00:02:43 evilmike: I personally hat eleeing myself. 00:02:50 It has always feltgrindy and annoying 00:02:56 Unstable branch on crawl.develz.org updated to: 0.10-a0-1160-g5c90495 (32) 00:03:00 however monsters fleeing is sometimes a useful escape mechanism 00:03:36 almost no one seems to like fleeing. getting rid of it would make the game a bit harder though, yeah 00:04:20 that's why I put the suggestion for breaking fleeing if a monster takes damage 00:04:36 The only problem with fleeing is not so much when to trigger it 00:04:38 But the actual effect. 00:05:10 Why not completely re-implement fleeing as "retreating"? Fleeing a short distance, then, if you'er running after it, stopping to fight you, repeat -- or if you don't chase, completely fleeing. 00:05:15 That doesn't... really help. 00:05:55 To be honest, I see absolutely no solution to fleeing but removing it. 00:06:22 one idea I had once was to make it so if a monster is "fleeing" it still attacks you if you're in range. The only time it runs away is if it can't attack you 00:06:41 thus if you melee a yak and stand next to it when it enters "fleeing", combat proceeds as usual. If you step back, it runs away 00:06:55 this seems complicated to me though, which is why I didn't put it on the wiki 00:07:09 Unstable branch on crawl.akrasiac.org updated to: 0.10-a0-1160-g5c90495 (32) 00:07:44 evilmike: oooh 00:07:48 evilmike: no but that's good 00:07:57 evilmike: if you are adjacent to a monter, it will not flee. 00:08:04 unless there are othe rmonsters adjacent to iit and you. 00:08:13 I guess I can add it to the page then 00:08:28 well, I'm not entirely certain but that seems like the best of both worlds... you can easily prevent monsters from fleeing 00:08:49 It still retains the ranged combat fleeing nerf (and it makes sense that a monster would flee if you were shooting at it) 00:09:04 But it also doesn't remove the fleeing element -- if you both mutually decide to flee, the monster skedaddles 00:09:05 -!- Zaba has quit [Ping timeout: 258 seconds] 00:09:20 So my first impression is that it is the best of all of the options, but I would need to examine it more 00:10:37 -!- Zaba has joined ##crawl-dev 00:10:48 * due currently working on dragon breath. 00:11:18 i heard something about a timeout on it? 00:12:44 yeah 00:12:49 currently 10 - (HD/3) 00:13:05 It's a little bit fussy because breath weapons are handled differently for the different dragons. 00:14:01 What about draconians? 00:14:22 Not handling them currnetly 00:14:29 their timeout needs to be slightly different 00:14:37 I'm thinking symmetrical to the player timeout would be best 00:15:29 Possibly. I would be a bit concerned about players "timing" it manually 00:15:38 For dragons? 00:15:42 it would be relevant for quicksilver dragons, at least 00:15:51 ver, very true 00:16:33 could work around that by making the timeout be as low as 1 turn, maybe 00:16:58 Wielding a vampiric weapon while not satiated warns of same, and then wields it anyway, resulting in death by starvation! (https://crawl.develz.org/mantis/view.php?id=4696) by Twilight 00:17:55 randomly, I mean... so the timeout could be from 1 to 10 - (HD/3) 00:19:03 Windows builds of master branch on crawl.develz.org updated to: 0.10-a0-1160-g5c90495 00:19:22 That is what I am now leaning to 00:19:28 1+random2(10-(HD/3)) 00:19:50 Which is min=1, max=4. 00:19:56 @??quicksilver dragon 00:19:56 quicksilver dragon (11D) | Speed: 15 | HD: 16 | Health: 68-116 | AC/EV: 10/15 | Damage: 25, 20 | Flags: sense invisible, fly, !sil | Res: 06magic(149), 12drown | Chunks: 07contaminated | XP: 3546 | Sp: b.quicksilver (3d20). 00:20:11 which is min 1 turn for quicksilver, max 5 turns.s 00:22:15 excellent 00:22:16 It seems to work 00:24:02 elliptic: or more specifically, pong here 00:24:12 hi 00:24:19 elliptic: dragon breath weapon! do you think 1+random2(10-(HD/3)) is reasonable? 00:24:30 (timeout) 00:25:07 would this affect the dragons that are completely pathetic in melee too? 00:25:10 @??mottled dragon 00:25:11 mottled dragon (13D) | Speed: 10 | HD: 5 | Health: 16-39 | AC/EV: 5/10 | Damage: 15, 6 | Flags: fly, !sil | Res: 06magic(20), 05fire, 03poison | Chunks: clean/none/unknown | XP: 155 | Sp: sticky flame splash (3d4). 00:25:15 hm 00:25:20 yes 00:25:21 oh that has damage 15? I didn't realize 00:25:27 @??death drake 00:25:27 death drake (15l) | Speed: 13 | HD: 9 | Health: 56-99 | AC/EV: 6/14 | Damage: 12 | Flags: cold-blooded, fly, !sil | Res: 06magic(36), 03poison, 04rot | Chunks: 04rot | XP: 829 | Sp: miasma (3d9). 00:25:30 not drakes 00:25:38 @??steam dragon 00:25:38 steam dragon (02D) | Speed: 10 | HD: 4 | Health: 20-40 | AC/EV: 5/10 | Damage: 12, 6 | Flags: fly, !sil | Res: 06magic(16) | XP: 183 | Sp: steam ball (3d10). 00:25:39 just MONS_DRAGON based monsters currently 00:25:48 ... 00:25:51 which includes swamp drakes 00:25:56 fire drakes and lindwurms. Hm. 00:26:01 and death drakes! bugger. 00:26:09 it sounds a little high to me maybe 00:26:32 for a normal dragon its max 6 min 1 turns timeout. 00:27:37 for a steam dragon it's max 9 min 1. 00:28:04 monster draconian is... 00:28:16 3 + random2(10) + random2(30 - XL) 00:28:41 Acid and steam get half that. 00:29:53 but the numbers can get tweaked easily 00:30:32 I'd tend to suggest something simpler like 1d5 for all dragons 00:30:35 but I don't really know 00:30:42 hm 00:30:45 well, that's possible 00:30:57 Okay, let's go with that 00:31:55 -!- MarvinPA_ has joined ##crawl-dev 00:31:59 MarvinPA_! 00:32:35 MarvinPA_, I am just implementing dragon breath weapon timeout. elliptic suggests 1d5 for all dragons, but is unsure; my original idea was 1+random2(10-(HD/3)), but I am thinking 1d5 could be a good starting point; do you have any thoughts? 00:34:07 -!- MarvinPA has quit [Ping timeout: 248 seconds] 00:34:13 oh. 00:34:18 I suppose it's like... 4am over there. 00:34:26 okay, 1d5 it is for now, subject to change 00:36:53 ... why is there code calling static_cast(mons->type)? 00:37:33 I'll leave yellow draconians as-is. 00:38:51 and static_cast(monster->number) O_o 00:39:05 ah 00:39:15 random2 is signed 00:48:42 -!- Pacra has quit [Quit: This computer has gone to sleep] 00:52:54 -!- Pacra has joined ##crawl-dev 01:05:46 -!- Zaba has quit [Ping timeout: 276 seconds] 01:15:57 -!- Zaba has joined ##crawl-dev 01:24:24 -!- Mottie has joined ##crawl-dev 01:24:28 Hello Friends 01:24:47 Does anyone know how to code in C? :V Im kinda doing homework and getting rediculous results 01:24:54 apparently getchar is a terrible, terrible thing 01:30:02 well 01:30:10 I develop DCSS but I don't really code much in C 01:30:18 whence is thine getchar? curses? 01:32:08 Just homeworks and the likes, Im working with strings and I just realized stdin DERP. 01:34:05 you may have more luck with ##Crawl-offtopic 01:34:40 True, thanks man 01:41:45 there's also ##c 01:45:44 -!- Pacra has quit [Quit: This computer has gone to sleep] 02:15:37 -!- ainsophyao has joined ##crawl-dev 02:22:22 -!- ortoslon has joined ##crawl-dev 02:22:35 -!- galehar has joined ##crawl-dev 02:23:11 why do clouds from vial traps leave travel exclusions? 02:23:22 because it's buggy 02:23:39 -!- ainsophyao has quit [Remote host closed the connection] 02:23:46 hi galehar 02:25:25 galehar: elliptic suggest 1d5 for drag/drac breath timouet, so I am going with this as an initial attempt 02:25:57 for monsters? 02:26:21 yeah 02:26:44 currently, there's none, right, 02:26:46 ? 02:26:49 yeah 02:27:07 I originally suggested 1+random2(10-(HD/3)), but that ended up being 02:27:10 well, then that sounds reasonable 02:28:27 quite high 02:28:33 whoops, misentered there 02:28:41 okay! I will go with that then. 02:29:56 it seems to work relatively perfectly 02:38:53 -!- syllogism has joined ##crawl-dev 02:52:08 -!- monqy has quit [Quit: hello] 02:52:45 -!- Mottie has quit [Ping timeout: 258 seconds] 02:55:48 -!- mikee_ has quit [Ping timeout: 245 seconds] 03:21:48 @??bone dragon 03:21:48 bone dragon (15D) | Speed: 8 | HD: 20 | Health: 205-274 | AC/EV: 20/4 | Damage: 30, 20, 20 | Flags: 07undead, evil, sense invisible, fly | Res: 06magic(106), 05fire, 02cold, 10elec, 03poison, 12drown, 04rot+++, 13neg+++, 13torm | Vul: 08holy++ | XP: 2352. 03:28:02 -!- capablanca has joined ##crawl-dev 03:28:06 -!- capablanca is now known as mikee_ 03:30:21 rebase time :9 03:31:14 morning :) 03:31:42 -!- Twinge has joined ##crawl-dev 03:33:17 * kilobyte sniff Napkin. 03:33:49 * Napkin continues to sip his coffee * 03:34:10 hi 03:34:12 Napkin: I looked at a bunch of vi clones... too bad, the three groups: simple ones, those with "rvi" and those with Unicode support are disjoint :( 03:34:14 oh excellent, no pain 03:34:27 looks like someone how enum.h other than me though 03:34:37 serves me right for coding so slowly 03:34:43 kilobyte: I am about ot push dragon breath weapons and cherubs 03:34:52 ouch, kilobyte 03:34:59 so if we want one, it's a task of taking a vi clone and ripping away commands that can access outside files 03:35:00 but thanks for checking up on that again :) 03:35:28 * due waves napking 03:35:36 and there is the problem.. since I don't even know all the commands which give outside file access 03:35:42 hey due :) 03:36:01 how could I possibly manage to rip them all out?! 03:36:03 it might take a while until I'll get to that... so if you could apply just the ee patch and install it, we'd have at least one way to use Unicode literals in monster glyphs 03:36:38 Napkin: yeah, modern clones of vi are hundreds of megabytes of complex commands 03:37:02 bah, upgrading dgl.. okok 03:37:03 remember when "emacs" meant "Eight Megabytes And Constantly Swapping"? 03:37:12 been postponing it ;) 03:37:21 hehe 03:37:23 i have bene postponing by music practice to finish these! 03:37:31 I feel bad for having neglected crawl so much :( 03:37:41 how did your project go, due? 03:40:10 awesomely! I finished it :D 03:41:09 -!- ais523 has joined ##crawl-dev 03:41:11 ahhh! I'm using ee and virus as external programs :D no need to fiddle with DGL upgrade :) 03:41:14 done, kilobyte 03:41:22 great, due :)) 03:42:34 could you please verify ee works, kilobyte? 03:44:25 * due grumbles 03:44:30 mon-speak.o and ... slowly 03:45:40 -!- Zao has joined ##crawl-dev 03:48:20 Napkin: yay! 03:48:30 excellent :) 03:50:42 -!- ais523 has quit [Read error: Connection reset by peer] 03:55:34 -!- ais523 has joined ##crawl-dev 03:56:57 kilobyte: nowadays emacs means "Eight-hundreds Megabytes And Constantly Swapping". I'm sure that it will mean s/hundres/thousand/ soon 03:58:17 -!- Zao has left ##crawl-dev 04:01:53 ehmacs? etmacs? 04:02:13 -!- st_ has joined ##crawl-dev 04:06:33 no, only capital letters are used for acronyms :) 04:08:28 Some small changes to a few abyss rune vaults I made (https://crawl.develz.org/mantis/view.php?id=4697) by evilmike 04:08:45 eventually it will be "exabytes" and it can be eemacs though 04:09:54 that day might come sooner than one would think 04:24:58 evilmike: as one of biggest vault makers, perhaps you would want to take a stab at the holy Pan level? I have a bunch of ideas but not anything for the whole level, and it should be good. 04:26:23 ideas include a pasture (green floor) of apises, a "throne room" rectangular church with pairs of every kind of angels, one on each side of the room, leading to an altar 04:26:51 Interesting. It's an encompass vault right? 04:27:18 Also, I am assuming it puts all the portals in one room, since there's no holy rune? 04:27:35 oh, and custom redefined angels. The stub in pan.des already have seraphs: rF+ cherubs with much more HD but only a blessed great sword of flaming as a weapon 04:27:39 and a tile 04:28:16 It would be interesting to have three temples (or maybe something smaller, like shrines), one for each of the holy gods 04:28:19 evilmike: my idea was an encompass one, yeah... but it could be just a serial vault, more in line with non-unique pan levels, yeah 04:28:43 yeah, what I had in mind were side shrines and one big temple 04:28:51 when I was trying to make big god-themed vaults I didn't bother with the good gods, because there was really nothing to do (couldn't figure out how to differentiate tso and zin either) 04:28:52 with Ely's shrine being more of that pasture 04:29:02 do silver stars work yet? 04:29:57 not yet, the plan is damn straightforward though 04:30:20 orb of fires with s/fire/silver/;s/bad mutations/removing good mutations/ 04:31:16 beh... seems so little work that I'll make a stab, hopefully before due gets his new cherubs ready 04:31:40 heh... with 3 altars I'm suddenly getting a picture of nethack's astral plane 04:31:47 ... 04:31:52 not that the holy pan level soudns anything like it, of course 04:32:01 see, this is an interesing idea 04:32:19 heck, it sounds so tempting we might have two levels on the first stab :) 04:33:10 who's Pest though? :p 04:33:47 * kilobyte fondly remembers playing as an unchanging green slime, then sliming + taming all the Riders on astral :) 04:34:27 haha, that sounds cool 04:35:24 so, how about giving gdr a conservation effect? (to even out the field between AC and EV based builds regarding item destruction) 04:35:26 I suppose if it's an encompass vault, the player would still get randomly placed somewhere, just like any other pan level 04:35:40 which would necessitate a rather open / nonlinear layout I think 04:40:50 is there any rule yet on which holies belong to zin and which are tso? 04:50:01 I don't think there needs to be any hard split... most can be shared; the ones with a strong exclusive allegiance are apises and silver stars so far 04:50:27 even daevas which are biased to TSO sometimes serve Xom instead... 04:51:17 hmm. I wonder why the chance for abyssal rune is inside the item placement loop with only 1 chance in 200. It makes the overall probability hard to figure out, and maybe too low. 04:51:36 SamB, the bias of big room placement is caused by laziness of the way its position and size is generated with :P 04:51:40 and moin 04:53:00 -!- ainsophyao has joined ##crawl-dev 04:53:42 Lua error gives away flagged portal (deprecated reaching brand) (https://crawl.develz.org/mantis/view.php?id=4698) by absolutego 04:57:25 kilobyte: cherubs are ready 04:57:27 i am about to push ;) 04:58:49 pushed :D 05:00:03 OSX builds of master branch on crawl.develz.org updated to: 0.10-a0-1160-g5c90495 05:00:32 03due * r530f8949f2f3 10/crawl-ref/source/ (8 files): Implement breath timeouts for dragons, draconians, and, slightly, drakes. 05:00:38 03due * rdfe1758b7f57 10/crawl-ref/source/ (17 files in 3 dirs): Overhaul Cherub flavour, attacks, messages, etc. 05:02:00 :D 05:04:12 -!- ainsophyao has quit [Remote host closed the connection] 05:08:18 -!- mikee_ has quit [Ping timeout: 245 seconds] 05:09:48 Mummy race shouldn't drown (https://crawl.develz.org/mantis/view.php?id=4699) by petzl 05:09:51 -!- cbus has quit [Ping timeout: 245 seconds] 05:10:38 -!- capablanca has joined ##crawl-dev 05:10:47 -!- Pingas has joined ##crawl-dev 05:22:00 -!- edlothiol has joined ##crawl-dev 05:23:04 -!- ais523_ has joined ##crawl-dev 05:23:25 -!- ais523 has quit [Ping timeout: 258 seconds] 05:35:09 -!- ainsophyao has joined ##crawl-dev 05:35:43 -!- ais523_ has quit [Read error: Connection reset by peer] 05:36:18 -!- alefury has joined ##crawl-dev 05:36:57 -!- ais523_ has joined ##crawl-dev 05:38:57 -!- Torokasi has joined ##crawl-dev 05:40:11 hi 05:42:58 Exiting ziggurat with abyss makes the game crash when I'm trying to escape from the abyss (https://crawl.develz.org/mantis/view.php?id=4700) by Surr 05:47:12 kilobyte: I've been thinking of ideas that could work in a holy pan level. What if the level still had some pan demons in it, but they're imprisoned or at least sealed away? Zin does seem fond of imprisonment after all, and it would give the angels a reason to be there... 05:47:27 I don't mean all over the level, just in one area of it 05:51:38 i think having several guaranteed but randomly placed vaults could be used for more of the special levels 05:52:13 do you mean serial vaults? 05:52:17 so there could be a prison on the level, a boss castle, some guarded loot rooms, etc 05:52:25 yeah, kind of, but for a special level 05:52:33 like a branch end or something 05:52:47 or a holy pan level :) 05:53:11 i've found it's hard to get serial vaults to really cooperate with stuff like that, at least when experimenting with them 05:53:13 could be nice for a crypt ending too 05:54:07 for a pan level, there's also the problem of exits: if the level has no rune, one of the main goals would be to just escape from the level. I think if its serial valut based it can normal portals 05:54:21 like, even if you put all the exits in a guaranteed vault, it could still put hatches in the random part 05:54:40 well, there would have to be a flag: no random portals 05:54:44 or no hatches 05:55:04 besides, there could be good loot 05:55:17 holies are certainly strong enough to guard some fancy things 05:56:11 -!- ainsophyao has quit [Remote host closed the connection] 05:57:32 I tried to make a serial vault once, it was a ruined village in the swamp. Problem was it never would place more than 3 or so of the buildings... couldn't find a good way to work around that 05:58:00 make them small 05:58:22 is there a way to guarantee all the vaults get placed? without overwriting each other? 05:58:46 well, yes 05:58:53 but you shouldn't insist on that 05:59:23 well, for a branch end it could just be part of randomness, just the rune vault would have to be guaranteed if there is a rune 05:59:45 same for pan level, the exit vault would have to be placed, the rest could be random 05:59:52 you can always use subvaults to restrict them to regions that are guaranteed to not overlap 05:59:58 like layout_roguelike does 06:00:15 is that the one with the rooms? 06:00:17 I suppose small is like 10x10? 06:00:19 well, you can't use subvaults right in the layout itself for everything 06:00:23 evilmike, no, way smaller 06:00:35 5x5 is pushing it 06:00:42 ah, I see 06:00:51 a serial vault wouldn't be suited for a ruined town then 06:01:00 you can't use subvaults in C++ layouts, not even with a stretch, and you can only use them for maps that are prepared to be used as subvaults 06:01:49 and you can't write layouts that fill a level with a primary vault in non-C++ 06:02:08 -!- ais523_ has quit [Ping timeout: 248 seconds] 06:02:15 Zaba: is there any reason for that restriction on lua-based layouts? 06:02:27 it's intereting that the gnoll castle serial vault uses stuff that's mostly larger than 10x10 06:02:53 gnoll camp, I meant 06:03:01 there usually only are about 3 gnoll rooms though, so this is consistent with what you noticed with your vault 06:03:20 hm 06:03:20 ah, good point 06:03:20 due, lua-based layouts are handled just like normal vaults 06:03:20 Zaba: ah, true. 06:03:34 Zaba: what kind of effect would layout-restriction (i.e., open levels) have on subvaults -- I';m guessing it would mean there's more space for the placer to work with, but I'm unsure. 06:03:39 due, other vaults on the level don't exist to them, it'd break encapsulation :P 06:04:25 due, subvaults in what? 06:04:35 erm 06:04:38 serial vaults, rather 06:04:43 -!- ortoslon has quit [Quit: bye] 06:04:47 ah 06:05:17 perhaps we could make bulk of a layout in post-processing? Ie, place the vaults first but with a margin between them, and only then make the layout itself. 06:05:19 if you mean bigger_room, well, vaults tend to be placed in its walls 06:05:35 kilobyte, some layouts really aren't prepared for working like that 06:05:42 kilobyte: most of the layouts can't... cope with that yeah 06:06:35 kilobyte, you could of course throw in basic MMT_VAULT awareness into them but vaults would mess them up just as much as they do now, and it would improve nothing 06:07:56 kilobyte, generalizing the layouts that are applied as post-processors would be nice, though 06:20:24 Zaba: of course, yeah, most existing layouts assume a blank map 06:20:26 -!- ais523_ has joined ##crawl-dev 06:26:14 -!- ainsophyao has joined ##crawl-dev 06:36:26 -!- ortoslon has joined ##crawl-dev 06:38:08 03dolorous * re27dbb28cdfa 10/crawl-ref/source/dat/database/monspeak.txt: Tweak wording. 06:40:25 -!- ais523_ is now known as ais523 06:50:14 -!- capablanca is now known as mikee_ 07:00:58 -!- casmith_789 is now known as casmith789 07:24:27 -!- ainsophyao has quit [Remote host closed the connection] 07:36:45 -!- Mottie has joined ##crawl-dev 07:40:08 -!- evilmike has quit [Read error: Connection reset by peer] 07:43:31 -!- ncampion has quit [Quit: Linkinus - http://linkinus.com] 07:46:49 -!- ais523 has quit [Remote host closed the connection] 07:50:36 -!- galehar has quit [Quit: Page closed] 07:54:29 -!- mikee_ has quit [Quit: HydraIRC -> http://www.hydrairc.com <- The professional IRC Client :D] 08:13:34 -!- Mu_ has joined ##crawl-dev 08:23:43 -!- MarvinPA has joined ##crawl-dev 08:24:53 -!- MarvinPA__ has joined ##crawl-dev 08:26:57 -!- MarvinPA_ has quit [Ping timeout: 255 seconds] 08:28:29 -!- MarvinPA has quit [Ping timeout: 256 seconds] 08:33:13 -!- valrus has joined ##crawl-dev 08:38:18 -!- MarvinPA__ is now known as MarvinPA 08:48:08 03kilobyte * rd2dbe0ea19d1 10/crawl-ref/source/ (enum.h spl-data.h spl-zap.cc zap-data.h): Rename "Light Beam" to "Holy Light" to match the BEAM and to disambiguate vs Sunray. 08:48:08 03kilobyte * r59152f3bb121 10/crawl-ref/source/ (abl-show.cc dat/database/help.txt): Whitespace fixes. 08:48:09 03kilobyte * reca2fcaf174c 10/crawl-ref/source/mon-stuff.cc: Stop the arena from immediately crashing. 08:48:09 03kilobyte * r9ca9327b7874 10/crawl-ref/source/ (9 files in 2 dirs): Silver stars, with their dreaded bolt of Zin spell. 08:57:34 -!- valrus has quit [Remote host closed the connection] 09:28:24 -!- alefury has quit [Quit: ChatZilla 0.9.85 [Iceape 2.0.11/20110701115916]] 09:31:29 -!- Twilight-1 has quit [Read error: Connection reset by peer] 09:31:49 -!- Twilight-1 has joined ##crawl-dev 09:46:08 03MarvinPA * r1e127255f126 10/crawl-ref/source/acquire.cc: Bump the quantities of ammo acquirement from scroll some more 09:46:16 03MarvinPA * ra7a16c747328 10/crawl-ref/source/acquire.cc: Make decks from misc acquirement much more likely to be legendary or ornate 10:01:21 -!- Zaba has quit [Ping timeout: 244 seconds] 10:12:05 -!- Zaba has joined ##crawl-dev 10:32:55 -!- MarvinPA has quit [Ping timeout: 255 seconds] 10:34:22 * SamB becomes frightened when he discovers his source/html directory missing for no obvious reason 10:34:53 * SamB is reassured that his compile.sh is still there; he thought maybe it had been removed by dpkg-buildpackage 10:35:04 * SamB backs it up into ~/.crawl just the same 10:36:14 -!- MarvinPA has joined ##crawl-dev 10:39:48 -!- Pacra has joined ##crawl-dev 10:40:35 kilobyte: so, did you see this yet: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=644580 ? 10:46:13 * SamB looks at the buildd logs to see what exactly was used... 10:56:31 -!- MarvinPA has quit [Ping timeout: 248 seconds] 11:04:06 SamB: yeah, 0.8.1-1 is long since replaced by 0.9-*, and never was a part of a stable release. 11:04:20 kilobyte: that's not the point at all 11:04:32 I want them *all* built with debugging symbols 11:04:40 -!- monqy has joined ##crawl-dev 11:05:09 a -dbg package might be added, but I really doubt Guus Sliepen would make a sourceful upload just for that 11:05:28 "sourceful"? 11:06:01 a binnmu may at most rebuild against a new set of libraries 11:06:09 ah 11:06:13 is that all you're allowed to do? 11:06:55 I directly can't do anything, at most send packages to Guus 11:07:32 I do run 0.8, 0.9 and trunk builds thrice a week, placing them on CDO's apt repository though 11:07:35 does debian have a strange definition of "uploader", then? 11:12:10 * SamB wishes he could google monospaced text 11:14:11 Luterac the Chiller (L3 DEIE) ASSERT(prog >= 0) in 'skills2.cc' at line 152 failed on turn 1914. (D:3) 11:16:30 sis, her husband and kid are here so I'm mostly AFK 11:16:44 pushed my current Debian trunk mess 11:17:21 Paralysis from zot traps/hell effects is seemingly fixed at 10 turns (https://crawl.develz.org/mantis/view.php?id=4701) by st 11:17:21 this mess is messy though, and I currently keep rebasing it so I don't keep it up-to-date on gitorious (sorry!) 11:18:08 Luterac the Chiller (L3 DEIE) ASSERT(prog >= 0) in 'skills2.cc' at line 152 failed on turn 1914. (D:3) 11:18:09 03kilobyte 07debian-trunk * r9a51e584e498 10/crawl-ref/.gitignore: Add ignores for contribs and prebuilt yaccage. 11:18:10 03kilobyte 07debian-trunk * re4b4c019c902 10/crawl-ref/debian/ (16 files in 2 dirs): Debian's 0.8.0-2 packaging, uupdated to 0.8.1 11:18:10 03kilobyte 07debian-trunk * r7a92fe2ba945 10/crawl-ref/debian/ (changelog rules): Bring back the workaround on gcc-4.4 bug on powerpc. 11:18:10 03kilobyte 07debian-trunk * rd86e332aedd6 10/crawl-ref/debian/NEWS: Remove an useless NEWS entry. 11:18:10 03kilobyte 07debian-trunk * r090aab688735 10/crawl-ref/debian/rules: Support cross-compiling. 11:18:10 03kilobyte 07debian-trunk * r30f4fbd7ee0e 10/crawl-ref/debian/rules: Update clua -> dlua. 11:18:10 03kilobyte 07debian-trunk * r45a5fe16f4a6 10/crawl-ref/debian/rules: Make the {build,binary}-{indep,arch} targets do what they are supposed to do. 11:18:20 03kilobyte 07debian-trunk * rc5529390c5ec 10/crawl-ref/debian/copyright: Remove the MT19937 license, it's no longer used. 11:18:21 03kilobyte 07debian-trunk * ra4ad5915fd46 10/crawl-ref/debian/changelog: Changelog from 0.9.0-1. 11:18:21 03kilobyte 07debian-trunk * r62ebc0d9de38 10/crawl-ref/debian/changelog: Changelog from 0.9.1-1. 11:18:21 03kilobyte 07debian-trunk * r5e9623752066 10/crawl-ref/ (.gitignore debian/source/options): Work around breakage of quiltless workflows in dpkg 1.16.1. 11:18:21 03kilobyte 07debian-trunk * r5008debb3309 10/crawl-ref/debian/rules: Obey flags set by dpkg-buildflags. 11:18:24 Luterac the Chiller (L3 DEIE) ASSERT(prog >= 0) in 'skills2.cc' at line 152 failed on turn 1914. (D:3) 11:30:57 -!- greatzebu has joined ##crawl-dev 11:32:38 kilobyte: are you using git-buildpackage? 11:32:54 ouch 11:33:19 looks like I need to go in and sort out the Python upgrade 11:35:31 kilobyte: oh, there's a CCACHE option we should probably set in Makefile when PCH is enabled 11:36:15 CCACHE_CPP2 11:37:33 tells ccache not to feed the output of its preprocessing run to the compiler 11:37:37 kilobyte: when you get a chance I'd like to talk about how to improve the level changing code. I'd like to fix it up, but there are parts I don't understand and I'm concerned about unintended consequences if I just dive in and start changing things 11:37:53 then, maybe ccache compiles would see a speedup with PCH, too 11:38:38 since, you know, the compiler would actually have a chance to *use* the PCH 11:43:43 -!- dpeg has joined ##crawl-dev 11:43:51 !seen evilmike 11:43:51 I last saw evilmike at Fri Oct 7 12:40:08 2011 UTC (4h 3m 43s ago) quitting with message Read error: Connection reset by peer. 11:44:09 !seen elliptic 11:44:09 I last saw elliptic at Fri Oct 7 05:54:59 2011 UTC (10h 49m 10s ago) saying drie on ##crawl. 11:44:22 !seen MarvinPA 11:44:22 I last saw MarvinPA at Fri Oct 7 15:56:31 2011 UTC (47m 51s ago) quitting with message Ping timeout: 248 seconds. 11:44:25 cache hit (direct) 0 11:44:25 cache hit (preprocessed) 13061 11:44:25 cache miss 17616 11:45:03 so it looks like something makes preprocessed output to never match 11:45:13 s/to // 11:46:28 kilobyte: huh? 11:47:06 did you -z before trying it? 11:48:07 * SamB wonders why wiki.debian.org doesn't remind you to use FirstnameLastname on the login screen 12:01:12 kilobyte: 12:01:12 cache hit (direct) 48504 12:01:12 cache hit (preprocessed) 368520 12:01:12 cache miss 246404 12:01:23 crawl@cdo 12:03:48 I quite rarely get a "direct" cache hit 12:06:40 kilobyte: note that ccache *will* still preprocess and hash to look for cache entries if it doesn't find one based on the "direct" method; the option just tells it not to compile the already-preprocessed output directly 12:07:18 and that preprocessing AppHdr.h{,gch} is going to be a lot cheaper than actually compiling it 12:08:48 (I'm not really sure if -E runs can use PCH or not) 12:10:17 grrr 12:13:10 SamB: ah, it will throw it away, right? 12:13:58 greatzebu: I'm unifying the code for branches, level_type and lua portals into one. 12:20:10 kilobyte: yes, after hashing it 12:22:41 hmm, what does level_type do now? 12:23:04 I mean, is it unique for each branch anyway? 12:23:32 -!- oberstein has quit [Ping timeout: 260 seconds] 12:24:00 -!- Brannock has joined ##crawl-dev 12:24:24 SamB: yeah, for any branch that you cannot autotravel to 12:24:49 those which are connected have level_type of LEVEL_DUNGEON 12:24:53 oh 12:24:56 okay 12:25:14 so you're doing what, then? 12:25:30 so there are two fields: either level_type == LEVEL_DUNGEON and branch names the place, or level_type is something else and branch is -1 12:25:46 okay 12:25:54 I'm removing level_type, so it'll be BRANCH_ABYSS or BRANCH_LABYRINTH 12:26:12 -!- oberstein has joined ##crawl-dev 12:26:13 what happens for those portalled areas? 12:26:37 there's also third case: LEVEL_PORTAL_VAULT, where the actual branch is known only to lua, causing "place" to have incomplete information 12:26:39 * SamB never understood whether the portal vault was the vault with the portal in it, or the thing on the other side of the portal 12:26:51 portal vault is the thing on the other side of the portal :P 12:27:17 which causes bugs like the recent ones... "Zig:16" cannot be mapped to a "place" since the only place is "Port" 12:27:51 BTW, it is astonishing how much code duplication can be removed 12:28:33 so, is this going to cause problems for henzell? 12:28:46 like, floor colour or default tiles: BRANCH_* has it set in branch-data.h, LEVEL_* in C++ code, portals in lua functions 12:29:04 hardcoding information about every portal vault type doesn't sound like fun, though 12:29:12 I was going to say 12:29:19 it should be possible to do in C++ or lua 12:29:37 but a more unified mechanism would obviously be better 12:29:58 yes 12:29:58 Zaba: you'll have to define an enum and an entry in branch-data, but otherwise you can do everything in lua 12:30:17 and especially, all the portal selection code can be replaced with DEPTH:WizLab 12:30:23 kilobyte, would be nice to be able to define those from lua somehow 12:30:25 or Zig:6-27 12:30:31 hmm, are incomplete enums allowed? 12:31:00 * SamB dreads all the rebuilds if not 12:31:29 SamB: you don't add a portal branch every week... 12:31:31 kilobyte, also, how are you going to handle &P 12:31:47 &Playout is very useful 12:32:11 -!- Danei has quit [Quit: Holy shit, this is going to be awesome.] 12:32:29 kilobyte: okay, *maybe* putting an enum entry in wouldn't be too bad 12:32:32 Zaba: it can select the vault then... 12:33:10 but what will you set the branch to for a portal vault created that way? 12:33:16 and you'll be able to &~ Zig:27 12:33:26 but what if you don't want to have to hardcode the colors and such? 12:33:41 SamB: LFLOORCOL 12:34:29 * SamB thought momentarily that kilobyte was laughing at him 12:34:57 (the L...OL may have caused that) 12:36:07 -!- MarvinPA has joined ##crawl-dev 12:37:41 oh, sweet 12:37:42 Zaba: actually, the question after entering &P is basically asking "which layout should the map just entered use?" 12:38:01 Zaba: which would be useful for any other place but old Port too 12:38:18 that aptitude bug with the "automatically installed" flag is supposed to be fixed in unstable 12:39:19 well, I'm still waiting for the "remove world, install a few unrelated packages, don't install the package you were asked to install" bug to be fixed, before I touch aptitude again 12:40:05 yeah, that one is kind of annoying 12:40:54 but I had already switched to manual resolution anyway before that was added 12:41:39 so it's not actively destructive, more of an annoyance 12:44:07 03kilobyte * r01f0e7b42a91 10/crawl-ref/source/mon-data.h: Use the recent flag M_NO_FLEE to have manticores have proper animal intelligence. 12:53:56 * SamB thinks maybe aptitude would do a better job at avoiding that if it assigned a -5000/n score to not doing each of n requested operations 12:56:27 -!- ortoslon has quit [Read error: Connection reset by peer] 13:01:51 -!- ussdefiant has quit [Ping timeout: 248 seconds] 13:07:19 Napkin: it looks like cdo is ailing? 13:07:57 My SSH sessions are authenticating and then dropping off abruptly for no evident reason 13:07:59 pardon? 13:08:40 turn off your torrents? ;) 13:08:54 and sgiratch says he has lots of lag 13:08:55 traceroute to you: 13:08:56 10. tge4-1.aggr1.phdl.pa.rcn.net 0.0% 18 88.6 88.6 88.5 88.8 0.1 13:08:59 11. port-chan1.upd-ubr1.trpr-upd.pa.cable.rcn.net 0.0% 18 89.4 89.3 89.2 89.8 0.1 13:09:03 12. 207.172.114.83 0.0% 18 351.1 512.6 167.4 747.6 164.3 13:09:49 hmm 13:09:53 you may have a point 13:10:11 but I'm pretty sure that's not why I'm getting dropped... 13:10:15 there is a webtiles player though, calls himself "hackum" 13:10:32 who manages to make his crawl process use up to 80% of one core 13:11:08 what's the memory usage like? 13:11:19 but still 13:11:20 $> uptime 13:11:20 20:11:08 up 177 days, 23:26, 24 users, load average: 0.70, 0.83, 0.61 13:12:58 7265340 free kb + 19gb of free swap 13:14:04 what is the xl27 mummy hackum doing in d2? 13:14:21 and how is it causing so much cpu usage? 13:14:25 -!- dpeg has quit [Quit: Lost terminal] 13:14:37 holding down "."? 13:14:58 crawl should really prevent some keys to be held down ;) 13:16:25 can it do that? 13:16:49 I guess it could just sleep between "."s or something 13:17:29 of course not prevent someone from holding down the key.. but preventing certain commands to be executed repeatedly every 3ms would be cool 13:18:35 anyways, I would guess on network, SamB - can't see usage which could cause severe log 13:18:47 *lag 13:22:54 Napkin: yeah, he has found he has 25% packet loss 13:23:07 damn 13:23:38 I've had worse but, yeah, that's bad 13:38:10 -!- Pacra has quit [Quit: This computer has gone to sleep] 14:02:42 oh, you know what I don't get? 14:03:30 why the only profiling build in our Makefile has debug spew (FULLDEBUG) enabled 14:04:02 that does seem rather counterproductive 14:04:31 i would guess there hasn't been much need for serious profiling so no one has addressed it 14:04:36 just delete it 14:05:12 greatzebu: there is a need, but I'm the only one among us who did any profiling in the last year, and I didn't even know about this target 14:05:41 it's too simple to CFOPTIMIZE="... -pg", and you need to if you want to control optimization flags anyway 14:06:59 greatzebu: I don't mean there's no point in profiling with FULLDEBUG on; not at all! 14:07:32 -!- dtsund has joined ##crawl-dev 14:07:40 just that it's not as useful in the end 14:08:52 did the guy who looked at my gnoll-castle-slowdown save not profile? or just not count as "among us"? 14:12:08 kind of the latter, but running a profiler would show iterators as being far more prominent than string mallocs 14:13:06 most of profiling in the past concentrated on sprint, though... including massive runs by greensnark and doy in 0.6 14:14:11 malloc itself is pretty cheap. the expensive thing is taking a lot of cache misses 14:14:20 -!- Pacra has joined ##crawl-dev 14:14:41 malloc+free isn't necessarily as cheap as you might think 14:15:07 oh, I did not apply the alloc-on-stack thingy yet 14:15:35 and aren't the cache misses kind of part of that? 14:15:49 it's surprising allocing that anew is faster than using a global 14:16:10 not really 14:16:17 stack cache is more likely to be hot 14:16:36 true 14:16:43 -!- jeanjacques has joined ##crawl-dev 14:17:06 most of my serious optimizing was in the 8086/80286 days so I kind of suck at that "cache" thing :p 14:17:18 yeah 14:18:00 back then it might have even been a good idea to use some of those CISCy opcodes, eh? 14:18:21 right 14:18:54 nowadays it just makes everyone go "huh? what does that do?" and look it up 14:19:01 programmer and CPU alike 14:19:37 heh, well described :) 14:21:05 this is why x86 decoders are a nightmare :) 14:21:22 greatzebu: disassemblers or decompilers? 14:21:41 SamB: i mean the decode unit of the cpu 14:22:13 well, as I said, they punt on the obscure CISCy operations 14:22:31 emulated in software 14:23:01 <|amethyst> but even so, the decoder still has to figure out how long the instruction is 14:23:12 <|amethyst> which in itself can be a pain 14:23:19 x86 in general is a monstrosity 14:23:22 of course, some operations may not actually run more slowly than they would have to do anyway if implemented thus 14:23:48 |amethyst: if it's going to punt to microcode, can't it let the microcode tell it that? 14:24:08 but I wasn't even talking about all of the addressing modes 14:24:10 <|amethyst> SamB: then it can't issue the next instruction until the microcode is done 14:24:57 <|amethyst> then again, maybe that's how it works 14:25:02 so your 286 code will only run 1000 times as fast, not 10000 times as fast. Big deal! 14:25:18 I don't know for sure, but I don't see any big obstacle here 14:25:20 <|amethyst> I'm also more familiar with the 386ish days than modern Intel processors 14:25:51 of course, if the oddest thing about the instruction is WHAT it does, then maybe the decoder can handle it fine 14:26:12 probably the more usual case, really 14:26:34 <|amethyst> aha, nowadays the intructions are translated to microcode independently of (and asynchronously from) the execution pipeline 14:26:44 yes 14:27:02 <|amethyst> and pretty much all the instructions are "obscure CISCy operations" from that point of view :) 14:27:45 <|amethyst> and apparently NetBurst actually caches micro-ops 14:27:54 <|amethyst> no idea about more recent architectures 14:27:55 by "obscure ciscy operations", I meant the ones that don't have quick \mu-op translations 14:28:00 i think they decided that was a bad idea, though 14:28:40 well, either way there are going to be some that are properly microcoded 14:28:42 there's no micro-op cache anymore, as far as i know 14:28:59 and some that are fast 14:38:53 IIRC they *did* use the Pentium M architecture as the basis for Core 14:39:13 oh, wait, I don't have to recall 14:39:46 Yeah, they basically say that they did. 14:40:35 anyway, I'm sure the decoders are a lot of trouble to get right 14:40:46 but I'm pretty sure that microcoded operations are NOT why 14:43:14 regardless, i think it's not a key issue determining whether crawl is fast or slow 14:44:59 probably not! 14:45:33 we were discussing kilobyte's experience with optimization ;-) 14:46:22 -!- Textmode has quit [Quit: Ex-Chat] 14:47:39 big words for "funny this theoretically slower operation is faster than the other version" :p 14:47:59 true, fair enough :-) 14:50:12 -!- galehar has joined ##crawl-dev 14:50:24 'evening 14:50:26 most of my work is very performance-sensitive and design-constrained so crawl is usually a nice break from that 14:51:08 -!- dtsund has quit [Quit: OWLS] 14:51:12 hello hello 14:51:15 I can imagine 14:52:08 03brendan * r6b40e00f7195 10/crawl-ref/source/terrain.cc: More lively lava damage messages 14:52:18 03MarvinPA * r223486ebe47b 10/crawl-ref/source/terrain.cc: Better messages for mummies drowning, tweak lava messages 14:54:22 -!- mong has joined ##crawl-dev 14:55:24 So I seem to have broken my CAO account (charname Blag) save file when transferring to the latest crawl version. It has been quite some time since I accessed that account, so I'm not sure what trunk version I had been playing on before. 14:56:09 I completed the save transfer as prompted, but now I get "ASSERT(prog >= 0) in 'skills2.cc' at line 152 failed." whenever I try to launch the game. 14:56:38 (I think, at least. The message only flashes onscreen very briefly, so I tried to transcribe it.) 14:57:32 MarvinPA: uhm... my dictionary lists "scorch" as "a surface burn", "make very hot and dry", "become superficially burned" -- which is so much weaker than the other verbs 14:57:40 and yet you made it the highest one 14:57:49 -!- ussdefiant has joined ##crawl-dev 14:59:30 hmm, i think scorch can also be used to mean "completely destroy" (with fire) 14:59:37 like, "scorched earth" or whatever 14:59:43 -!- mong_ has joined ##crawl-dev 15:00:10 roasting is kind of problematic as well, since it connotes slow cooking via indirect heat 15:02:30 -!- ussdefiant has quit [Ping timeout: 255 seconds] 15:03:08 -!- mong has quit [Ping timeout: 252 seconds] 15:03:09 <|amethyst> MarvinPA: if you think about it, "scorched earth" only destroys the *surface* of the earth :) 15:03:19 -!- mong_ is now known as mong 15:03:25 <|amethyst> s/it,/it in a certain way,/ 15:03:27 mong_: what happens after the message? terminal closes? 15:03:38 well, i'm not desperate to defend it or anything, they all sound like reasonable synonyms to me though 15:03:52 |amethyst: whatever you say, obi-wan 15:03:57 what with the context being "you've just fallen into some lava" :P 15:04:18 oh, a recursive crash too 15:04:21 "slagged" is hardly appropriate for non-metalic forms 15:04:28 the crash handler tries to display skill progress... 15:04:50 SamB: terminal returns me to the trunk game menu on cao 15:04:54 ah 15:05:00 okay, I can't do anything about that 15:05:20 kilobyte: oops that sounds like my doing 15:05:27 kilobyte: you have access on CAO? 15:05:32 galehar: yes, the word "skill" clued me in 15:05:48 MarvinPA: another dictionary (wordnet) has 7/8 definitions say "superficially" (or equivalent, like "discoloration"), one has "completely" 15:05:59 yeah, I did a couple of skill hacks ;) 15:06:10 "couple of small hacks"? 15:06:45 btw, I think the new skill system is finished 15:06:46 You reworkd practically the whole skill system code 15:07:03 MarvinPA: I'm just a dirty foreigner so you're certainly so much more likely to be right about obscure uses... yet if I did not hear that 1-out-of-8 version, it's likely a bunch of other foreigners didn't too 15:07:04 and it's not finished until that crash is fixed ;-P 15:07:25 or for that matter, those who don't look at the context too closely 15:07:42 galehar: no access to CAO, just reading http://crawl.akrasiac.org/rawdata/Blag/ 15:07:51 SamB: there will always be some fixes and adjustments, but I'm quite satisfied with the system as it is now 15:07:52 I'm a big fan of the new skill system, by the by. Good work to all involved in that one. 15:07:53 true, well it's no problem to tweak the order 15:08:12 could just go with scorch -> sear -> burn ... as was suggested on the mantis item? 15:08:19 mong: were you the one who said it was worth enduring the food reform for ? 15:08:33 Yeah :x 15:08:52 when most of us were hiding in -dev? 15:09:45 hey, some people like the food reform! I have hard evidence about that! 15:10:14 how many signed? 15:11:16 galehar: who's the lying sucker-up? :p 15:11:18 and are their PGP keys actually validated as belonging to distinct people? 15:12:07 03MarvinPA * r77e86a4c656d 10/crawl-ref/source/spl-miscast.cc: Make ench miscast paralysis use the standard 2-7 turns instead of being fixed at 10 turns 15:12:17 03MarvinPA * rd99eae9062c5 10/crawl-ref/source/terrain.cc: Tweak lava messages again 15:12:30 sanka said "The food changes I good in my experince, thank you for implementing." 15:12:44 smoke said "BTW, I love the food reform. It makes sense and it's simple. Very nice." 15:13:00 people like it? time to nerf 15:13:12 yes! 15:13:36 also, you guys are talking about searing and burning? 15:13:40 i haven't played trunk, but i think the idea behind food reform is very good 15:13:41 did cooking patch get added in finally 15:14:02 Eronarn: that's actually on the list of things we aren't going to do 15:14:25 kilobyte: maybe it would be better to not call you.skill from dbg-assert.cc and just dump the raw values. To avoid recursive crashes. 15:14:29 probably mostly because it would be a pain if it did anything useful 15:14:51 galehar: most likely, yes 15:17:36 galehar: definitely, yeah. Or divide it by hand. 15:17:36 a smooth user experience is more important than a delicious quokka burger 15:18:24 a good crash is better than a recursive crash 15:22:14 -!- ussdefiant has joined ##crawl-dev 15:23:52 btw, I was wondering how to force a crash when I wanted to test changes to the crash dump. The easiest I found: kill 15:24:59 -!- ais523 has joined ##crawl-dev 15:26:44 -!- ussdefiant has quit [Ping timeout: 252 seconds] 15:37:02 galehar: let's show it as as "xp above last/xp needed", like "112/119" 15:38:08 hmm... I'm getting 1 above the number it should be for skills at 27 15:38:31 needed 35378, skill_points 35379 15:38:43 kilobyte: I'm not sure it's worth bothering. Dumping the raw data is enough, we can always do the calculation by hand if needed. 15:39:26 I did that locally, will push later but want to have a look at a couple of patches on mantis before 15:40:53 oh dammit 15:41:14 sorry, pushed my stab just a second ago :( 15:41:46 it's ok, I don't mind :) 15:42:07 03kilobyte * r9196d74e9344 10/crawl-ref/source/xom.cc: Remove some obsolete Xom joy about losing runes. 15:42:08 03kilobyte * rc170fb1e10c7 10/crawl-ref/source/ (externs.h store.cc store.h): Add prop[const char*] to avoid mallocs of std::strings, add inlines. 15:42:08 03kilobyte * rc32100602fe2 10/crawl-ref/source/terrain.cc: Swap "scorches" and "scalds". 15:42:09 03kilobyte * r60f23966a412 10/crawl-ref/source/dbg-asrt.cc: Dump the skill progress as a rational number rather than a real one. 15:42:12 compare if your version is better, if so, just replace it 15:44:22 my version is certainly not better, I didn't bother calculating progress 15:49:09 -!- ussdefiant has joined ##crawl-dev 15:49:17 About 4 weeks from now, if there suddenly appeared a suggestion for a new branch - including concept art, complete tile pack, detailed description of every monster and feature, lore and an original, cool concept - what are the chances it would be up for development? 15:49:29 Assuming it's wickedly awesome. 15:49:34 And fun. 15:50:02 well, it would help if there was some kind of mathematical description of how to layout levels for it 15:50:14 easiest would be if it was all vaults 15:51:40 if people like it, i'd imagine the chances are very good 15:51:48 Pingas: hard to tell. It just depends if one of the dev like the idea. 15:51:56 but, why would the tiles exist at suggestion time? 15:52:17 probably it would be better to get more feedback before doing all that work though 15:52:36 also wickedly awesomeness and fun-ness seems hard to determine before actual implementation, yeah 15:52:41 I doubt if you'd have artists that misguided ;-) 15:52:44 Pingas: also, coding a whole branch is a huge amount of work. They usually start by being a portal vault (like spiders) 15:52:56 It was merely hypothetical. And yep. 15:53:34 of course, when I said "mathematical description of how to layout levels", I meant "code that generatese layouts" 15:53:36 -!- ussdefiant has quit [Ping timeout: 258 seconds] 15:53:39 starting with a portal vault is an excellent suggestion 15:53:42 those are actually the same thing 15:54:03 of course, vaults count as code ;-) 15:55:04 If only I could code well. 15:55:30 I just draw stuff. 15:57:30 good art is hard to come by 15:57:35 anyway, you could start with designing vaults--that doesn't require programming expertise 15:57:59 I'd love to learn how. 15:58:04 vaults don't count as code if there's no lua. There's 0% programming. 15:59:00 Pingas: https://crawl.develz.org/wiki/doku.php?id=dcss:help:maps 15:59:01 there is some vault writing documentation in docs/develop/levels 15:59:13 galehar: Will check out, thanks! 15:59:24 and mumra conveniently moved all of it on the wiki for easier reading 16:00:40 so, when the abyssal rune spawns, it has only 1 chance in 10 of using one of the many nice vaults that evilmike did. What a shame. How about we revert it? 9 chances in 10 for a vault, 1 in 10 for floor. 16:00:47 galehar: we aren't sure what we're going to do about preventing duplication of effort with that, actually 16:01:27 SamB: no. dpeg and kilobyte discussed it, but I don't when the end of it. 16:02:23 mumra and I, at least, seem to be agreed that it would be nice to keep a wiki copy and an in-tree copy in the same syntax 16:02:48 and that it would be best to use something that looks reasonably nice in a text editor 16:03:17 (such as reST or markdown) 16:03:55 galehar: i definitely vote for using abyssal rune vaults more often 16:03:56 that way, technical aspects can be updated in synch with the accompanying code changes 16:05:14 mumra is not at all tied to using docuwiki for this, and he hadn't actually started work on any dokuwiki->text conversion code yet when I talked to him 16:07:04 I think the idea of turning the whole whole docs/ tree of some clone into a wiki may have been mentioned 16:07:53 arg, I suck at vault syntax. Do I just need to change the abyss_rune dummy chance from 90% to 10%? 16:13:21 -!- MarvinPA_ has joined ##crawl-dev 16:16:27 -!- MarvinPA has quit [Ping timeout: 244 seconds] 16:18:11 03galehar * r9f240e8a4664 10/crawl-ref/source/wiz-fsim.cc: Fix a calculation error in ranged combat fsim. 16:18:11 03galehar * r03059b7e87ff 10/crawl-ref/source/wiz-fsim.cc: Fsim: put a decimal in the average time. 16:18:19 03evilmike * r99ff13cff705 10/crawl-ref/source/dat/des/branches/abyss.des: A few changes to abyss rune vaults. 16:18:19 03galehar * r40ec3f0981fd 10/crawl-ref/source/wiz-fsim.cc: Some simplifications. 16:18:19 03galehar * r794ff840803a 10/crawl-ref/source/ (mon-behv.cc mon-movetarget.cc): Fix the retreating behaviour. 16:25:52 kilobyte: I upgraded http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=570377 to important 16:41:27 -!- dpeg has joined ##crawl-dev 16:42:10 !tell due Thanks for holy work! 16:42:11 dpeg: OK, I'll let due know. 16:52:12 Would anyone here object to occasional postings of cute? Such as this: http://i.imgur.com/4md4N.jpg 16:57:30 Objection! 16:57:33 * SamB hopes he won't regret looking at this ... oh, good, an actually kitten 16:59:10 -!- ussdefiant has joined ##crawl-dev 16:59:58 SamB is a brave man :) 17:00:16 dpeg: what, for looking at stuff that Aerdan posts? 17:00:49 yes! 17:00:52 I don't post gruesome shit. When I do, I make it absolutely clear what it is. 17:01:11 This isn't really the place for pictures of any kind, though. 17:01:18 You didn't give much of a proper warning for this abomination. 17:01:25 dpeg: but ... but ... the tiles? 17:01:56 mine is cuter anyway :p 17:02:01 SamB: okay, the tiles. I have to admit that I cannot distinguish tiles. Well, perhaps wall tile from monster tile but no more than that. 17:02:04 kilobyte: <3 17:02:08 and tiles, well, are an abomination too :p 17:02:26 A large abomination, I may point out. 17:02:36 yes, but one we have evidently decided to put up with 17:02:47 and accept additions to 17:02:52 But... But... Tiles are awesome. 17:03:07 they at least help us to indoctrinate new players 17:03:38 I guess not updating webtiles is Napkin's way of incentivizing the switch to console ;-P 17:04:00 dpeg: do you think 10% is good for dummy abyss_rune? 17:04:05 yes 17:04:23 what about the chance to spawn the rune? 17:04:33 at all, I mean 17:04:39 hm, no idea -- I've never been a garbage collector 17:04:44 not updating? O_o 17:04:55 Napkin: Guten Abend! 17:05:00 Nabend :) 17:05:57 !seen Thasero 17:05:58 Sorry dpeg, I haven't seen Thasero. 17:07:09 -!- st_ has quit [Ping timeout: 255 seconds] 17:07:18 kilobyte: reason for not always showing inventory weights: fewer numbers == good (unless you want the numbers); also space. 17:08:27 but we now would need to mention that Ctrl-W toggles weights, removing the spaces savings :p 17:08:50 kilobyte: the toggle should be mentioned in the help page, nowhere else 17:09:24 kilobyte: I think of new players. Them looking at the inventory should not convey the idea that weights are crucial information. I think the toggle is fine, especially if we could save state. 17:18:05 kilobyte: do you see what I mean? (I think you can turn on all weights with an option since forever) 17:19:19 * Napkin still thinks weight is crucial.. since especially new players pick up everything! 17:19:50 Napkin: it is >20 numbers per screen. This is really bad psychologically. 17:20:49 the interface of the patch is great. There's an option which you can toggle in-game. 17:21:04 they are so far on the right, i don't really see the impact 17:21:25 galehar: does it work on tiles, too? weights didn't show up there for a long time 17:22:09 The patch fixes that too 17:22:14 \o/ 17:22:20 well there's 2 patches actuallyy 17:22:31 one for the tiles fix, and one for ctr+w 17:22:49 Napkin: the human brain can only evaluate so much information at once. For example, the % screen is extremely dense which is fine -- a new player will not discover it right away. The inventory is used all the time. 17:22:50 the tiles fix is a bit hacky and I haven't tested it 17:23:09 dpeg: first game my GF played.. after 3 levels: so, how do I know what to drop? how can I see the weight? 17:23:19 dpeg: are you discussing what it should default to? 17:23:49 -!- Pingas has quit [Ping timeout: 240 seconds] 17:23:56 galehar: yes. I reason that default should be off. 17:24:28 Napkin: the help screen will explain that. I think it is already there, _ iirc. 17:25:07 well, we can agree to disagree ;) 17:25:14 yes, no problem 17:25:28 Napkin: there is a weight indicator in the top line, though 17:25:56 people don't go to the help screen for figuring out how to turn on weight-display 17:26:31 Napkin: it seems a natural place to answer your girlfriend's question!? 17:26:37 using the weight indicator to figure out weights is quite reverse logic ;) 17:27:12 Napkin: it is much fewer numbers, though 17:27:24 especially, because you are "burdened" at 75/100 - people don't understand that either ;) 17:27:44 yeah, well, true 17:28:39 you only really need to know a few basic facts, like wands being rather heavy, and the top line suffices for that 17:28:53 heck, I learned the lesson without any help at all =) 17:29:06 it's not intuitive enough 17:29:33 * SamB thinks the default should be on 17:29:44 i bet hiding this ctrl+w in the help will give you lots of comments like "ahh, i found the secret option to show weights!" 17:29:57 (that way, those who dislike it will look for the option 17:29:58 -!- Pingas has joined ##crawl-dev 17:30:03 -!- Pingas_ has joined ##crawl-dev 17:30:03 -!- Pingas_ has quit [Client Quit] 17:30:08 and those who like it, won't have to!) 17:30:15 nah, newbies don't think like that, SamB 17:30:22 * dpeg feels misunderstood in his quest for the newbie but it's been like this since 0.2. 17:30:24 and really don't expect to be able to hide it 17:30:29 newbies are the onse who want it though 17:30:42 SamB: no 17:31:04 dpeg: look, which of us is closer to newbie level? 17:32:09 Sigh 17:32:33 You can look up every item's weight in the item description, too. 17:32:43 tedious... 17:33:32 dpeg is right, that a complete newbie doesn't care about the weights 17:33:46 okay, probably so 17:33:50 Each number we display tells a new player "this is important information, you have to understand and use it". And we show tons of numbers. Weights are simply not that important. If you're burdened, you drop stuff until not burdened anymore. Later on, a player knows that weights are not as important as AC etc. Then the weights will not distract anymore. 17:33:57 but someone who managed to get burdened a few times will 17:34:04 okay, new idea 17:34:20 isn't there a patch to show your new burden status in the pickup/drop menus? 17:34:22 Napkin: we all start as pack rats. You have to learn dropping stuff anyway. No big deal, really. 17:34:28 alongside the show-weights patch? 17:34:29 Hint system tells users about option when they get burdened, or don't pick something up because they would if they did 17:34:53 assuming that goes in then it seems pretty reasonable to not bother showing weights by default 17:34:58 but only if they're in the tutorial or in hints mode 17:35:26 -!- ais523 has quit [Remote host closed the connection] 17:35:39 MarvinPA_: that would be much better than the curent A>B C stuff we currently have in the top line. 17:35:52 hmm 17:35:54 I dunno 17:36:06 right, i've not looked at the patch myself but it sounds good 17:36:37 I kinda want to get some idea *how long* a particular drop will keep me unburdened... 17:36:54 * dpeg suspects SamB is a number fetichist :) 17:37:37 well, I guess *with* show the weight toggle, yeah 17:37:40 that should be fine 17:38:19 it's not so much that I'm a number fetishist as that the words are so chunky 17:38:34 it's like a bad Atari 2600 game 17:39:06 * galehar galehar thinks show_weights should default to auto. Only shows weight when at 90% burden) 17:39:13 oooh 17:39:22 that's a really good idea 17:39:30 -!- stabwound has quit [Read error: Operation timed out] 17:40:24 galehar: interesting 17:40:45 SamB: the words are great. Reducing a 0...400.0 scale to three words ==> awesome! 17:41:12 dpeg: it's just I'd like to have some idea of how much wiggle room I've gained, you know? 17:41:24 if there was "almost burdened", that would work better 17:42:34 well you can have. It's when you're at 90% and weights appear. Status can display almost burden. 17:43:20 -!- stabwound has joined ##crawl-dev 17:44:14 it might also be good to have the inventory say *how much* you are over/under burdened when you're over 90% ? 17:44:34 -!- MarvinPA has joined ##crawl-dev 17:45:00 seems to be really overcomplicated 17:45:06 dunno 17:45:16 I mean, the whole auto-hiding thingy 17:45:27 the toggle is awesome in itself 17:45:30 even I'd use it 17:45:45 okay, fine 17:45:50 either the player wants to know basic info, or is a masochist... but sometimes showing it sometimes not would only cause confusion 17:45:52 there is the option, and we're spending ages on the ancient question of the default 17:46:04 lets just forbet about defaults and leave the default at whetever the user had hard-wired it to ;-) 17:46:09 kilobyte: I'm the masochist? :) 17:46:21 dpeg: 99% players use the default because they don't know about the other option 17:46:23 and make sure the hotkey is shown prominently somewhere 17:46:44 like, the hint mode or tutorial 17:46:58 SamB: they should learn about options in general 17:47:04 dpeg: you may already know item weights by heart... but new players don't 17:47:31 which, incidentally, might be a good idea to add to CAO/CDO, though I don't think that deaths should be announced in ##crawl and whatnot ;-) 17:47:32 kilobyte: galehar recommended easy_eat_chunks to me, which does not seem to be newbie material. So we could have an advanced.txt option set, doing it all at once for the players. 17:47:46 -!- MarvinPA_ has quit [Ping timeout: 256 seconds] 17:48:05 kilobyte: I know I am often alone in trying to suppress numbers, but I think I know what I talk about. 17:48:43 huh, that's an idea 17:48:58 suggest adding "include = advanced.txt" 17:49:11 then the user dislikes some of the settings, decides to investigate 17:52:49 -!- greatzebu has quit [Quit: greatzebu] 17:52:58 oh dear 17:53:12 it looks like I may have just broken dpkg :-( 17:53:16 well, not personalyl 17:54:16 * SamB hopes tar is not actually aborting, just returning a failure status... 17:54:36 -!- Pingas_ has joined ##crawl-dev 17:55:42 -!- Pingas has quit [Ping timeout: 256 seconds] 17:57:52 !messages 17:57:52 (1/1) dpeg said (1h 15m 42s ago): Thanks for holy work! 17:57:59 dpeg: welcome! 18:08:28 -!- dtsund has joined ##crawl-dev 18:11:36 -!- Brannock has quit [Read error: Connection reset by peer] 18:14:05 Well, at least I've realized that I don't have to muck around with ar to fix it... 18:14:21 ... I can just use emacs to extract the new tar binary ;-) 18:15:59 -!- Brannock has joined ##crawl-dev 18:24:39 due: do you have further plans with holies? 18:25:05 dpeg: kilobyte is doing some work on silver stars, I believe 18:25:06 dpeg: I need to do a review 18:25:10 cool 18:25:36 This is relevant for ziggurats. Need to update them only at the end :) 18:25:42 -!- ixtli has quit [Quit: Textual IRC Client: http://www.textualapp.com/] 18:26:35 due: we have very good vault makers by now (e.g. evilmike), tell me when we have enough guys to ask for a holy Pan lord. 18:27:01 dpeg: we can start about, let's say, right now 18:27:07 ok! 18:27:14 I see no reason not to start now 18:27:36 more holies can be better, but there's a bunch already 18:27:45 @??cherub 18:27:46 cherub (12A) | Speed: 10 | HD: 9 | Health: 63-92 | AC/EV: 10/20 | Damage: 15, 8 | Flags: 08holy, see invisible, fly | Res: 06magic(96), 05fire, 10elec, 03poison, 04rot, 13neg+++, 08holy | XP: 752. 18:27:49 done... 18:27:53 angel, daeva, also done. 18:28:01 I'll review them either today or tomorow 18:28:03 what would the threat of a holy level be? 18:28:11 late-game 18:28:12 which gap does the pan lord have to fill? 18:28:14 @??pearl dragon 18:28:14 pearl dragon (08D) | Speed: 12 | HD: 18 | Health: 97-139 | AC/EV: 10/15 | Damage: 3508(holy), 2008(holy) | Flags: 08holy, sense invisible, fly, !sil | Res: 06magic(168), 12drown, 04rot, 13neg+++, 08holy | XP: 5312 | Sp: holy breath (3d36). 18:28:43 due: I am talking about Pan, maybe the only one 18:28:53 -!- syllogism has quit [] 18:29:38 daeva / angel / cherub / pearl dragon / shedu / ophan / apis / silver star / [holy swine] / cherub name:seraph n_rpl n_spe col:lightmagenta hd:25 tile:mons_seraph ; blessed great sword ego:flaming 18:31:29 blessed toes need just the summon_holy spell adjusted and are ready to go too 18:31:32 @??blessed toe 18:31:32 blessed toe (08*) | 04UNFINISHED | Speed: 10 (move: 200%) | HD: 14 | Health: 77 | AC/EV: 50/1 | Flags: 08holy, see invisible, lev, !sil | Res: 13magic(immune), 02cold, 10elec++, 03poison, 04rot, 13neg+++, 08holy | XP: 4953 | Sp: holies, heal other, holy word. 18:32:30 AC 50? Heck, I want that the next time I stub it. 18:32:44 @??curse toe 18:32:45 curse toe (08z) | Speed: 12 (move: 200%) | HD: 14 | Health: 77 | AC/EV: 50/1 | Flags: 07undead, evil, see invisible, lev, !sil | Res: 13magic(immune), 05hellfire, 02cold, 10elec++, 03poison, 12drown, 04rot+++, 13neg+++, 13torm | Vul: 08holy++ | XP: 7890 | Sp: summon undead, mushrooms, torment symbol. 18:34:52 unfinished: phoenix, spirit, paladin 18:36:34 hrm, not so good... blessed toes would need something NOT top tier to summon 18:36:45 yes 18:37:05 and the weakest holies are around dire elephants 18:37:28 i still really dislike blessed toes 18:37:48 yeah... me too 18:37:48 Mu_: your toes are cursed? 18:38:07 100% copy of curse toe 18:38:09 -1 for blessed toes 18:38:49 * dpeg wonders if it was possible to add death cobs these days. 18:39:27 I wouldn't want a "life cob" monster at least 18:39:29 it's not that they're silly, it's that they only exist because curse toes do :P 18:40:03 we could brainstorm what a blessed toe should do 18:40:40 would be more interesting if they weren't just an exact parallel to curse toes, yeah (likewise silver stars and orbs of fire, sort of) 18:40:53 escort halflings on their way out? :p 18:41:53 that reminds me 18:42:07 silver stars are near-exact opposities to orb of fires... so let's not overdo that 18:42:16 how can we give the smallish guys a chance at Zot Defence? 18:42:20 irresistible orbs of fire :/ 18:42:53 -!- galehar has quit [Remote host closed the connection] 18:43:08 MarvinPA: I toned down the chance to 1/3 and made amulet of rMut work against them... still NASTY 18:43:22 well, the 3d40 bolt of holy is what worries me 18:43:37 ah, that too 18:43:40 @??orb of fire 18:43:40 orb of fire (05*) | Speed: 15 | HD: 30 | Health: 150 | AC/EV: 20/20 | Flags: 11non-living, see invisible, lev, !sil | Res: 13magic(immune), 05hellfire, 02cold, 11elec+++, 09poison+++, 12drown, 04rot+++, 13neg+++, 13torm | XP: 11437 | Sp: b.fire (3d40), polymorph other, fireball (3d43). 18:43:43 @??silver star 18:43:43 silver star (15*) | Speed: 15 | HD: 30 | Health: 150 | AC/EV: 20/20 | Flags: 08holy, see invisible, lev, !sil | Res: 13magic(immune), 05fire, 02cold, 11elec+++, 09poison+++, 04rot+++, 13neg+++, 08holy | XP: 11437 | Sp: holy light (3d40), silver blast (3d43). 18:44:09 -!- greatzebu has joined ##crawl-dev 18:44:19 bolt of fire is resistable, holy damage can be at most multiplied (without any way for the player to mitigate that!) 18:44:21 I'm not sure why orbs of fire have a holy counterpart... it isn't like orbs of fire are evil (unless you ask zin) 18:44:43 elliptic: the idea of a holy star is attractive 18:45:16 sure, but do we have to base it on orb of fire? it just doesn't make much sense to me 18:45:26 in addition to the general problems with being a copy 18:45:27 right now the only counterparts are silver stars and blessed toes... and the latter have only dpeg defending them (due didn't say a word yet) 18:45:50 yeah, they can copy the idea not exact stats 18:46:00 blessed toes should be good for the player 18:46:13 except if the player is undead 18:46:33 right, possibly the fact that they have the exact same speed/hd/ac/ev etc contributes to it bugging me :P 18:46:57 Could silver star have a huge corona? Or disappear in a sanctuary in order to protect someone, a Pan lord say? 18:46:58 they should actually be armour that prevents the player from tripping 18:47:16 dpeg: only 32x32 ;-P 18:48:20 dpeg: I made the bolt of Zin work mostly like usual Zin stuff, with regular evil being only quite weakly affected 18:48:50 good 18:49:31 ie: mutated, chaotic or transformed? You're screwed. Merely undead? Only some damage. 18:50:08 Good old Zin and his priorities :) 18:50:18 the holy light has regular holy properties, though 18:51:04 -!- Pingas_ has quit [Read error: Connection reset by peer] 18:51:09 Wasn't there the idea of a holy monster that sacrificies itself to mass heal holies? 18:51:23 paladin 18:52:11 ah 18:54:02 due: what about nerfing silver stars, especially the holy light? 18:55:20 note that silver blast currently does zilch if you lack a single mutation, are not transformed, don't worship a chaotic god and are alive 18:55:35 -!- edlothiol has quit [Ping timeout: 260 seconds] 19:02:02 elliptic: thank you! 19:03:12 I've asked elliptic and MarvinPA about removing the inconsistency between weapon enchantments and slaying rings. elliptic suggests to move slaying out of the product into the enchantment die roll. 19:03:28 That's a slaying nerf but not as big as players will expect (i.e. quite late). 19:04:31 -!- ussdefiant has quit [Ping timeout: 248 seconds] 19:05:05 would you also nerf randart attributes like Dam+3 the same way? 19:06:43 not sure that would be needed... the Dam+3 should go straight into the +dam enchantment instead 19:06:49 Is weapon damage currently something like base damage * skill_and_stat_multiplier + enchantment die roll? 19:06:56 dpeg: it's a buff! it makes EW scrolls as good as slaying :) 19:07:00 -!- rejuxst has joined ##crawl-dev 19:07:09 Eronarn: that is the proper way of judging it, yes! 19:07:19 btw, I'd also suggest that we make randarts unlikely to have crappy enchantments 19:07:35 they don't need to be over the top, but why come up with -5 stuff? 19:07:43 -!- elliptic has quit [Ping timeout: 248 seconds] 19:07:52 who are you and what did you do with the real dpeg 19:08:09 haha 19:09:14 dpeg: what do you mean that wouldn't be needed? i thought Dam+3 currently went towards base damage just like slaying, unlike +dam enchantment 19:09:25 (it is very likely I was mistaken!) 19:11:26 krel: the point is make those two work the same way 19:13:01 ah, I understand 19:13:18 in which case +dam on randarts would be a bit superfluous... it'd only affect auxes differently 19:14:21 Eronarn: you mean on randart *weapons*, right? 19:14:23 yes, hence my line 19:14:58 Is this enchantment die roll before multipliers like holy kick in? 19:15:33 I'm trying to get some sense of how big a slaying nerf it would be. 19:16:43 jle: After, I think? 19:16:47 ??damage 19:16:47 damage[1/2]: See {weapon damage}, {unarmed combat} 19:16:52 ??weapon_damage 19:16:53 weapon damage[1/3]: Approximately: 1d(base*strength_bonus + slaying)*skill_bonus + 1d(to_dam) + various bonuses. Weapon {brand} is applied after monster AC reduction. 19:17:07 Ah, before. 19:17:18 ??skill bonus 19:17:18 skill bonus[1/1]: Weapon skill bonus to damage is (25 + 1d(skill + 1))/25 or about 4% per point of skill. 19:17:55 At skill 27, this means slaying is about twice as powerful as enchantment. 19:18:15 Yes. 19:18:30 but early on, enchantment is stronger? 19:18:45 what are the numbers at skill 0? 19:18:50 No, early on they're about equal. 19:18:58 Exactly equal with skill 0. 19:19:10 as long as it isn't a nerf at skill 0, *I* don't mind ;-) 19:19:20 Okay, not *exactly* equal, but the expected values are the same. 19:19:42 SamB: at skill 0, 1d(base*strength_bonus + slaying) * 1 + 1d(to_dam), so exactly equal 19:19:43 Slaying has a higher variance. 19:20:08 Enchantment makes it more weighted toward some intermediate value. 19:20:19 Difference between, say, rolling a d8 and rolling 2d4 19:20:23 skill_bonus at skill 0 is 26/25 which is pretty close to 1. 19:20:34 Oh, skill+1, not skill, yeah. 19:20:43 Essentially the same. 19:20:49 I was kinda discounting that for these purposes 19:21:04 ??strength_bonus 19:21:04 weapon damage[2/3]: Strength bonus is usually < 1.5 (see {str}), skill bonus (both from weapon and Fighting skills) is randomized, averaging about 2.0 at max skill 19:21:10 Yes, I doubt 4 % will make much of a difference for anything. 19:21:10 err, assuming 1d is distributive... 19:21:38 krel: As far as expected value goes, it pretty much is, but the distributions will look different. 19:21:40 krel: it isn't, actually 19:21:56 neat! today I learned a new thing about math. 19:22:06 see, 1d isn't gaussian, so... 19:22:22 -!- elliptic has joined ##crawl-dev 19:22:31 -!- evilmike has joined ##crawl-dev 19:24:53 Regeneration and Death's Door as monster-castable spells (https://crawl.develz.org/mantis/view.php?id=4702) by 78291 19:25:22 <3 death's door nergalle 19:25:39 oh, zipcode strikes 19:25:45 thats cool 19:27:13 is it strange that the first google result for 78291 is http://crawl.chaosforge.org/index.php?title=78291 19:27:16 Death's Door on a monster sounds tedious 19:27:35 dtsund: why? Only if it is used too often. 19:27:39 "Welp, guess I can't hurt this thing, time to retreat and wait it out" 19:27:49 I think hexes should still work 19:27:57 Walk away, come back in 15 turns to magic dart it to death when it's at 13 hp? 19:28:14 hopefully you can't just walk away because it's smite-targeting summons on you 19:29:22 if I'm reading the patch right nergalle gets ddoor as an emergency spell 19:29:31 so it will be cast in a similar way to how khufu casts ToD 19:29:42 ToD can be trivially broken with dig 19:30:01 and still it is cool 19:30:20 Yeah, I'm fine with ToD for Khufu. 19:30:32 well, one of the things to note is how rare it is 19:30:38 I think I've seen khufu cast ToD maybe once 19:30:54 nergalle gets ddoor in place of teleport self. I've never seen nergalle teleport 19:31:28 I think monster ddoor will be more interesting and less annoying than monster fleeing and monster teleporting 19:31:44 yeah its way better than teleporting away 19:32:12 was teleport self in her escape slot? monsters seem not to use those spells very often at all. 19:32:18 evilmike: indicates that emergency is triggered not often enough 19:32:24 -!- rejuxst has quit [Quit: Page closed] 19:32:24 yes that 19:32:53 bears rarely berserk...have you ever seen ice fiends summon demons? gretell says they have summon demon. 19:33:14 my only real thought on the slaying change/nerf is that it hurts melee/ranged people in aftergame, and they're the ones who have the biggest problems with it already. 19:33:28 the only monster notorious for using its escape spell, to my knowledge, is asmodeus. 19:33:48 elliptic: will write traps pt 2 tomorrow... some interesting ideas from you guys. I thought you'd like that I address your concern =) (which is really an issue for me, just like labyrinth drawing was) 19:34:56 dpeg: your idea is certainly interesting, I just think that trying to design traps so that they are all just as interesting with no monsters around as with monsters around is pretty hard 19:35:34 and running into traps while traveling through cleared levels doesn't seem very flavorful to me 19:36:36 dpeg: you're probably right that it's not triggered enough. If it's triggered more often, I think emergency spells would need to be looked at though (in particular, a bunch of enemies should lose teleport self) 19:36:48 I'm thinking of asmodeus in particular, he can be annoying when he does that 19:36:59 mnoleg too 19:37:21 elliptic: absolutely agree it is hard... would've contemplated culling traps more if it weren't for shafts. Can't sacrifice shafts! 19:37:54 shafts and teletraps are the only types of trap I actually like currently I think 19:37:58 elliptic: ah yes, that is the part I want to resolve before sending pt 2. Cleared areas should be somehow safe of traps, new areas dangerous. I need to formalise that. 19:38:29 elliptic: Spider webs will be fine b/c there's always monsters nearby 19:38:45 dpeg: well, that was what my suggestion was for :) you might have a better idea of how to do it though 19:39:31 webs with spiders around would be reasonable of course 19:39:38 elliptic: but for new, random traps, I believe that using my proposal is good -- it is scum free 19:39:58 I agree that applying it to long-cleared areas is not interesting 19:40:20 no need to provoke action where there shouldn't be any -- this ties with redoing monster respawning 19:40:35 dpeg: well, I was serious about hating it that usually there are no monsters nearby when you run into a trap while exploring 19:40:51 I think this is a large problem independent of the problem with long-cleared areas 19:41:58 elliptic: I agree. But we can solve both problems at once. 19:42:24 Your reason is why I wanted to remove mechanical traps (outside of vaults). 19:42:33 it isn't just mechanical traps though 19:43:02 yes, but those are worst because they sometimes kill you 19:43:21 sure, I agree traps shouldn't kill people directly 19:47:54 -!- MarvinPA has quit [Ping timeout: 256 seconds] 19:50:02 -!- ussdefiant has joined ##crawl-dev 19:50:03 kilobyte: sounds fine 19:51:53 nights 19:51:55 -!- dpeg has quit [Quit: leaving] 19:52:26 night 19:56:39 -!- Twinge has quit [Ping timeout: 258 seconds] 19:56:51 early mechanical traps are useful, though ... 19:57:12 are you going to up the spawn rates of their respective ammo? 19:58:34 -!- Torokasi has quit [Ping timeout: 258 seconds] 20:01:38 -!- mikee_ has joined ##crawl-dev 20:04:53 samb: nobody uses them for that, really 20:05:14 not even really early? 20:05:26 what about, uh, me? 20:05:59 well, the thing is that you don't get arrow/bolt traps really early 20:06:05 true 20:06:05 and darts are really really common 20:06:18 not as common as dart traps 20:06:37 yeah, but you quickly get more darts than you will ever want 20:06:59 true, true 20:07:01 -!- evilmike has quit [Read error: Connection reset by peer] 20:07:22 -!- evilmike has joined ##crawl-dev 20:27:01 -!- greatzebu has quit [Remote host closed the connection] 20:27:09 -!- greatzebu has joined ##crawl-dev 20:43:35 -!- Twinge has joined ##crawl-dev 20:43:59 -!- MarvinPA has joined ##crawl-dev 21:08:00 -!- greatzebu_ has joined ##crawl-dev 21:09:00 -!- ainsophyao has joined ##crawl-dev 21:11:36 -!- greatzebu has quit [Ping timeout: 258 seconds] 21:11:44 -!- greatzebu_ is now known as greatzebu 21:17:23 -!- Wensley has joined ##crawl-dev 21:17:25 -!- Zaba has quit [Ping timeout: 240 seconds] 21:28:24 -!- Zaba has joined ##crawl-dev 21:31:33 -!- Twilight-1 has quit [Read error: Connection reset by peer] 21:31:48 -!- Twilight-1 has joined ##crawl-dev 21:32:06 -!- greatzebu has quit [Ping timeout: 258 seconds] 21:32:55 -!- valrus has joined ##crawl-dev 21:35:39 -!- greatzebu has joined ##crawl-dev 21:40:18 -!- greatzebu has quit [Ping timeout: 260 seconds] 22:06:04 -!- greatzebu has joined ##crawl-dev 22:24:39 -!- mong has quit [Quit: Page closed] 22:25:45 berserk able to activate while very hungry seems new (and not intended?) 22:26:42 almost a suicide button for berserkers without a warning message 22:29:09 -!- valrus has quit [Remote host closed the connection] 22:44:09 -!- jeanjacques_ has joined ##crawl-dev 22:45:26 -!- Mu_ has quit [Quit: Defecator, may everything turn out okay so that you can leave this place.] 22:46:18 -!- jeanjacques has quit [Ping timeout: 260 seconds] 22:47:05 -!- valrus has joined ##crawl-dev 23:04:23 -!- jeanjacques_ is now known as jeanjacques 23:18:08 -!- Brannock has quit [Ping timeout: 256 seconds] 23:42:21 -!- greatzebu has quit [Quit: greatzebu] 23:42:46 -!- greatzebu has joined ##crawl-dev 23:43:35 jeanjacques: the berserk thing looks like a definite bug--you can berserk even at near starving. i'm creating an issue and associated patch on mantis 23:46:18 Prevent berserking when very hungry or hungrier (https://crawl.develz.org/mantis/view.php?id=4703) by GreatZebu 23:58:55 -!- Pacra has quit [Quit: This computer has gone to sleep]