutely awful because it makes the colors bleed into it 09:26:44 but what you wrote in the bug report was that you _want_ gray background 09:27:04 A dark gray charcoal like background is a significant step up 09:27:11 uhm, if colors bleed, then that's a problem other than merely being too bright 09:27:27 I think you're getting the wrong idea 09:27:42 it doesn't /actually/ bleed, but it appears to. 09:29:25 I pointed out in the feature request that an option for a lighter color scheme would be appreciated. I don't think I would be the only one who'd use it - most websites nowadays use light colors after all. 09:29:40 Wouldn't have to be a white background. 09:29:43 I do have an opposite problem: in my office, between ~12-14:00, dark blue and to a lesser degree, other dark colours, all mimic black -- but they do turn black, instead of black turning an other colour 09:30:37 Well, that's another good reason for a lighter color scheme option. 09:31:13 -!- Cryp71c has joined ##crawl-dev 09:31:18 the background is not a problem, foreground is 09:31:33 I want bigger contrast, you want smaller 09:31:55 A lighter color scheme would fix that issue all the same though. 09:32:57 Well, since we're on the topic, IS there a way to change the background color of the tiles version to a charcoal (50 r/g/b)? I have the sneaking suspicion the source would have to be edited for that, most I've been able to do is change the background color of the game world. 09:33:28 uhm, why? lighter blue on gray would be little different in contrast to dark blue on black 09:34:07 Okay, sure. That might work too. Charcoal fixed the issue for me though, so that's what I've been using. 09:51:26 -!- MarvinPA has quit [Ping timeout: 252 seconds] 10:02:32 in my local build ever monster i examine as the 'radiating silence' descriptor when i use x to look at it 10:06:01 -!- ortoslon has quit [Ping timeout: 246 seconds] 10:06:21 -!- ortoslon1 has joined ##crawl-dev 10:06:26 -!- ortoslon1 is now known as ortoslon 10:22:53 Mu_: Which revision? My 0.8.0-a0-599-g4e21c95 is not doing that 10:23:50 0.8.0-a0-604-g93e49fb 10:26:44 Ok, I'll compile the latest and try. 10:27:13 -!- MarvinPA has joined ##crawl-dev 10:27:56 -!- Textmode has joined ##crawl-dev 10:30:29 Nope, not seeing that (same revision, tiles) 10:30:52 weird 10:31:15 Here: A giant newt (radiating silence) 10:38:24 -!- mruuuu has quit [Ping timeout: 264 seconds] 10:38:56 -!- QQCucumber has left ##crawl-dev 10:44:33 "colorless scales" is not proper Australian English spelling (https://crawl.develz.org/mantis/view.php?id=2267) by reid 10:47:24 So its official then, Australian English is the norm? If so, I'll go ahead and fix ^^ 10:47:59 It's Linley's game! 10:48:45 everybody who wants american spelling can fork and call it dungeon crawl stone sop! 10:50:28 And add aluminum golems to it 10:51:26 Aluminium* 10:51:39 03Cryp71c * re045741e0933 10/crawl-ref/source/mutation-data.h: Mantis 2267 : Colorless => Colourless 10:51:40 Hm, looks like it's the -pipe flag that causes msysGit's GCC to segfault. 10:51:42 Cryp71c: You fail your test for American english 10:51:57 greensnark, lol :) 10:52:03 Cryp71c: Americans are not supposed to correct that spelling :P 10:52:35 I actually tend to write "colour" and "colorless", because "color" is so colourless 10:52:36 Maybe there should be a mutation that Americanized the spelling... 10:53:09 Anym: That would cause players to quit! 10:56:33 23 points to #4 clan spot! 10:58:37 unfortunately I threw away a nice HEFE because I confused Roxanne with a giant beetle 10:58:48 Ouchie. 10:58:56 @whereis Keskitalo 10:58:56 Keskitalo the Nimble (L10 KeAs), a worshipper of Kikubaaqudgha, is currently on Lair:1 after 18128 turns. 11:00:38 -!- Ekaterin has quit [Quit: Leaving] 11:31:55 -!- valrus has joined ##crawl-dev 11:35:30 hey greensnark thanks for the help 11:35:43 Ah, did you get it to build? :) 11:36:59 I ended up removing the -isysroot from the gcc lines instead, and it worked :) 11:37:18 (this was before I saw your messages) 11:37:20 Ah, that's what APPLE_GCC=n does 11:37:25 oh, I see 11:37:27 cool 11:37:32 Yeah, I don't know why the makefile wants to use 10.4 SDK at all 11:37:57 on first generation macbooks? 11:38:11 by: The makefile tries to use 10.4 SDK all the time 11:38:17 Without checking if it exists :) 11:38:24 was it just in there when 10.4 was current and never removed? 11:38:26 Whereas it could just use the current SDK and build 11:38:35 yes, I thought the check is for machine architecture 11:38:52 valrus: Things Just Work if you don't try to force the SDK 11:39:12 why not change it, then? 11:39:28 I don't know the intent of whoever wrote that :) 11:39:37 fair enough 11:39:57 maybe it was to make it so the uninitiated can't compile :P 11:40:17 I think it might happen only on 32bit intel macs 11:40:41 what might? 11:40:46 I'm on a 64bit intel 11:40:50 oh 11:40:59 by: Yeah, same here 11:41:11 The makefile forces 10.4u SDK and i386 compile 11:41:20 there's a check for "uname -m" vs x86 11:41:26 That also causes mayhem in rltiles if you have SDL installed from MacPorts 11:42:30 i.e. the main build forces a 32-bit build and then uses pkg_config and tries to link against the 64-bit MacPorts libs 11:43:07 But I haven't really spent much time on rltiles for obvious reasons :) 11:44:51 Treating ARCH as uname -m is wrong in the makefile 11:45:05 Because a lot of 64-bit Macs still use the 32-bit kernel 11:45:13 Especially laptops 11:46:10 cdo seems to be lagging heavily all of a sudden 11:46:55 I don't see any significant activity 11:47:04 Although the tremulous server seems reasonably popular :P 11:47:04 fine again, maybe the connection? 11:47:22 Sounds like a connection issue 11:48:11 -!- ortoslon has quit [Disconnected by services] 11:48:11 -!- ortoslon1 has joined ##crawl-dev 12:07:59 -!- ortoslon has joined ##crawl-dev 12:08:54 -!- Moriasc has quit [Remote host closed the connection] 12:09:23 -!- ortoslon1 has quit [Ping timeout: 260 seconds] 12:12:34 -!- Textmode has quit [Ping timeout: 265 seconds] 12:18:45 -!- tenach has quit [Read error: No route to host] 12:21:52 -!- Textmode has joined ##crawl-dev 12:34:02 -!- Textmode has quit [Ping timeout: 240 seconds] 12:54:01 -!- Siber has quit [Ping timeout: 276 seconds] 12:56:49 -!- Textmode has joined ##crawl-dev 13:06:29 !tell due volcano_bunker seems to have MONS: dragon, but no glyph that places it.. 13:06:30 Keskitalo: OK, I'll let due know. 13:10:09 -!- Anym has quit [Disconnected by services] 13:10:15 -!- Anym has joined ##crawl-dev 13:12:35 -!- Siber has joined ##crawl-dev 13:22:37 -!- ortoslon has quit [Ping timeout: 276 seconds] 13:26:39 -!- Pseudonut has joined ##crawl-dev 13:36:02 are slime creatures not supposed to split up if they have the space? 13:40:39 they're supposed to, but they don't always do it 13:40:50 they usually will eventually 13:43:39 Old BR for monsters resisting hellfire when they only have rF and whatnot: https://crawl.develz.org/mantis/view.php?id=873 13:44:29 Was looking into it and wanted you guys' feedback on whether monsters (who have rF but not rHellfire) should only resist hellfires not cast by the player, or if their resists should be the same regardless of who cast the hellfire. 13:44:33 s/you/your 13:47:00 same regardless 13:47:57 that's what I felt, but obviously that has implications for monsters that get caught near, say...hellfire island, or cerebov, and don't have rHellfire, so I wanted to make sure my perspective wasn't entirely unique. 13:48:26 should keep an eye out for that, but yeah, same regardless 13:48:41 we could also make monster hellfire smart and not damage anyone it's not supposed to hurt :) 13:48:45 kind of un-demonic, but.. 13:49:16 Well, at least for cerebov, everything in his vault is rHellfire, I believe, as are most of the greater demons he can summon. 13:49:19 @??Executioner 13:49:20 Executioner (151) | Speed: 20 | HD: 12 | Health: 49-83 | AC/EV: 10/15 | Damage: 30, 10, 10 | Flags: 05demonic, evil, see invisible, !sil | Res: 06magic(144), 05fire, 02cold, 10elec++, 03poison | XP: 2373 | Sp: pain (d14), haste. 13:49:21 @??Fiend 13:49:22 Fiend (041) | Speed: 10 | HD: 18 | Health: 72-120 | AC/EV: 15/6 | Damage: 25, 15, 15 | Flags: 05demonic, see invisible, fly, !sil | Res: 06magic(288), 05hellfire, 03poison | Vul: 12cold | XP: 3702 | Sp: hellfire blast (3d20), torment symbol. 13:49:24 @??Pit Fiend 13:49:24 Pit Fiend (071) | Speed: 8 | HD: 19 | Health: 93-146 | AC/EV: 17/5 | Damage: 28, 21, 21 | Flags: 05demonic, see invisible, fly, !sil | Res: 06magic(304), 05hellfire, 02cold, 10elec++, 03poison | XP: 1869 | Sp: hellfire blast (3d20), torment symbol. 13:49:48 -!- flowsnake has quit [Quit: Leaving] 13:51:52 @??Killer Klown 13:51:52 Killer Klown (04@) | Speed: 15 | HD: 20 | Health: 128-174 | AC/EV: 10/15 | Damage: 3012(klown) | Flags: see invisible, regen, !sil | Res: 06magic(160) | XP: 6948 | Sp: blink. 13:52:21 Huge xp! 13:53:31 -!- flowsnake has joined ##crawl-dev 13:53:42 Alrighty, I'm keeping AC reduction in, since even Cerebov's hellfire checks AC 13:54:36 -!- Anym has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]] 13:57:16 cerebov's hellfire? 13:57:35 @??cerebov 13:57:35 Cerebov (05&) | Speed: 10 | HD: 21 | Health: 650 | AC/EV: 30/8 | Damage: 60 | Flags: 05demonic, see invisible, !sil | Res: 06magic(168), 04hellfire+++, 03poison | XP: 15000 | Sp: fire storm (8d16), iron shot (3d36), haste, greater demon. 13:57:37 he doesn't have hellfire does he? 14:02:29 -!- Pseudonut has quit [Remote host closed the connection] 14:10:43 he does not have hellfire 14:11:16 and if he did, it wouldn't check AC 14:31:49 yeah, meant fire storm, but fire storm checks AC 14:31:57 That's intentional, yeah? 14:32:16 ??fire storm 14:32:22 fire storm[1/1]: The most powerful fire conjuration, dealing 35% resistible fire damage in a large radius and leaving behind fire clouds and {fire vortices}. Like its frigid counterpart, it makes a godawful racket when cast. Unlike ice storm, can be placed anywhere in line of sight and ignores obstacles. Don't be tricked -- it CAN reach you 3 squares away! 14:32:34 yeah 14:35:01 yes, fire storm is not hellfire 14:35:05 completely different things 14:38:33 -!- Pseudonut has joined ##crawl-dev 14:39:38 -!- Zaba has quit [Ping timeout: 240 seconds] 14:46:07 -!- Zaba has joined ##crawl-dev 14:49:59 03Cryp71c * r7db9b672f21f 10/crawl-ref/source/ (beam.cc monster.cc monster.h): Mantis 873 : Monster rF counted as rHellfire 14:51:38 wait, what? 14:51:50 doy, what what? 14:52:15 that commit message implies that now monsters use rF to resist hellfire damage? 14:52:17 Sorry, that commit title was a bit misleading, it made sense when I typed it up, rF was counting as rHellfire, that commit split them up. 14:52:27 ah, okay 14:52:37 forgot "was" 14:52:48 at least the commit text is clarifying 15:08:31 -!- Pseudonut has quit [Quit: Computer went to sleep] 15:15:59 Cryp71c: hellfire resist is just 0 or 1, so that "if (resist > 2)" is always false 15:16:12 in particular, it looks like say imps only partially resist hellfire now 15:18:08 fire+++ used to also be hellfire resistant? 15:20:54 monsters with hellfire resistance got fire+++ automatically 15:22:30 right, but the other way around used to be true too 15:22:47 yes, in particular fire+ was one level of hellfire resistance 15:23:18 by, right, but that is not appropriately true, since we have both rFire and rHellfire 15:23:45 Cryp71c: I don't understand that sentence 15:23:46 by, though I'm confused on if hellfire resist is binary, why there are variable levels of rHellfire for various monsters. 15:24:00 by, rF should not convey rHellfire, to any degree. 15:24:01 what monsters have different levels of hellfire? 15:24:05 @??Cerebov 15:24:06 Cerebov (05&) | Speed: 10 | HD: 21 | Health: 650 | AC/EV: 30/8 | Damage: 60 | Flags: 05demonic, see invisible, !sil | Res: 06magic(168), 04hellfire+++, 03poison | XP: 15000 | Sp: fire storm (8d16), iron shot (3d36), haste, greater demon. 15:24:07 @??Fiend 15:24:08 Fiend (041) | Speed: 10 | HD: 18 | Health: 72-120 | AC/EV: 15/6 | Damage: 25, 15, 15 | Flags: 05demonic, see invisible, fly, !sil | Res: 06magic(288), 05hellfire, 03poison | Vul: 12cold | XP: 3702 | Sp: hellfire blast (3d20), torment symbol. 15:24:16 @??Hellion 15:24:16 hellion (052) | Speed: 13 | HD: 7 | Health: 25-53 | AC/EV: 5/10 | Damage: 10 | Flags: 05demonic, !sil | Res: 06magic(65), 05hellfire, 03poison | Vul: 12cold | XP: 597 | Sp: hellfire burst (3d15). 15:24:24 Cerebov is hellfire immune (presumably?) 15:24:31 whereas fiends and hellions only take partial damage from it. 15:24:38 Or at least, that's the convention I understood. 15:25:54 -!- valrus has quit [Remote host closed the connection] 15:26:10 it's just cerebov that has this weird triple hellfire resistance, but as far as I can tell, that didn't use to make a difference 15:26:33 fiends and hellions are immune to hellfire also I think 15:26:42 yes, hellfire res is just binary 15:26:45 as far as i know 15:26:53 Its implementation isn't, actually 15:26:57 The only reason they were immune before 15:27:01 if it's not in the code, it isn't actually checked 15:27:11 but looking at the code, it didn't used to be checked 15:27:11 so 15:27:34 was because hellfire conveyed rF+++ and monster's were checking res.fire instead of res.hellfire when it came to hellfire damage. 15:27:44 right 15:27:44 I don't think we want fiends partially vulnerable 15:27:50 so hellfire+++ doesn't make sense 15:27:54 we don't 15:28:00 doy, right, well..that's a separate discrepency, I think. 15:28:09 Cryp71c: but one that doesn't actually matter 15:28:17 yeah, hellfire resistance *should* be binary I feel 15:28:24 just treat hellfire+++ the same as hellfire+, since that's what used to happen 15:28:27 k, that's my misinterpretation, then. I'll make a quick change. 15:29:08 and probably give cerebov hellfire+, just for less confusion in the future 15:32:34 03Cryp71c * r42465c0cb4c1 10/crawl-ref/source/ (beam.cc mon-data.h): Change rHellfire into binary resist 15:32:58 Sound alright with you guys that confused giant eyeballs should sometimes not paralyze anything (even if they would've otherwise), and other times paralyze other monsters, instead of the player, out of their confusion? 15:33:14 I FR'd it a while ago, but its gone mostly unnoticed, and I've got an implementation for it done 15:33:56 i think it should follow the behavior for zapping wands when confused 15:34:08 just pick a random tile in los and shoot a tracer at that 15:35:29 Well, paralyze is smite-targetted, and nothing happens if nothing is on the tile, which is kind of what I'm doing anyways: 15:36:26 If the eyeball is confused && coinflip, iterate through all monsters in los and if (one_chance_in(++chance)), paralyze them...so it becomes increasingly unlikely for any target to be paralyzed 15:36:41 This would allow for mult-targetted paralyze, but could just as easily only paralyze one target. 15:36:42 yeah, i don't particularly like that implementation is my point 15:36:48 I gotcha 15:55:09 Yeah, that implementation is a waste of time, since there are no graphical effects for paralyze, I would be better off just coding paralyze to fail when the eyeball is confused, I sat there for a hundred turns with 4 giant eyeballs confused (in a fairly narrow space) and I couldn't get any of the eyeballs to paralyze any of the other eyeballs because of the rarity of eyeballs taking action + the chance of rng picking a ti 15:55:09 le with something actually on it, etc. 15:57:04 eyeballs can't paralyze other eyeballs, can they? 15:57:33 doy, right now, they only paralyze the player, but when confused, they should have a chance to paralyze something else in their LOS. 15:57:41 @??giant eyeball 15:57:41 giant eyeball (16G) | Speed: 3 | HD: 3 | Health: 9-24 | AC/EV: 0/1 | Flags: lev | Res: 06magic(12), asphyx | XP: 3. 15:58:11 I don't see why they shouldn't be able to paralyze other eyeballs (and they can, they were doing it under my first implementation) 16:12:46 03Cryp71c * rb50dee43503e 10/crawl-ref/source/mon-abil.cc: Mantis 2194 Confused giant eyeballs 16:21:49 the "if (one_chance_in(++count))" idiom chooses a random element of a list fairly; if you start at 3, I don't believe it's fair 16:22:30 also, your probability math is off 16:23:58 actually, it may be fair after all 16:24:17 by, yeah, I forgot I changed base value from 2 to 3, making the starting chance 1/4 not 1/3 16:24:17 also, why can't the player be chosen? 16:24:40 by, they can, if the chance fails, the player may still be paralyzed, even by a confused eyeball. 16:25:23 i still don't think just making up probabilities here makes very much sense 16:26:00 doy, I ran through it, testing as appropriate and came up with the probabilities that I used, they feel quite fair, but rare enough as to not be problematic on balance, I think. 16:26:37 It seemed better than making confused eyeballs do nothing while confused, which was my initial plan of implementation. 16:29:25 just arbitrarily choosing some random monster in view at some fixed chance seems like it doesn't really make any sense though 16:30:55 really? I thought it made reasonable sense, with a confused monster stumbling around, but still trying to kill the player, they would use their abilities / weapons still. In the giant eyeball's case, when thinking of this entire scenario thematically 16:30:58 doy: why? It's a simple, fair version. 16:31:21 Imagined it as the eyeball thinknig "There! That guy is the guy I'm killing *attempt to paralyze*" 16:31:34 whether or not its the player is a matter of rng 16:31:47 it may make a bit more sense to bias towards selecting nearby targets, but just selecting a random one isn't bad either 16:32:00 why is it any different from a monster zapping a wand? 16:32:24 hmm, it might be good to not reward having many vs no alternate targets 16:32:33 I used that code, actually, and it selects a target so rarely, you might as well save the codespace and cancel eyeball-paralyze if its confused. 16:32:36 after a thought, you do have a point 16:33:32 my main point here is that we should really be consistent in how we treat these things, not just implement something different for every case 16:33:55 perhaps confuse a random target along a random ray, similar to wands? That still gives a decent chance of paralyzing an adjacent player. 16:34:18 -!- elliptic has quit [*.net *.split] 16:34:20 -!- jld has quit [*.net *.split] 16:34:47 -!- jld has joined ##crawl-dev 16:35:05 -!- jld has quit [*.net *.split] 16:35:28 -!- jld has joined ##crawl-dev 16:36:30 -!- Twinge has quit [*.net *.split] 16:36:31 -!- CIA-42 has quit [*.net *.split] 16:36:31 -!- MarvinPA has quit [*.net *.split] 16:36:35 -!- phyphor has quit [*.net *.split] 16:36:37 -!- lorimer has quit [*.net *.split] 16:36:38 -!- Zao has quit [*.net *.split] 16:36:38 -!- felirx has quit [*.net *.split] 16:36:40 -!- flowsnake has quit [*.net *.split] 16:36:41 -!- Cryp71c has quit [*.net *.split] 16:36:42 -!- by has quit [*.net *.split] 16:36:42 -!- rax has quit [*.net *.split] 16:36:44 -!- Adeon has quit [*.net *.split] 16:36:46 -!- Vandal has quit [*.net *.split] 16:36:49 -!- RjY has quit [*.net *.split] 16:36:50 -!- casmith789 has quit [*.net *.split] 16:36:51 -!- bhaak has quit [*.net *.split] 16:36:52 -!- paxed has quit [*.net *.split] 16:36:57 -!- joosa has quit [*.net *.split] 16:36:58 -!- due has quit [*.net *.split] 16:38:37 -!- Keskitalo has quit [Ping timeout: 267 seconds] 16:39:38 -!- flowsnake has joined ##crawl-dev 16:39:38 -!- MarvinPA has joined ##crawl-dev 16:39:38 -!- Cryp71c has joined ##crawl-dev 16:39:38 -!- Vandal has joined ##crawl-dev 16:39:38 -!- Twinge has joined ##crawl-dev 16:39:38 -!- casmith789 has joined ##crawl-dev 16:39:38 -!- by has joined ##crawl-dev 16:39:38 -!- CIA-42 has joined ##crawl-dev 16:39:38 -!- rax has joined ##crawl-dev 16:39:38 -!- phyphor has joined ##crawl-dev 16:39:38 -!- bhaak has joined ##crawl-dev 16:39:38 -!- lorimer has joined ##crawl-dev 16:39:38 -!- paxed has joined ##crawl-dev 16:39:38 -!- Zao has joined ##crawl-dev 16:39:38 -!- Adeon has joined ##crawl-dev 16:39:38 -!- felirx has joined ##crawl-dev 16:39:38 -!- joosa has joined ##crawl-dev 16:39:38 -!- due has joined ##crawl-dev 16:39:38 -!- RjY has joined ##crawl-dev 16:39:58 -!- Keskital1 has joined ##crawl-dev 16:40:50 -!- purge has quit [*.net *.split] 16:40:50 -!- Eronarn has quit [*.net *.split] 16:40:51 -!- kilobyte has quit [*.net *.split] 16:40:53 -!- philsnow has quit [*.net *.split] 16:41:08 -!- Eronarn has joined ##crawl-dev 16:41:24 -!- Cryp71c_ has joined ##crawl-dev 16:41:26 -!- Cryp71c has quit [Write error: Broken pipe] 16:41:37 Even with a room full of eyeballs (ever square possible square occupied via &M monster creation) and all confused (mass confusion), it still took ~10 turns for even 1 eyeball to paralyze another eyeball. 16:41:41 -!- Cryp71c_ is now known as Cryp71c 16:42:06 -!- purge has joined ##crawl-dev 16:42:39 -!- philsnow_ has joined ##crawl-dev 16:42:46 -!- kilobyte_ has joined ##crawl-dev 16:47:04 -!- kilobyte_ is now known as kilobyte 16:47:44 -!- elliptic has joined ##crawl-dev 16:48:10 anyways.. 16:48:13 Even with a room full of eyeballs (ever square possible square occupied via &M monster creation) and all confused (mass confusion), it still took ~10 turns for even 1 eyeball to paralyze another eyeball. 16:48:23 Sorry for the repeats, for anyone who has seen them 16:48:26 Silly splits. 16:49:06 well they are speed 3 16:49:28 and they don't even attept to paralyze things every turn they take 16:50:39 03j-p-e-g * r33f764fe6911 10/crawl-ref/source/rltiles/ (3 files in 2 dirs): Add Poor Yurik's new minotaur doll tile, minus horns. 16:50:41 03j-p-e-g * rcc48ed47a29c 10/crawl-ref/source/rltiles/ (16 files in 3 dirs): Replace Xom's altar tiles with coolio's new design. 16:50:42 03j-p-e-g * r79f2b8674223 10/crawl-ref/source/tilemcache.cc: Fix #2232: Don't draw ogre mages with a weapon tile. 16:50:42 doy, correct, which is part of why the same code as the confused wand usage doesn't work (the othe reason is because paralyze is smite targeting) 16:51:05 It just results in super, rediculously rare effects 16:51:42 i don't think making them artificially more common is a reasonable solution either 16:51:47 In similar matters, how does monster_iterator loop through things in LOS? Randomly? 16:53:07 -!- Pseudonut has joined ##crawl-dev 16:57:21 Iterating through the monsters in sight is closer to the foe_selection than a random-ray or random-tile implementation. 16:57:31 s/foe_selection/foe_selection of non-confused eyeball behavior 17:00:12 I think it makes more sense that confused eyeballs can't paralyze anyone 17:00:21 they need to be unconfused to focus their stare 17:01:06 That's simlarly fine with me, I thought of that for my initial implementation, but it seemed rather boring, and unlike other confused behavior. 17:01:43 confused casters can't cast spells 17:01:46 I'll implement that and we can discuss this sometime tomorrow, or later on...if there's a big desire for eyeballs to paralyze other things. 17:01:47 how is this different? 17:01:52 Its an ability 17:01:55 not a spell 17:02:09 confused dragons don't breathe fire afaik 17:02:58 03j-p-e-g 07stone_soup-0.7 * r7d3a96fddfcf 10/crawl-ref/docs/changelog.txt: Belatedly add 0.7 highlights to the changelog. (Mantis #2120) 17:03:03 nope, they don't 17:03:05 confused phantoms blink 17:03:28 i think everything can blink while confused, can't it? 17:03:51 marvinpa: only things with af_blink or whatever it is called 17:03:57 ah ok 17:04:01 blink frogs and such 17:04:05 yes 17:04:11 03j-p-e-g * r65b4968f2ba5 10/crawl-ref/source/dat/database/FAQ.txt: Update reference to the tutorial in the FAQ. 17:04:21 I'm not sure whether confused imps can blink 17:04:45 I don't know, I always just EH them 17:04:47 right 17:05:08 K, there! 17:05:16 see ya later. 17:05:39 03Cryp71c * r6854232195c8 10/crawl-ref/source/mon-abil.cc: Confused Giant Eyeball revert & reimplementation 17:05:53 @?giant eyeball 17:05:53 giant eyeball (16G) | Speed: 3 | HD: 3 | Health: 9-24 | AC/EV: 0/1 | Flags: lev | Res: 06magic(12), asphyx | XP: 3. 17:06:22 I wonder whether it might be a good strategy in theory to enslave giant eyeballs 17:06:52 and haste them 17:07:09 and paralyze orbs of fire with them 17:09:01 Summon Eyeballs as a player spell! 17:09:36 Summon dissolution for Jiyvites! 17:10:13 Summon Orbs of Fire and Nequexecs for Zinnites! 17:10:46 -!- Cryp71c has quit [Quit: Leaving] 17:10:59 go to pan, find a random lord with summon eyeballs, lead the lord to cerebov's level, enslave some eyeballs and watch them paralyze everyone 17:11:08 sounds like a good plan 17:33:45 -!- eith has quit [Ping timeout: 240 seconds] 18:03:07 -!- by has quit [Ping timeout: 265 seconds] 18:11:25 -!- by has joined ##crawl-dev 18:13:55 -!- pointless_ has joined ##crawl-dev 19:00:26 -!- Pseudonut has quit [Remote host closed the connection] 19:11:50 -!- syllogism has quit [] 19:14:58 -!- nrook has joined ##crawl-dev 19:49:01 -!- nrook has left ##crawl-dev 19:53:50 -!- Eronarn has quit [Quit: Reconnecting] 19:54:08 -!- Eronarn has joined ##crawl-dev 20:07:05 -!- Mu_ has quit [Quit: Defecator, may everything turn out okay so that you can leave this place.] 20:26:11 -!- mruuuu has joined ##crawl-dev 21:05:57 hey, I won a game a couple hours ago, but it's not showing up on the tournament scoring 21:06:00 !lg . t 21:06:01 115. reid the Backstabber (L19 SpEn), worshipper of Cheibriados, escaped with the Orb and 3 runes on 2010-08-12, with 1660248 points after 50229 turns and 4:43:52. 21:08:05 several more recent wins have been scored 21:08:39 they forgot to announce that chei wins are much slower to get scored :P 21:08:47 you should !tell greensnark 21:09:08 !tell greensnark Hi, I won a game a couple hours ago, but it's not showing up on the tournament scoring: !lg reid t 21:09:08 rwbarton: OK, I'll let greensnark know. 21:13:16 maybe the code that was supposed to determine streaks got confused 21:13:39 -!- nrook has joined ##crawl-dev 21:13:53 -!- nrook has left ##crawl-dev 21:16:34 -!- ortoslon has joined ##crawl-dev 21:28:12 Possible tile for blessed toe (https://crawl.develz.org/mantis/view.php?id=2268) by minmay 21:42:06 -!- elliptic has quit [Ping timeout: 240 seconds] 21:42:49 -!- elliptic has joined ##crawl-dev 21:50:50 -!- MarvinPA has quit [Ping timeout: 252 seconds] 21:53:36 -!- pointless_ has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]] 22:09:49 -!- Textmode has quit [Quit: Over the ages, man has been certain of a great many things. some of these things were even true.] 22:13:19 ooh tiles 22:13:19 due: You have 1 message. Use !messages to read it. 22:13:23 !messages 22:13:23 (1/1) Keskitalo said (9h 6m 53s ago): volcano_bunker seems to have MONS: dragon, but no glyph that places it.. 22:13:40 !tell Keskitalo Yes, I never ended up using the dragon. 22:13:40 due: OK, I'll let Keskitalo know. 22:28:31 please update windows builds 23:05:21 -!- Pseudonut has joined ##crawl-dev 23:11:09 -!- valrus_ has joined ##crawl-dev