00:00:43 Stable (0.13) branch on crawl.s-z.org updated to: 0.13.0-13-g27bfa31 00:02:39 -!- Datul has quit [Ping timeout: 250 seconds] 00:05:05 -!- Tarquinn has joined ##crawl-dev 00:05:18 anyone got a sec to help me understand the guts of augmentation? 00:05:20 ??augmentation 00:05:21 augmentation[1/1]: Demonspawn mutation: gives scaling spell power and slaying bonuses with HP total: bonuses are equal to max(0, ceiling(mutlevel*(2*hp - maxhp)/maxhp))*(0.4*spell power + 4 slaying). 00:05:41 is there a typo at the end with '4 slaying'? I'm trying to rig up an excel sheet 00:05:41 -!- st_ has quit [] 00:05:52 Unstable branch on rl.heh.fi updated to: 0.14-a0-305-g056651f (34) 00:06:13 not sure if that means I need to multipy my current 'slaying' value by 4 00:06:27 johnstein: ((0.4*spell power) + 4) slaying, possibly? 00:06:40 maybe 00:06:45 * Grunt goes to dig into code... 00:06:47 Unstable branch on crawl.s-z.org updated to: 0.14-a0-305-g056651f (34) 00:06:48 but it enhances spell and slaying 00:06:59 yea. I was about to try that too. 00:08:13 but I'm not familiar enough with the code to search efficiently yet 00:08:33 Call the first chunk "augmentation_amount" (that's what it comes out as in the code). 00:09:04 ...then the spell power boost is 0.4 * augmentation_amount, and the slaying boost is 4 * augmentation_amount. 00:09:12 ah 00:09:14 That's what the code suggests to me. 00:10:07 yea. I see that now. in 'slaying_bonus' 00:11:06 Stable branch on crawl.develz.org updated to: 0.13.0-13-g27bfa31 (34) 00:11:31 -!- N78291 has quit [Quit: null] 00:13:02 -!- rossi has quit [Read error: Operation timed out] 00:14:49 -!- ystael has quit [Ping timeout: 272 seconds] 00:16:25 Unstable branch on crawl.develz.org updated to: 0.14-a0-305-g056651f (34) 00:20:55 -!- magistern has quit [Quit: This computer has gone to sleep] 00:29:10 -!- Quashie has quit [Ping timeout: 245 seconds] 00:35:24 I think if I'm following the code, looks like at mutlevel= 1, augmentation_amount() = 2 at full health, and 1 below that (to half health) 00:35:46 -!- Zifmia has quit [] 00:35:51 so at full health, I'm getting aug_amount * 4 = 2 *4 = +8 slaying 00:36:01 which seems pretty rad 00:36:41 Windows builds of master branch on crawl.develz.org updated to: 0.14-a0-305-g056651f 00:44:39 -!- Gamesmaster has quit [Quit: Page closed] 00:51:56 -!- Bloaxor has quit [Quit: Error: Bloax not found] 00:56:38 -!- Stossel has quit [Quit: Page closed] 01:01:20 -!- bh has quit [Quit: Page closed] 01:11:18 has there ever been a conversation regarding spell damage, to be specific about the crazy range of die values that exist? 01:11:25 ice dragon (16D) | Spd: 10 | HD: 12 | HP: 74-107 | AC/EV: 10/8 | Dam: 17, 17, 1707(trample) | cold-blooded, fly, !sil | Res: 06magic(48), 02cold++, 03poison, 12drown | Vul: 04fire | XP: 986 | Sp: cold blast (3d24) | Sz: Huge | Int: animal. 01:11:25 %??ice dragon 01:11:42 3d24? it can do 3 damage or 72 damage? 01:11:51 ??throw icicle 01:11:52 throw icicle[1/1]: A 4th level Ice/Conjuration spell. Throw Icicle casts a single-target projectile of high-velocity ice, doing mixed cold and physical damage. 40% of the damage can be mitigated by cold resistance. 3d20 at max power. In the Book of Frost. 01:12:03 same thing, it can do 3 damage or 60 damage? 01:12:11 melee can do 1 damage or about 60 damage 01:13:04 -!- kingbuddyboy has quit [Quit: Page closed] 01:13:16 doesnt that seem like quite a large range, given that monsters do fixed damage and you can either evade the damage altogether or AC reduce the damage which is, once again such a large range 01:14:02 wait what 01:14:18 an AC 60 character could possibly only mitigate 1 damage if ive read it right 01:14:24 yes 01:14:29 or 0 01:15:38 that doesn't happen that much, and it certainly doesn't happen much in combination with a high damage roll on the spell 01:15:42 so why the large range of values? a guy with 200 HP and 60 armor could be demolished in 5 turns by a boulder beetle if the RNG chooses so... 01:15:59 I should note that the way spell dice work, mostly they are less swingy than melee 01:16:10 3 dice tends towards the middle much more than 1 die 01:16:18 -!- iasov has quit [Remote host closed the connection] 01:16:56 plus the RNG is a cruel bitch, ive had many times where beginning a magic character with only 3 or 4 mp have had 2 miscasts (while at 6% or so) and then my sucessful casts only dealth minimal damage 01:17:45 which leaves a low level frail magic user with no mp and a gnoll or some other fucker just plucking away 01:17:46 nevermind that increasing the tendency to the center would have balance implications for AC 01:18:07 but thats besides my point of i would much ranther have 6d10 than 3d20 01:18:25 so again, why are you arguing this for spells but not melee 01:19:02 i will argue this for melee as well if that is the case 01:19:27 honestly i have not looked that close at melee damage calcs 01:20:57 anyway swinginess serves to keep you on your toes pretty well, right 01:21:02 when there are things like OOD monsters and packs of gnolls, jackals, and ugly things, the potential for a losing situation develops more when the chance of taking a turn to do such small damage makes things.... shitty for lack of better term 01:22:34 it can be extremely frustrating when youve got shock or freeze at max power, and are frying goblins and shit left and right, then all of a sudden you have to spend 10 freaking casts to eliminate one stupid gecko or something because the RNG wills it 01:22:41 So just to be clear you do actually realize that sometimes you're supposed to die in this game, no matter how hard people like me try to avoid it 01:22:44 -!- myrmidette has quit [Quit: Leaving.] 01:22:50 !won 01:22:51 Naruni has won once in 254 games (0.39%): 1xSpEn 01:22:57 yeah i know :) 01:23:37 for agrument's sake, lets look at what you were saying about melee 01:23:45 I mean... I'm trying to read what you're saying and it sounds something like "sometimes this game kills me partly because a roll doesn't go my way -> damage should be less random" 01:23:57 which feels like a weak argument to me so maybe something is getting lost here 01:25:16 when you get to a level that you can dish out 60 damage in a turn... is it realistic to expect to have to hack away at say some troll for 13 turns while the RNG fucks you over and over? with higher skills maybe the curve should narrow a bit? 01:25:34 sure it's realistic to expect it, it just won't happen very much 01:26:15 i can understand fighting 27 and maces 0 then you hit a guy with a mace to cause a wild range of damage, but 27 fighting and 27 maces you should be a pretty damn good judge of what damage you can do 01:26:43 turns out when I have a +7 great mace and 20 M&F/Fighting I do indeed kill ogres a lot faster than I do with, like, a +0 flail, 6 maces, and 3 fighting 01:27:49 since my accuracy is higher so I hit a lot more, and my damage range is higher so I do enough damage to 1shot or 2shot them much more often 01:28:02 (read: ever, that second case would not 2shot ogres very much) 01:28:57 I mean... it's not like combat is -actually- you and the enemy lining up politely and taking turns swinging at each other in an ordered fashion where you can line up your perfect blow, right 01:29:38 even if in a turn-based game it feels like this is the case 01:30:18 adjust that concept to magic though 01:30:25 gnoll (08g) | Spd: 10 | HD: 2 | HP: 8-18 | AC/EV: 2/9 | Dam: 9 | 10weapons, 10items, 10doors | Res: 06magic(8) | Chunks: 07contam | XP: 14 | Sz: Medium | Int: normal. 01:30:25 %??gnoll 01:30:31 ogre (07O) | Spd: 10 | HD: 5 | HP: 16-40 | AC/EV: 1/6 | Dam: 17 | 10items, 10doors | Res: 06magic(20) | Chunks: 07contam | XP: 119 | Sz: Large | Int: normal. 01:30:31 %??ogre 01:30:37 realism of magic is pretty hard imo 01:30:46 touche 01:30:50 but let's put it like this, magic is hard and getting magic to be perfect is -really- hard 01:31:26 lol 01:32:40 anyway the early game is generally the only place where you are going to have any real problems with randomness 01:32:50 so if getting magic to be perfect is -really- hard, should perfect magic (aka a max power throw icicle) be weak enough to cause only 3 damage? 01:32:59 you don't have perfect magic 01:33:12 I'm pretty sure you don't have 0% miscast on throw icicle 01:33:26 you might have it on throw frost and freeze if you work hard enough 01:33:39 you have to get to level 81 before magic is perfect 01:34:02 and even if you do get it to 0, surely you're not so far removed from the point where you would occasionally mess up even getting it to work 01:34:20 so having it occasionally fail to work very well maybe not surprising 01:36:00 anyway later in the game this is not really a problem at all if you're playing with it in mind, because your rolls are more favorable and you have some amount of access to defensive (and occasionally offensive) options that are 100% reliable in some ways 01:36:11 with a 30 int, 26 ice, 26 conj, and 26 spc, you have 100/100 power throw icicle 01:36:21 not unacheiveavble in real game 01:36:39 ok but what is your point, also 01:37:02 in what way do you think making damage dice tend more towards the mean will improve Crawl 01:37:15 with all of that investment players should not be submitted to only doing 3 damage! 01:37:26 cast it again and you'll probably do like 40 01:37:35 alternately use a better spell 01:37:58 SwissStopwatch> in what way do you think making damage dice tend more towards the mean will improve Crawl 01:38:02 let me think a moment 01:38:17 how about we make miscasts worse so it could just kill you now 01:38:25 -!- ussdefiant has quit [Ping timeout: 272 seconds] 01:38:33 already plausible if you miscast juuuuust so 01:38:53 also damage dice do tend very heavil towards the mean so 01:38:54 24% fail bolt spells not very safe 01:39:14 just simply miscasting as a early magic user can be death in itself from the opponent, regardless of the miscast effect 01:39:39 that's an entirely different thing, for one! 01:39:48 and usually you can walk away to get more MP to try again 01:41:01 Naruni: do you know how likely it is for 3d20 to come out 3? 01:41:10 (hint: not very) 01:42:08 fr: inner flame produces clouds in trees 01:42:24 granted, it's something like 3x as likely to be 4 01:42:34 well the 3x is also not very often 01:42:38 -!- _dd has joined ##crawl-dev 01:42:50 I don't know how many x it is for 5 01:43:07 -!- bmfx has quit [Ping timeout: 248 seconds] 01:44:20 for 5 it is 6/8000 I think, or 10/8000 for 3 to 5 01:44:39 ok here goes: making the dice tighter on the averages based on skills accomplishes a few things: AC to begin with, a character with 40 AC wont get one shotted by a lucky ogre hit, since the ogre wont have the skill necessary to find the 'chink' to cause a killing blow. likewise, a character in melee with a high AC monster would need high skill to cause more damage. 01:44:57 40 ac is not going to be oneshotted by an ogre are you crazy 01:45:10 how are you getting 40 AC with that little HP :P 01:45:27 and you already need higher skill to do more damage! 01:45:35 situations develop... this is what im talking about 01:45:38 including to high AC monsters 01:45:49 Dispater (06&) | Spd: 10 | HD: 16 | HP: 450 | AC/EV: 40/3 | Dam: 50 | 05demonic, 10weapons, 10items, 10doors, fighter, see invisible, !sil | Res: 13magic(immune), 05hellfire, 04fire+++, 02cold, 11elec+++, 03poison, 04rot, 13neg+++, 13torm | Vul: 08holy++ | XP: 9455 | Sp: greater demon, crystal spear (3d35), iron shot (3d29), hellfire (3d20) | Sz: Large | Int: high. 01:45:49 %??dispater 01:46:07 If I try to hit this dude with low-skilled +0 mace I'm not going to hurt him much 01:46:19 ogre (07O) | Spd: 10 | HD: 5 | HP: 16-40 | AC/EV: 1/6 | Dam: 17 | 10items, 10doors | Res: 06magic(20) | Chunks: 07contam | XP: 119 | Sz: Large | Int: normal. 01:46:19 %??ogre 01:46:33 ??giant spiked club 01:46:34 giant spiked club[1/2]: It looks *really* painful. Damage rating: 22 Accuracy rating: -7 Base attack delay: 18, (pierce) damage, never generated with a brand unless it's an artefact or god gift. Can only be wielded by trolls, giants, and ogres. The best twohander in the game. 01:46:39 and situations are supposed to develop I'm pretty sure 01:46:58 like do you actually object to the game trying to kill you sometimes 01:47:07 no not at all 01:47:09 sometimes even trying extra-hard to do so 01:47:23 then... what does "situations develop" actually mean 01:47:49 just the insane RNG occurances which cause death is a pain 01:48:06 I think you should find that there are not actually that many of those 01:48:06 if that is roguelike religion then i accept it, but it sucks :) 01:48:09 those basically aren't a thing 01:48:09 like really, for real 01:48:22 oh yeah? 01:48:22 and even when things are going against you if you play right you can avoid death 01:48:26 skilled play is death resistance 01:48:47 !lg . teae ckiller=ogre 01:48:47 everyone who gets killed by the first ogre they see says 01:48:47 1. Naruni the Cloud Mage (L11 TeAE), worshipper of Vehumet, demolished by an ogre (a +0,+0 giant club) on D:11 on 2013-10-16 23:40:48, with 8969 points after 14111 turns and 1:41:13. 01:48:52 "damn he got some lucky hits and" 01:48:59 !lg . teae ckiller=ogre -log 01:49:00 1. Naruni, XL11 TeAE, T:14111: http://crawl.akrasiac.org/rawdata/Naruni/morgue-Naruni-20131016-234048.txt 01:49:01 but he didn't 01:49:23 You could see Urug, Grum, 3 yaks, a hippogriff, a troll, an ogre and 4 wights. 01:49:39 ok first off that's a ridiculous situation and that happening without some highly questionable play is -incredibly- unlikely 01:49:45 and you were fighting in view of these guys why? 01:50:24 i had just walked down the stairs... which the other stairs were flooded by a pack of about 6 centaurs firing ice and flame arrows at me 01:50:29 -!- gnum has quit [Quit: Leaving] 01:51:01 sure, that's bad 01:51:14 now i do not argue, i could have played that better, but it really sucked 01:51:29 it's not likely that all those dudes will line up just so and instantly kill you going down the stairs 01:52:00 you can watch the tv if you like, i didnt have any options 01:52:15 either that staircase or the centaur stairs 01:52:36 i even commented in the game "arent there supposed to be 3 staircases?" 01:52:52 and there are, of course 01:52:57 enclosed areas. 01:53:14 anyways, we got off topic, i just wondered if anyone had ever thought about narrowing the ranges a bit 01:53:43 im not saying it must be done im just wondering if it is in people's heads 01:56:06 -!- Stelpa has quit [Remote host closed the connection] 01:56:13 -!- xFleury has quit [Quit: ChatZilla 0.9.90.1 [Firefox 24.0/20130910160258]] 01:57:51 all this came up because i see an ice dragon on lair:3 01:57:55 ice dragon (16D) | Spd: 10 | HD: 12 | HP: 74-107 | AC/EV: 10/8 | Dam: 17, 17, 1707(trample) | cold-blooded, fly, !sil | Res: 06magic(48), 02cold++, 03poison, 12drown | Vul: 04fire | XP: 986 | Sp: cold blast (3d24) | Sz: Huge | Int: animal. 01:57:55 %??ice dragon 02:02:34 and there it is... approx 30 max power shocks double bounced wouldnt take him out 02:03:23 well... by lair:3 you should have -something- else 02:03:25 isn't shock like Level 1? 02:03:30 ??shock 02:03:31 shock[1/1]: Level 1 conjuration/air spell which fires a weak bolt of lightning. It's still lightning though - prepare for {bolt bouncing} (and exploit it, if possible) 02:03:34 yeah, it is 02:03:35 also you can, you know, not kill it 02:03:41 why would you expect that to work? 02:04:09 because it shouldnt do 1 damage 02:04:15 every time 02:05:03 ontoclasm: yeah, i had plenty of sting and meph cloud, which the ice dragon resists 02:06:00 hm 02:06:46 seems only fair that a level 1 spell wouldn't be reliably effective against an ice dragon 02:06:47 lightning bolt? melee? 02:06:51 even without going into any mechanics 02:06:54 wands? 02:07:50 scroll of teleport or something like that? 02:08:23 ontoclasm: so a level 0 evoc wand should be more reliable/harmful than a conj 12 level 1 spell? 02:10:13 it just seems wild, if the dragon has 90 hit points, why wouldnt 18 casts of 5 point damage take it out? 02:11:08 Naruni: that's the whole "level 1" thing 02:11:18 also regen 02:11:24 regen i understand 02:12:25 is that the difference between a level 1 spell and a level 4 spell? the possibility of damage is higher? 02:13:23 the difference between a level 1 spell and a level 4 spell is that the level 4 spell does a lot more damage 02:13:23 why cant i rely on a well-trained level 1 spell (shock) which is dual school conj/air to tackle foes 02:13:31 -!- Snufkin has quit [Ping timeout: 250 seconds] 02:13:46 because we SAID so 02:14:03 and our forebears implemented things that way 02:14:48 -!- ystael has quit [Ping timeout: 240 seconds] 02:19:56 it's been such a long time since I played 02:21:21 -!- Quashie has quit [Ping timeout: 240 seconds] 02:35:20 -!- Kalir has quit [Quit: CADENZA-CLASS CLOCKWORK KNIGHT--OPERATION TERMINATED] 02:51:13 -!- Valarioth has quit [Ping timeout: 250 seconds] 02:58:54 -!- bhaak has quit [Ping timeout: 268 seconds] 02:59:11 -!- Pepe has quit [Ping timeout: 245 seconds] 03:00:16 -!- bhaak has joined ##crawl-dev 03:04:21 -!- VolteccerJack has quit [Ping timeout: 248 seconds] 03:07:09 -!- Crehl has quit [Quit: Leaving] 03:11:02 -!- ystael has quit [Ping timeout: 240 seconds] 03:12:42 -!- MarvinPA has joined ##crawl-dev 03:13:07 -!- eb has quit [] 03:17:03 -!- Cronoth has quit [Quit: ChatZilla 0.9.90.1 [Firefox 24.0/20130910160258]] 03:22:57 -!- flappity has quit [Read error: Connection reset by peer] 03:24:07 -!- Cronoth has quit [Quit: Leaving] 03:24:28 -!- magicpoints has quit [Ping timeout: 240 seconds] 03:24:43 -!- dtsund has quit [Ping timeout: 245 seconds] 03:25:42 -!- Ragdoll has joined ##crawl-dev 03:25:55 -!- qoala has quit [] 03:28:51 -!- nubcakes has quit [Quit: Page closed] 03:29:11 -!- Nstar has quit [Ping timeout: 245 seconds] 03:35:46 -!- Stelpa has quit [Remote host closed the connection] 03:38:38 -!- Stelpa has quit [Changing host] 03:42:33 -!- syllogism has joined ##crawl-dev 03:54:29 -!- simmarine has quit [Quit: Leaving] 03:56:29 -!- pythonsnake has quit [Ping timeout: 272 seconds] 04:05:37 -!- bmfx has quit [Ping timeout: 272 seconds] 04:14:04 -!- Helmschank has quit [Ping timeout: 240 seconds] 04:14:10 -!- MP2E has quit [Quit: leaving] 04:18:41 -!- crate has quit [Read error: Connection reset by peer] 04:20:20 -!- minqmay has quit [Quit: Leaving] 04:23:12 -!- jeffro has quit [Read error: Connection reset by peer] 04:23:38 -!- jeffro has joined ##crawl-dev 04:26:02 -!- alefury has joined ##crawl-dev 04:29:42 -!- fanaatikko has quit [Ping timeout: 256 seconds] 04:35:00 -!- Gotham has joined ##crawl-dev 04:37:05 -!- st_ has joined ##crawl-dev 04:38:28 -!- Somefellow has quit [Ping timeout: 240 seconds] 04:44:34 -!- elliptic has quit [Quit: Leaving] 04:54:52 -!- Roarke has quit [Quit: Page closed] 04:59:28 -!- pythonsnake has quit [Ping timeout: 265 seconds] 05:10:07 -!- orionstein_away is now known as orionstein 05:11:27 -!- pantaril has quit [Read error: Connection reset by peer] 05:22:26 -!- WalkerBoh has quit [Remote host closed the connection] 05:23:02 -!- orionstein is now known as orionstein_away 05:25:29 -!- blabber has joined ##crawl-dev 05:30:09 -!- bonghitz has quit [Remote host closed the connection] 05:30:28 -!- blabber has quit [Ping timeout: 264 seconds] 05:30:29 -!- blabber_ has joined ##crawl-dev 05:46:19 -!- Cronoth has quit [Quit: Leaving] 06:03:09 -!- raskol has quit [Ping timeout: 272 seconds] 06:18:10 DrKe (L13 VpCK) ASSERT(item.sub_type == CORPSE_BODY) in 'misc.cc' at line 145 failed. (Lair:6) 06:19:26 DrKe (L13 VpCK) ASSERT(item.sub_type == CORPSE_BODY) in 'misc.cc' at line 145 failed. (Lair:7) 06:23:45 -!- Blazinghand has quit [Ping timeout: 245 seconds] 06:23:58 -!- elliptic has joined ##crawl-dev 06:26:55 -!- blabber_ has quit [Quit: leaving] 06:28:11 -!- heteroy__ has joined ##crawl-dev 06:28:17 DrKe (L14 VpCK) ASSERT(item.sub_type == CORPSE_BODY) in 'misc.cc' at line 145 failed. (Lair:8) 06:29:09 -!- heteroy has quit [Ping timeout: 248 seconds] 06:29:20 -!- heteroy__ is now known as heteroy 06:39:25 -!- Hailley has quit [Quit: Petra is closed!] 06:50:38 -!- ackack has joined ##crawl-dev 06:54:01 -!- zyx has quit [Client Quit] 06:56:33 DrKe (L15 VpCK) ASSERT(item.sub_type == CORPSE_BODY) in 'misc.cc' at line 145 failed. (Orc:4) 07:00:01 DrKe (L15 VpCK) ASSERT(item.sub_type == CORPSE_BODY) in 'misc.cc' at line 145 failed. (Orc:4) 07:19:35 -!- Bloax has joined ##crawl-dev 07:21:34 -!- Ladykiller69 has joined ##crawl-dev 07:23:30 does x->v not show ghost movement speed? 07:23:42 because it REALLY should 07:28:39 -!- elliptic has quit [Ping timeout: 272 seconds] 07:29:17 -!- jeffro has quit [Ping timeout: 272 seconds] 07:32:49 -!- jeanjacques has joined ##crawl-dev 07:35:50 -!- pantaril has quit [Ping timeout: 245 seconds] 07:43:30 Contined lua lines in .des files cannot be parsed when using flex 2.5.4 on FreeBSD by bojan 08:05:31 -!- orionstein_away has quit [Ping timeout: 248 seconds] 08:05:36 -!- magicpoints has quit [Ping timeout: 256 seconds] 08:07:34 -!- Daekdroom has quit [Quit: Saindo] 08:08:11 -!- SamB has quit [Ping timeout: 248 seconds] 08:11:37 -!- atomicthumbs has quit [Ping timeout: 272 seconds] 08:11:57 -!- SamB has joined ##crawl-dev 08:18:52 Hydra does not come closer by sanka 08:18:57 -!- Quashie has quit [Ping timeout: 272 seconds] 08:33:22 -!- chlorine has quit [Ping timeout: 256 seconds] 08:37:43 -!- xFleury has joined ##crawl-dev 08:51:44 -!- yxhuvud has quit [Remote host closed the connection] 08:52:39 -!- orionstein_away is now known as orionstein 08:52:50 -!- _D_ has quit [Quit: Lost terminal] 08:55:05 -!- Comradin has quit [Remote host closed the connection] 08:57:15 -!- Fhqwhgads_ has quit [Ping timeout: 248 seconds] 08:57:26 -!- Fhqwhgads__ is now known as Fhqwhgads_ 09:00:11 |amethyst: I did an 11 hour game ( http://pastebin.com/ufunZAyn ) game with the MSVC build running in the debugger. And the only buggyness I encountered was some rare " 'it monster is not fully coded ". 09:05:11 do scythes already have cleave? 09:13:37 -!- blabber has joined ##crawl-dev 09:16:57 -!- paulsomebody has quit [Ping timeout: 272 seconds] 09:17:22 absolutego (L13 HaMo) (Lair:4) 09:20:51 -!- atomicthumbs has joined ##crawl-dev 09:21:25 -!- moxian has joined ##crawl-dev 09:31:04 !tell TZer0 clan webtiles died. 09:31:07 moxian: OK, I'll let tzer0 know. 09:31:09 wow! 09:31:09 TZer0: You have 1 message. Use !messages to read it. 09:31:11 okay 09:31:32 supposedly at that latest crash of absolutego 09:31:39 Webtiles server stopped. 09:31:44 Somebody (L27 GrFi) (Dis:3) 09:31:53 Webtiles server started. 09:33:06 damn it.. server being laggy 09:33:08 hold on 09:33:53 Webtiles server restarted. 09:33:58 !messages 09:34:00 (1/1) moxian said (2m 52s ago): clan webtiles died. 09:34:02 should be up now 09:34:11 thanks! 09:34:20 I hope this was satisfactory responce time xD 09:34:57 sure ^_^ 09:35:07 -!- HellTiger has quit [Ping timeout: 248 seconds] 09:38:36 -!- Sorbius has quit [Read error: Connection reset by peer] 09:39:09 -!- ketsa has quit [Remote host closed the connection] 09:47:27 -!- reaver has joined ##crawl-dev 10:01:40 -!- ackack has quit [Ping timeout: 245 seconds] 10:04:06 -!- duralumin has quit [Read error: Connection reset by peer] 10:06:51 -!- stuntaneous has quit [Ping timeout: 245 seconds] 10:07:41 -!- duralumin_ is now known as duralumin 10:14:05 -!- G-Flex has joined ##crawl-dev 10:15:48 -!- G-Flex has quit [Client Quit] 10:17:10 -!- G-Flex has joined ##crawl-dev 10:23:48 -!- Fhqwhgads_ has quit [Ping timeout: 240 seconds] 10:23:52 -!- Fhqwhgads__ is now known as Fhqwhgads_ 10:33:48 -!- reaver has quit [] 10:34:21 -!- Silurio has quit [Ping timeout: 245 seconds] 10:34:53 -!- eith has quit [Read error: Connection reset by peer] 10:35:06 -!- eith has joined ##crawl-dev 10:35:06 -!- valtern has quit [Read error: Connection reset by peer] 10:38:13 -!- Soadreqm has quit [Ping timeout: 248 seconds] 10:39:49 -!- Silurio1 has quit [Ping timeout: 248 seconds] 10:40:11 -!- eith has quit [Ping timeout: 272 seconds] 10:40:43 -!- Silurio has quit [Ping timeout: 248 seconds] 10:42:54 -!- ussdefiant has joined ##crawl-dev 10:44:56 -!- Silurio1 has quit [Ping timeout: 240 seconds] 10:45:48 -!- Soadreqm has quit [Ping timeout: 240 seconds] 10:46:03 -!- Silurio has quit [Ping timeout: 248 seconds] 10:47:01 -!- TacoSundae has quit [Ping timeout: 250 seconds] 10:49:40 -!- ShadeTornado has quit [Max SendQ exceeded] 10:56:51 -!- eith has joined ##crawl-dev 10:58:47 -!- Pacra has joined ##crawl-dev 11:03:42 -!- ainsophyao has joined ##crawl-dev 11:05:03 -!- duralumin has quit [Read error: Connection reset by peer] 11:05:10 Jesus is lantea being bitchy. 11:08:43 -!- Bcadren has quit [Read error: Connection reset by peer] 11:11:29 -!- blackcustard has quit [Ping timeout: 272 seconds] 11:12:28 -!- blabber has quit [Ping timeout: 264 seconds] 11:14:57 -!- LexAckson has joined ##crawl-dev 11:22:29 -!- ussdefiant has quit [Ping timeout: 248 seconds] 11:23:23 -!- blackcustard has joined ##crawl-dev 11:24:32 -!- duralumin has quit [Read error: Connection reset by peer] 11:25:31 ahahahahah 11:25:35 -!- orionstein is now known as orionstein_away 11:25:38 I got the Mark-status in E:3 11:25:40 best thing ever 11:25:46 everyone came running at me one by one :D 11:33:45 -!- Silurio1 has quit [Ping timeout: 245 seconds] 11:43:12 -!- Pacra has quit [Ping timeout: 256 seconds] 11:43:49 -!- rossi has quit [Ping timeout: 248 seconds] 11:46:16 -!- ackack has joined ##crawl-dev 11:50:43 -!- dead_ has quit [Ping timeout: 250 seconds] 11:51:06 -!- GuraKKa has quit [Read error: Connection reset by peer] 11:51:39 -!- Tux[Qyou] has quit [Read error: Connection reset by peer] 11:59:04 -!- dtsund has joined ##crawl-dev 12:02:15 -!- raskol has quit [Ping timeout: 272 seconds] 12:03:28 -!- VolteccerJack has quit [Ping timeout: 240 seconds] 12:05:47 03MarvinPA02 07* 0.14-a0-306-g0752fc5: Update manual from wiki 10(4 minutes ago, 1 file, 12+ 11-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=0752fc585101 12:05:47 Cherry-picked 1 commit into stone_soup-0.13 12:05:47 03MarvinPA02 07[stone_soup-0.13] * 0.13.0-15-g9204fa6: Don't mention Lava Orcs or Djinn in the manual 10(3 minutes ago, 1 file, 0+ 47-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=9204fa67a638 12:05:49 -!- dtsund has quit [Quit: dtsund] 12:05:49 -!- dtsund has joined ##crawl-dev 12:08:50 Stable (0.13) branch on crawl.akrasiac.org updated to: 0.13.0-13-g27bfa31 12:12:39 -!- alefury has quit [Read error: Connection reset by peer] 12:14:02 -!- blabber has joined ##crawl-dev 12:15:01 -!- blabber has quit [Client Quit] 12:15:19 -!- blabber has joined ##crawl-dev 12:17:15 -!- bh has joined ##crawl-dev 12:17:31 Can someone give me an idea of how tight the coupling is between Webtiles and Crawl? 12:17:39 -!- bh is now known as Guest45686 12:17:53 -!- Daekdroom has quit [Quit: Saindo] 12:17:54 -!- Guest45686 has quit [Changing host] 12:17:54 -!- Guest45686 has joined ##crawl-dev 12:18:09 -!- Guest45686 is now known as bh 12:20:38 -!- magistern has quit [Quit: This computer has gone to sleep] 12:21:48 -!- ShadeTornado has quit [Ping timeout: 240 seconds] 12:24:50 Unstable branch on crawl.akrasiac.org updated to: 0.14-a0-306-g0752fc5 (34) 12:25:41 bh: not very tight, it's mostly a bidirectional pipe. Crawl in webtiles mode produces json data as text, python-tornado stuff ferries it over websockets to the browser. 12:26:06 then there's javascript code which duplicates a good part of tiles UI 12:26:17 kilobyte: I'm weighing rolling my own webtiles or chopping them out of crawl 12:26:55 would that json be enough for you? 12:27:26 -!- evilmike has joined ##crawl-dev 12:28:16 kilobyte: what do you mean? 12:29:55 -!- mizu_no_oto has quit [Quit: Computer has gone to sleep.] 12:30:24 you could insert your interface in one of two places: 1. taking existing json data (without changing Crawl itself), or 2. implementing a whole new interface, basing on map_knowledge and stuff 12:30:50 -!- myrmidette has joined ##crawl-dev 12:30:52 is the prompt-for-swap option deprecated? https://gitorious.org/crawl/crawl/source/0752fc5851015ec43040d63dd2b004b4ab006032:crawl-ref/docs/options_guide.txt#L1071 12:31:34 kilobyte: I don't necessarily want to replace crawl webtiles. I'm trying to create generic webtiles for any roguelike 12:31:50 ah 12:31:59 I haven't looked deeply into Crawl's json 12:32:10 -!- dtsund has quit [Quit: dtsund] 12:32:16 you might want to spy on the webtiles socket and see what kind of data it sends 12:32:18 -!- ontoclasm has quit [Quit: Leaving.] 12:32:26 (and receives) 12:33:59 -!- Kalir has quit [Changing host] 12:36:10 bh: I think Brogue would benefit more from having the ability to watch other players play from with-in the game-client rather than having to use a browser. But maybe that's just me. 12:36:39 -!- eith has quit [Ping timeout: 240 seconds] 12:37:02 -!- Blazinghand has joined ##crawl-dev 12:39:13 03kilobyte02 07* 0.14-a0-307-g11d5d5e: Slightly simplify. 10(27 hours ago, 1 file, 49+ 47-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=11d5d5e26e59 12:39:13 03kilobyte02 07* 0.14-a0-308-g9384bbd: Separate monster_descriptor() into proper functions. 10(41 minutes ago, 9 files, 73+ 93-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=9384bbde4e7b 12:39:13 03kilobyte02 07* 0.14-a0-309-g6f9563e: Make monster fast regeneration a flag rather than a list. 10(28 minutes ago, 3 files, 21+ 44-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=6f9563e63611 12:39:13 03kilobyte02 07* 0.14-a0-310-gbd08a7f: Unify some duplicated monster regen code. 10(3 minutes ago, 4 files, 20+ 24-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=bd08a7f36084 12:41:00 xFleury: I can't even run brogue on my machine 12:41:16 -!- eith has joined ##crawl-dev 12:44:02 -!- SamB has quit [Ping timeout: 264 seconds] 12:45:13 myrmidette: probably although it's hard to tell since there's still a whole bunch of mostly-unnecessary butchery code 12:47:57 i vaguely remember something about a weird edge case if you wield a non-weapon and transform and then get prompted to rewield it afterwards, maybe technically that option still does something there 12:49:23 most useful option crawl has ever had 12:49:36 MarvinPA, when playing a fetm I kept getting asked whether I wanted to switch back my claws (from ice fists) 12:49:44 nice 12:49:51 can you report that on mantis? 12:54:21 so drop that bird before butchering, rather than tossing it in the air and around :p 12:54:26 Water engulfing and sticky flame. by dck 12:54:39 see, it turns out such butcher woes are pretty realistic 12:54:52 -!- SamB has joined ##crawl-dev 12:55:53 SamB: turns out gprof lies about its passive timings being mostly accurate; it seems to me they MASSIVELY overestimate the cost of function calls 12:55:59 SamB: am I doing something wrong? 12:56:32 -!- tkappleton has quit [Ping timeout: 265 seconds] 12:58:41 I guess that's quite hard to do if you do both active and passive profiling at once like gprof does 12:58:43 what do you mean by "the cost of function calls"? 12:59:01 -!- Voker57 has quit [] 12:59:04 you mean, the calls themselves? 12:59:08 -!- MaxFrost has quit [Ping timeout: 240 seconds] 12:59:35 obviously the active instrumentation does increase the cost of entry/exit to/from a function 12:59:39 as in, if you have a single big function doing CPU-costly work, it doesn't suffer the cost of gprof's instrumentation as much 13:00:51 is there a way to make the game prompt casting non-targeted spells? 13:00:58 -!- blabber has quit [Quit: leaving] 13:03:27 it is doable: if, when the interrupt comes, the running code is gprof's accounting, that sample should be disregarded 13:03:35 westexas (L12 LOFi) ASSERT(!feat_is_solid(grd(ctarget))) in 'cloud.cc' at line 567 failed. (D:6) 13:04:05 that would eliminate most of inaccuracies not related to CPU cache, etc 13:04:19 kilobyte: hmm, what if it's the prologue/epilogue code to call into gprof? 13:04:51 I suppose it would be possible to mark that somehow ... 13:05:13 not much harder to detect than doing it per-function: you list regions of the code that were added by gprof 13:06:00 and I guess gprof(1) wouldn't even need changing there, just the library code would need to know about it somehow ... 13:06:54 my fault is: I believed that if gprof claims to do passive profiling, it does it semi-accurately. 13:07:23 the data it provides seems pretty bad other for call graph/counts :( 13:09:32 is there a way to run gprof passively only? 13:10:13 -!- WalkerBoh has quit [Remote host closed the connection] 13:10:25 -!- etenil has quit [Client Quit] 13:14:19 -!- Crehl has quit [Ping timeout: 248 seconds] 13:20:57 TKT (L6 LONe) ASSERT(!feat_is_solid(grd(ctarget))) in 'cloud.cc' at line 567 failed. (D:5) 13:21:57 -!- Vizer_ has quit [Ping timeout: 248 seconds] 13:21:58 -!- blabber has joined ##crawl-dev 13:23:19 -!- ackack has quit [Ping timeout: 272 seconds] 13:24:01 -!- dtsund has joined ##crawl-dev 13:31:33 -!- Blazinghand has quit [Ping timeout: 248 seconds] 13:31:35 -!- tupper has quit [Client Quit] 13:34:59 -!- Tarquinn has quit [Ping timeout: 272 seconds] 13:41:19 -!- bonghitz has quit [Remote host closed the connection] 13:43:28 -!- rossi has quit [Ping timeout: 240 seconds] 14:00:19 -!- ontoclasm has joined ##crawl-dev 14:03:44 -!- absolutego has quit [Quit: leaving] 14:04:05 -!- Bcadren has quit [Ping timeout: 248 seconds] 14:04:52 -!- simmarine has joined ##crawl-dev 14:09:39 kilobyte: hmm, where does the profiling library itself live? 14:13:44 -!- ackack has joined ##crawl-dev 14:18:05 SamB: in gprof? no sure, but I assume it's a regular library that the instrumentation calls; most of the logic is in the "gprof" tool that analyzes the dump 14:20:25 we already got some code to produce a stack dump, this can be easily turned into a simple but accurate profiler 14:21:05 kilobyte, there are tools like oprofile and pmcstat 14:21:54 and valgrind --tool=callgrind 14:22:05 jilles: looking at that oprofile tool, I see it does mostly explicit events 14:22:25 callgrind, like gprof and about all profilers I see, does active profiling 14:23:13 I don't want a call graph as gprof already does that part well 14:23:32 oprofile uses the cpu's performance counters 14:24:47 taking a sample whenever the number of events of a certain kind (such as unhalted cycles) reaches some value 14:25:45 with callgrind, the active profiling should not interfere with the results as much, because it simulates various parts of the CPU (this also makes it very slow) 14:29:48 yeah, I was assuming that kilobyte didn't want callgrind's overhead (and/or the fictional cost model) 14:32:21 huh, surprisingly the things that have been said while I've been waiting for a chance to finish typing that (why is there never enough RAM?) did not make it look silly 14:33:13 -!- Arkaniad has quit [Read error: Operation timed out] 14:34:30 -!- ainsophyao has quit [Remote host closed the connection] 14:35:23 -!- UseBees has quit [Ping timeout: 248 seconds] 14:35:30 -!- Nightbeer has joined ##crawl-dev 14:35:41 SamB: overhead is ok, as long as it hits everything equally 14:36:09 SamB: we're talking about test runs on the order of 10 seconds... 14:37:33 -!- bh has quit [Quit: Page closed] 14:38:01 jilles: after a few mins of browsing oprofile's docs, I did not find yet what I want 14:38:30 and that's simple: the backtrace of parts that take the most user time 14:39:00 -!- Ladykiller69 has quit [Ping timeout: 265 seconds] 14:39:45 I'm not saying oprofile can't do that, just that its focus and thus the docs focus mostly on whole-system profiling that doesn't provide detail on the code that is running 14:44:03 kilobyte: callgrind/cachegrind use a (presumably simplistic) cost model for the target CPU, rather than anything related to the actual cost of executing the instrumented code 14:44:43 -!- Crehl_ has quit [Ping timeout: 240 seconds] 14:45:05 so, any innacurate bias against short functions would be due to that cost model, not the profiling overhead 14:45:29 (obviously cachegrind's is more sophisticated) 14:46:21 -!- fooobarrr has quit [Quit: Lost terminal] 14:48:25 I doubt foogrind would differ more than real-life architectures already differ between themselves 14:49:01 it would presumably bear more resemblance (at least in complexity) to older incarnations 14:50:04 about oprofile: http://packages.qa.debian.org/o/oprofile/news/20111228T001254Z.html 14:51:45 hmm, did that disappear? 14:52:12 * SamB reads now 14:53:33 that was pretty good RM bug, too 14:54:23 concrete where it matters most 14:56:40 -!- ystael has quit [Ping timeout: 245 seconds] 14:59:09 -!- Leafsnail has quit [Quit: ASCII a stupid question, get a stupid ANSI!] 14:59:30 - Alternatives exist (perf) 15:01:02 -!- sildraith has quit [Quit: Leaving] 15:01:12 yeah, but perf seems even less adequate here (according to the very little I know about it) 15:02:30 it at least seems likely to work with current kernels though 15:02:41 it can give you reports as detailed as "most of battery power was spent in this asm instruction here" 15:03:09 i am impressed with the technical knowledge of the crawl devs here =] 15:03:15 if oprofile doesn't work with current kernels, is there any point in shipping it? 15:03:31 s/with/by/ 15:03:38 what I'm looking for, though, is where the function was called from 15:03:58 -!- ontoclasm has quit [Read error: Connection reset by peer] 15:05:02 and not just as a count like gprof does, but: radius_iterator was guilty of 10% of CPU when called from builder_goofy from generate_level, and guilty of 1% of CPU when called from mon_behv 15:08:38 I *think* callgrind does that as long as recursion doesn't screw things up 15:09:20 I forget how accurate it is though 15:11:45 heh, look like backtrace_symbols_fd() called from SIGVTALRM would do just what I'm looking for, with some postprocessing 15:12:10 there might be better ways though... 15:12:21 surely there would be 15:20:30 well, it might not be IMPLEMENTED yet 15:23:55 -!- kingbuddyboy has quit [Ping timeout: 250 seconds] 15:25:40 -!- blabber has quit [Quit: leaving] 15:31:17 -!- Nightbeer is now known as Drahbeg 15:31:34 -!- Quashie has quit [Ping timeout: 256 seconds] 15:40:49 -!- Wolfram has quit [Ping timeout: 250 seconds] 15:41:23 -!- duralumin has quit [Quit: Leaving] 15:45:07 -!- emeraldemon has quit [Quit: emeraldemon] 15:47:55 -!- simmarine has quit [Ping timeout: 248 seconds] 15:52:21 -!- Ragdoll has quit [Quit: Houdoe] 15:54:03 -!- ystael has quit [Ping timeout: 272 seconds] 15:56:51 -!- sumguy has quit [Ping timeout: 250 seconds] 15:59:58 -!- simmarine has joined ##crawl-dev 16:00:17 okay, I'm trying to start making an Emacs mode for .des files again; what should I call it? crawl-map-mode? crawl-des-mode? crawl-vault-mode? 16:04:26 -!- shummie has quit [Quit: Page closed] 16:05:02 -!- xFleury_ has joined ##crawl-dev 16:05:09 -!- dtsund has quit [Quit: dtsund] 16:06:13 -!- nonethousand has quit [Ping timeout: 248 seconds] 16:06:59 -!- xFleury has quit [Ping timeout: 272 seconds] 16:07:12 -!- xFleury_ is now known as xFleury 16:10:23 -!- Stelpa has quit [Remote host closed the connection] 16:14:48 -!- ystael has quit [Ping timeout: 240 seconds] 16:16:41 -!- Kellhus has quit [Quit: Page closed] 16:17:11 -!- dtsund has joined ##crawl-dev 16:19:48 -!- nonethousand_ has quit [Ping timeout: 240 seconds] 16:20:01 -!- jeffro has joined ##crawl-dev 16:20:01 -!- jeffro has quit [Changing host] 16:20:01 -!- jeffro has joined ##crawl-dev 16:21:34 78291 (L16 GhAE) (Lair:3) 16:26:40 -!- valtern has quit [Read error: Connection reset by peer] 16:26:49 -!- Blazinghand has joined ##crawl-dev 16:26:49 -!- ShadeTornado has quit [Ping timeout: 240 seconds] 16:28:27 !lm * crash -log 16:28:28 5516. 78291, XL16 GhAE, T:52127 (milestone): http://dobrazupa.org/morgue/78291/crash-78291-20131019-212133.txt 16:30:10 78291 using tiles? That's treason! Or more likely, that's the result of attaching tile observers to a normal, faithful to Xom, console player. 16:31:10 hard to explain this one without the core file 16:37:09 -!- brian_ has quit [Ping timeout: 248 seconds] 16:37:09 -!- irctc034 has joined ##crawl-dev 16:38:01 -!- myrmidette1 has quit [Ping timeout: 272 seconds] 16:38:02 -!- G-Flex has quit [Quit: Time for us to bug out.] 16:39:07 -!- absolutego has joined ##crawl-dev 16:43:43 -!- G-Flex has joined ##crawl-dev 16:44:12 -!- irctc034 is now known as reaver_ 16:44:27 -!- reaver_ is now known as arcane_marksman 16:45:58 -!- SteampunkDuck has quit [Quit: Lost terminal] 16:54:20 -!- Zifmia has quit [Ping timeout: 265 seconds] 16:54:52 -!- myrmidette1 has quit [Ping timeout: 256 seconds] 16:54:55 -!- arcane_marksman is now known as reaver_ 16:59:11 -!- chlorine has quit [Read error: Operation timed out] 17:02:12 Enemy health bar display logic reversed by tmhedberg 17:05:27 -!- fanaatikko_ has quit [Ping timeout: 240 seconds] 17:06:00 -!- Morik has quit [Read error: Connection reset by peer] 17:07:09 hrm yeah... some folks go for the quality of code changes rather than quantity 17:07:47 -!- Soadreqm has quit [Ping timeout: 272 seconds] 17:09:08 -!- Zifmia has quit [] 17:09:35 -!- hurdos has quit [Ping timeout: 245 seconds] 17:10:03 03tmhedberg02 {kilobyte} 07* 0.14-a0-311-gd782aa2: Correct reversed logic for displaying enemy health bars 10(19 minutes ago, 1 file, 3+ 3-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=d782aa280379 17:10:05 -!- zrot has quit [Remote host closed the connection] 17:10:13 -!- gammafunk_ has joined ##crawl-dev 17:10:44 -!- scummos^ has quit [Quit: Konversation terminated!] 17:12:41 -!- lessens has quit [Ping timeout: 245 seconds] 17:13:55 -!- Tux[Qyou] has quit [Disconnected by services] 17:14:05 -!- Tuxedo[Qyou] is now known as Tux[Qyou] 17:16:27 -!- Crehl has quit [Ping timeout: 248 seconds] 17:16:57 kilobyte: what would you think of having an optional way to try and get a GDB stack trace when crashing? 17:17:00 -!- orionstein_away is now known as orionstein 17:20:42 "gdb -p $PID -ex bt -ex quit" or so should work nicely, with appropriate packages installed, and perhaps with crawl-gdb.py installed alongside the binaries (with a corresponding name) 17:21:00 -!- Stossel has quit [Quit: Page closed] 17:21:21 at that point crawl's state might be deep in the loo, so it would need to be done carefully, but that's doable 17:22:17 what would you want to get? Just the backtrace? If so, what's that python script for? 17:22:48 it might help print the backtrace 17:23:49 -!- _dd has quit [Remote host closed the connection] 17:23:50 then again sometimes it might make the backtrace less useful 17:24:08 basically, for printing argument values 17:25:15 Program terminated with signal 11, Segmentation fault. 17:25:15 #0 cell_is_solid (c=..., c@entry=) at terrain.cc:318 17:26:13 sounds like it's either a compiler bug, or a bug in GDB's interpretation of DWARF expressions 17:26:37 my idiocy in a sed job: 17:26:37 bool cell_is_solid(const coord_def &c) 17:26:38 { return cell_is_solid(c); 17:26:38 } 17:26:49 heh 17:26:50 but gdb should handle that 17:26:59 -!- TacoSundae has quit [Ping timeout: 250 seconds] 17:27:12 -!- ystael has quit [Ping timeout: 265 seconds] 17:27:26 -!- reaver_ has quit [Quit: Page closed] 17:28:03 mm, wasn't someone playing with a dwarf compressor? I could imagine that screwing up this case because nobody taught it to watch out for this case and not point something at itself 17:28:08 there's no loop anywhere in the debugger, including pretty-printers 17:28:29 geekosaur: reverted 17:29:12 -!- Ur-Quan has quit [Read error: Connection reset by peer] 17:29:13 -!- lessens has quit [] 17:29:35 -!- orionstein is now known as orionstein_away 17:29:59 geekosaur: kilobyte informed me that it broke building of Windows executables (presumably cross-building since why else would dwz be installed?), and that it was rather slow (which I sort of knew already) 17:30:48 being slow is okay, taking triple the time of a hot ccache build is not 17:30:55 in any case I suspect this loop is encoded in the debug symbol information in the binary 17:31:07 not using dwz if the target doesn't support it is easy 17:31:24 that, if you're not using something like dwz, strikes me as a compiler or linker bug 17:33:39 -!- myrmidette1 has quit [Quit: Leaving.] 17:34:18 SamB: does dwz reasonably improve things on optimized builds? 17:36:29 ??shadow creatures 17:36:30 shadow creatures[1/1]: L5 summoning, Book of Summonings and Dreams. When you cast it, summons one always-friendly level appropriate monster (some will come with an appropriate pack). When monsters cast it, summons up to 8. Good to use in places where normal monsters are nasty, like Zot. 17:36:39 don't actually know; it's likely that there are ways to make things better using just GCC flags anyway ... 17:38:49 -!- athros has quit [Quit: athros] 17:44:04 -!- nubcake has quit [Quit: Page closed] 17:45:28 -!- ZChris13 has joined ##crawl-dev 17:47:34 -!- Drahbeg has quit [Quit: Verlassend] 18:03:52 -!- Kalir has quit [Quit: CADENZA-CLASS CLOCKWORK KNIGHT--OPERATION TERMINATED] 18:04:41 03kilobyte02 07* 0.14-a0-312-gb8a6f58: A yet another rock worm + cloud fix. 10(53 minutes ago, 2 files, 9+ 3-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=b8a6f581a755 18:05:00 03kilobyte02 07* 0.14-a0-313-g982c3f3: Extinguish sticky flame when a water elemental engulfs someone. 10(5 minutes ago, 5 files, 14+ 2-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=982c3f3fd301 18:05:34 -!- Aponym has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/] 18:05:35 -!- Tarquinn has joined ##crawl-dev 18:05:59 I really do need to see if I can build local crawl with a macports gcc and optimization. APPLE_GCC and no -O can be pretty slow... 18:14:01 ...? 18:14:21 for playing (as opposed to running stress tests), I'd say a raspberry pi is fast enough 18:14:49 obviously, this assumes you don't put a compiler anywhere close to it 18:15:12 only 168 minutes for an -O2 build 18:15:20 I don't mean build time 18:15:55 like, I've had it take over a minute to start a new game (granted, on a somewhat loaded system) 18:16:08 ah, des cache build 18:16:19 * SamB has one on permanent loan, but no way to install an OS image for it 18:16:25 geekosaur: was this a G3? 18:16:42 nope 18:17:11 2007 iMac, 3.33GHz Core 2 Duo, 16GB RAM 18:17:18 -!- moxian has quit [Quit: Leaving] 18:17:23 not the world's fastest machine but not a particularly slow one either 18:18:00 it's just that some things got much slower when optimization was turned off for APPLE_GCC builds 18:18:18 <|amethyst> let me see about a core dump of 78291's game 18:23:59 <|amethyst> http://dobrazupa.org/saves/crash-78291-20131019.core 18:24:17 <|amethyst> I imagine you also need a binary? 18:24:43 probably just a backtrace 18:24:48 -!- HellTiger has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/] 18:25:08 I don't know that code really, it's just segfaults being typically trivial to fix 18:25:11 <|amethyst> it's in _etc_waves but the variables are all 18:27:20 Lair:3 doesn't have a heightmap, right 18:27:54 <|amethyst> http://s-z.org/neil/tmp/78291-bt.txt is the bt full 18:28:45 can you assign an elemental colour to something via the rc file? 18:30:03 <|amethyst> yes, but I don't see "waves" there 18:31:37 <|amethyst> kilobyte: tile ordering issue I think 18:31:47 <|amethyst> else if ((orig == TILE_SHOALS_SHALLOW_WATER 18:31:47 <|amethyst> || orig == TILE_SHOALS_DEEP_WATER) 18:31:47 <|amethyst> && element_colour(ETC_WAVES, 0, gc) == LIGHTCYAN) 18:32:39 if the colour can be used by the player somehow, it shouldn't crash 18:33:07 http://pastebin.com/hDnnE7i8 18:33:10 Well isn't this cute. 18:33:24 <|amethyst> sure enough *bg is 2057 == TILE_SHOALS_SHALLOW_WATER 18:33:28 -!- JaGGedTK has quit [] 18:33:33 in this case, the tile is TILE_SHOALS_SHALLOW_WATER or TILE_SHOALS_DEEP_WATER, both of which can be assigned only in actual Shoals 18:33:57 where are the arguments for why spiny worms should corrode your equipment just because 18:34:27 because, uhm... 18:34:28 just because 18:34:37 well equipment corrosion is horrible 18:34:47 almost as bad as consumable destruction 18:35:09 and here comes the one enemy that's so special that bites you and corrodes your equipment 18:36:07 One thing is slime and zot. 18:36:14 Where weird things like that are supposed to happen. 18:36:26 BUT SPINY WORMS - they are special! 18:36:39 They too should have some fun bugging the shit out of you! 18:37:02 instead of like iunno 18:37:11 doing the same thing as mimics? 18:37:18 you get entangled in the worm's spines! 18:37:23 holy shit help 18:37:43 -!- Blazinghand has quit [Ping timeout: 272 seconds] 18:38:19 you mean that great Lair:8 vault? It's other pleasantries include luring every rock worm one by one into open so you can kill them. 18:40:54 Oh no, spiny worms also spawn in vaults. 18:45:30 -!- syllogism has quit [] 18:49:28 -!- sbluen has quit [Quit: Leaving] 18:52:37 -!- xFleury has quit [Ping timeout: 248 seconds] 18:53:10 -!- PolkaDot has quit [Quit: Lost terminal] 18:55:06 -!- halv has quit [Quit: Leaving] 18:55:40 -!- absolutego has quit [Quit: leaving] 18:57:15 -!- Tux[Qyou] has quit [Quit: Excess flood] 18:58:45 -!- NeremWorld has quit [Ping timeout: 245 seconds] 19:01:02 -!- xFleury has joined ##crawl-dev 19:08:56 -!- dtsund has quit [Quit: dtsund] 19:08:57 -!- evilmike has quit [] 19:10:12 -!- Xenobreeder has quit [Quit: Page closed] 19:14:00 -!- hhkb has quit [Ping timeout: 252 seconds] 19:16:44 -!- mizu_no_oto has quit [Quit: ["Textual IRC Client: www.textualapp.com"]] 19:16:48 -!- hhkb has joined ##crawl-dev 19:17:25 -!- wheals has joined ##crawl-dev 19:17:32 -!- reaver has joined ##crawl-dev 19:19:21 -!- scummos^ has quit [Read error: Operation timed out] 19:28:06 -!- Daekdroom has quit [Quit: Saindo] 19:28:26 -!- paulsomebody has quit [Ping timeout: 256 seconds] 19:29:35 -!- ystael has quit [Ping timeout: 245 seconds] 19:37:24 -!- mikee_ has joined ##crawl-dev 19:39:03 -!- kingbuddyboy has quit [Quit: Page closed] 19:56:20 Wrong monsters have fast regeneration by tmhedberg 19:58:35 kilobyte: ^ 19:59:57 -!- Bloaxor has joined ##crawl-dev 20:00:05 Oh, kilobyte's not on this channel. That patch should probably be pushed before the serves rebuild, though. 20:00:44 Or he is, except he's on the top like all channel admins. Oops. 20:01:54 <|amethyst> got it 20:02:08 03tmhedberg02 {|amethyst} 07* 0.14-a0-314-g3e3f616: Correct inverted fast regeneration check 10(19 minutes ago, 1 file, 1+ 1-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=3e3f6160ba6f 20:03:33 -!- Bloax has quit [Ping timeout: 240 seconds] 20:05:09 -!- ystael has quit [Ping timeout: 248 seconds] 20:05:52 -!- LexAckson has quit [Quit: Leaving] 20:11:40 -!- nubcakes has quit [Quit: Page closed] 20:13:05 -!- fooobarrr has quit [Quit: leaving] 20:17:14 -!- dtsund has joined ##crawl-dev 20:18:56 * SamB thought we just had +v, not even half-ops 20:19:26 -!- st_ has quit [Ping timeout: 256 seconds] 20:19:59 yeh, only g*lehar is +o 20:20:51 and rax, but she's not here 20:21:01 I meant currently in channel 20:25:58 -!- dtsund has quit [Ping timeout: 245 seconds] 20:27:44 -!- Zermako has quit [] 20:29:13 -!- WalkerBoh has quit [Remote host closed the connection] 20:30:19 -!- elliptic has joined ##crawl-dev 20:31:01 hmm, what does it mean to pass -O0 to gcc for linking? 20:31:56 -!- Valarioth has quit [Quit: Page closed] 20:33:38 some linkers support optimization (I don't know if gnu ld or gold do) 20:34:05 what that means exactly depends on the platform 20:36:38 but for example a linker may be able to notice that a platform can use a short jump instead of a long one for a call across source files, which the compiler can't optimize unless it's a whole-program compiler (neither gcc nor clang is) 20:37:48 probably that is usually the assembler's job within a compilation unit 20:37:55 at least, on the GCC side 20:38:40 nope 20:38:53 assembler, like compiler, operates at the level of individual source files 20:39:04 I side "within a compilation unit" 20:39:27 right, but I specified "across source files". i.e. not in the same compilation unit 20:40:01 WTF, two reversed flag bugs in a single commit 20:40:02 yeah I'm nitpicking about how it's done in the local case 20:40:08 lookz like I'm trying hard... 20:40:25 anyway such things are more interesting on RISC platforms so more or less irrelevant for practical purposes 20:41:27 geekosaur: actually, after they fixed the compile-twice bug (they claimed it was "by design" for long...), LTO can be not by much slower than regular builds 20:42:19 are we back to believing I was talking about compile time earlier? 20:42:31 I'm sorry I mentioned anything 20:42:38 kilobyte: in what version did they fix that? 20:43:05 does this mean libpython.a will no longer have useless LTO objects? 20:43:17 er, sections 20:43:19 or is this something else? 20:43:25 4.7, I think. And double compilation is still the default, unless you pass -fno-fat-lto-objects 20:43:25 oh 20:43:37 kilobyte: darn 20:44:06 and there doesn't seem to be a sane way to strip LTO out of an object file 20:44:22 .o files get both GIMPLE and useless actual code, then during the link it gets compiled again 20:44:22 No keyword 'files' 20:44:28 different but related issue 20:45:02 SamB: it shouldn't be much slower to just compile twice with both flags, I think 20:46:53 so gcc+gld has its own link time optimization on top of the RISC thing where the linker effectively peephole optimizes across compilation units 20:47:20 you mean, even without -flto? 20:47:37 that is what I was describing earlier, yes 20:47:38 not lto 20:47:45 not *this* lto at least 20:48:37 <|amethyst> --relax you mean? 20:48:39 on some (mostly risc) platforms there are peephole optimizations that the linker can apply across modules. I think Itanium was also intended to do this 20:48:43 kilobyte: you want to patch the Python package to do that? 20:49:06 geekosaur: I don't get how you can peephole like that 20:49:13 it's more or less irrelevant in an all Intel world so it's mostly forgotten 20:49:19 how do you make the output smaller than the input? 20:50:27 also I thought intel had shorter and longer ways to code some jumps, too 20:50:37 (effectively) 20:51:05 so, for example, one of the RISC processors (MIPS?) had 3-byte short address jumps that saved a CPU clock, if other things were relative addressed (as they typically are) the linker can substitute that instruction for a cross-module call and move the saved instruction byte to the end of the module segment 20:51:27 hmm, oh, I guess that's conditionals which aren't usable for calls ... 20:52:07 -!- dtsund has joined ##crawl-dev 20:52:14 geekosaur: ah, so it requires an invariant that can't be expected to hold on Intel 20:52:14 (or insert a noop in place for an unconditional, but I think there was no such thing on mips? delayed instruction slots) 20:52:17 yes 20:52:28 also intel lacks the 30byte form and 2 byte is rarely useful across modules 20:52:33 and 1 byte almost never is 20:52:39 *3-byte 20:52:56 come to think of it I don't think Intel does 2-byte, just 1-byte and full size 20:53:04 -!- jeanjacques has quit [Quit: quit] 20:53:30 anyway these kind of optmizations generally only apply to RISC so nobody bothers much any more 20:53:55 geekosaur: anyway I thought MIPS had at least 2-byte alignment for instructions? 20:54:27 well, use full LTO and this problem becomes meaningless 20:54:36 as everything is a single compilation unit 20:55:10 MIPS may not have been the arch I'm thinking of; if it's 2 byte aligned then a 3 byte address seems unlikely 20:55:27 the 3 byte relative address was somewhat weird even for the RISC world 20:55:39 and I don;t recall which one had it 20:56:13 -j9 -O2: 444.25user 29.77system 1:06.32elapsed, same+LTO: 494.13user 38.67system 1:34.61elapsed 20:56:24 gcc-4.8 still spends too much time in single-threaded parts 20:56:58 geekosaur: does gcc use that 3 byte thing within a compilation unit? 20:57:26 don't recall, wouldn't surprise me if it knew about it and used it 20:57:47 but it's specifically link-time peephole optimizations that I recall encountering it in 20:58:23 the search for speedups on RISC processors got seriously strange at times 20:59:17 also I have no idea if gnu ld ever bothered with it. vendor linkers did 20:59:49 pretty sure it does 21:00:10 well, maybe not that exact thing but it at least does do things that involve patching instructions 21:00:24 which "gnu ld"? bfd? gold? 21:00:40 doh, no gold on mips 21:01:04 right, this also predates gold --- RISC has been dead for a while now, practically 21:01:52 those Loongson guys are making noise about reviving mips 21:02:09 although with those hundreds of new instructions that's hardly RISC anymore 21:03:09 actual RISC architectures stopped meaning what the acronym means a long time ago too :) 21:03:49 -!- fungee has quit [Ping timeout: 248 seconds] 21:06:11 gcc-snapshot (-j6 -O2 on a slower machine): 812.18user 18.19system 2:25.87elapsed, LTO 883.44user 21.36system 2:58.74elapsed 21:06:40 so the penalty for LTO is about half of what it was in 4.8 21:06:54 ugh, no, the number of cores is smaller here 21:10:42 oh, I think mips might have gotten away with 3 byte by making part of the address part of the "opcode" space. although I don't know if they actually did that I do recollect that being one trick used to get around alignment issues. (if first byte of opcode was in a certain range, second could be treated as data. I think even Intel's instruction decode has been known to do this... in the 8080 days >.> ) 21:11:13 (and, for 8080, at the bit level) 21:11:41 these days, they instead stuff parts of opcodes into operand space ;-P 21:12:00 -!- eeviac has joined ##crawl-dev 21:13:58 re-checked, gcc-snapshot's penalty for LTO is 33% compared to 22% on the same machine (by real time clock) 21:17:24 -!- Gotham has quit [Read error: Connection reset by peer] 21:21:26 -!- Froggeryz has quit [] 21:27:30 According to the Wiki, 0.13 was likely going to feature a "Lava Orc" and the removal of the priest background. 21:27:47 xFleury: LO didn't make the cut but pr is gone 21:28:06 Yah, pr got canned, but that got me curious about the fate of Lava Orc. 21:28:11 lava orcs will hopefully be in 0.14, though! 21:28:25 i think there's been some balancing of them in trunk 21:28:39 they always have the same speed now, i think 21:28:42 -!- Gamesmaster has quit [Quit: Page closed] 21:28:56 LO is just banned from non-alpha builds at the moment 21:29:11 wheals: yes, same speed. 21:29:21 or possibly just in 0.13 21:29:23 there were some more complicated balance changes that i made that needed refactoring, and haven't gotten in; because of that, some other balance changes like speed were made 21:29:25 not sure exactly 21:29:43 but anyway they have not been removed from the code 21:29:43 also there are still complaints about tension having problems 21:30:15 there is a better formula out there, but we also really need better data on tension 21:31:23 How did the "heat" mechanic work? Was it like, when you got hit for damage, your "heat" raised? 21:31:49 how about you try playing one in trunk and find out? 21:32:32 they aren't some relic of ancient crawl history, after all ;) 21:33:24 well, the speed changes are, of course, gone 21:41:56 -!- jeffro has quit [Ping timeout: 240 seconds] 21:43:49 -!- rossi_ has quit [Ping timeout: 240 seconds] 21:44:16 does jumping make any noise? 21:45:36 doesn't seem to 21:56:04 -!- eeviac has quit [Quit: ChatZilla 0.9.90.1 [Firefox 24.0/20130910160258]] 21:57:08 -!- Rebthor has quit [Ping timeout: 240 seconds] 21:57:47 I've got a strange bug where I press "G>" but the game can't find the downstairs. I have two staircases excluded, one up and one down, but it should know to go down the third. Should I file a bug report or let someone spectate it first? 21:58:06 The third staircase is the only one I haven't seen on the lower level. 21:59:02 -!- oberstein has quit [Ping timeout: 264 seconds] 22:00:29 sounds vaguely familiar 22:01:36 is the third staircase in the same connected region as you? 22:02:14 also, can you get to the third staircase with X> 22:02:15 ? 22:02:16 elliptic: Yes, I'm standing in the same room. 22:02:34 Also, can get it with X> 22:02:57 Using X> to get it does not fix the bug. 22:03:05 it's probably actually a bug then... I've noticed weirdnesses with G> previously I think, but I almost never use it because X> tends to be better 22:03:12 I'm on CSZO. 22:03:43 -!- qoala has joined ##crawl-dev 22:03:44 (X> is better because you can rest before going downstairs, can cast buffs, eat, whatever) 22:04:21 That's a good point. I've had some trouble forgetting to read ID my scrolls before moving on. 22:04:55 So I guess I'll file a bug report. 22:05:57 PS somebody should resolve #7650 the patch has been pushed to Trunk. 22:05:59 from what I remember of the code, G> actually behaves as "travel to the level below" if you've been to that level before, and interlevel travel doesn't use unknown stairs 22:06:15 whereas if you haven't been to the level before, it will use any stairs 22:06:43 -!- oberstein has joined ##crawl-dev 22:07:52 how come G< is normally okay going up untried stairs? 22:08:15 elliptic: That explain the bug. 22:08:18 would 22:08:48 -!- ZebTM has quit [Ping timeout: 256 seconds] 22:09:28 SamB: it is? the above was just from my memory, but G> or G< using untried stairs in any situation where there are nonexcluded already used stairs sounds like a bug to me 22:10:01 and here I've been complaining whenever G< *won't* go to the untried stairs 22:10:13 the question is just what should happen when all the already used stairs are excluded 22:10:30 SamB: do you mean X well, not actually complaining to anybody but grumbling about it 22:10:37 elliptic: no 22:10:42 why would you want this? 22:11:09 -!- TAS-2012v has quit [Ping timeout: 272 seconds] 22:11:26 untried stairs are dangerous, a shortcut for "go up a level" shouldn't do dangerous things :P 22:11:39 if I knew where I was going on the level above I wouldn't use G< 22:12:01 if I was trying to flee something I wouldn't be able to use G anyway 22:12:26 I don't understand, unless you were shafted past a level you should always know where you are going on the level above 22:12:42 or at least know a little bit about what the level above is like 22:12:55 um, I actually just wanted to try those stairs to know which other staircase they went with 22:13:15 right, how does G< help with trying all three staircases? that is what X<<< is for 22:13:36 I try to try them soon after encountering them 22:13:44 oh, I see 22:15:01 anyway I haven't actually looked at the relevant code in a year or something, I just remember it behaving weirdly somehow :P 22:15:29 I have tried to look; it did not really help much :-( 22:15:31 -!- Lawman0 has quit [Ping timeout: 260 seconds] 22:16:02 Oh that's weird. If I unexclude the staircase excluded from above, and then press "G>", it will go down the unexplored staircase. 22:16:04 well, what I was trying to look at was for G<, or actually I guess I tried to look at all the ways you can go downstairs ... 22:16:40 <|amethyst> I know there's something different between, on the one hand "Travel to this level one level away", and on the other hand both "Travel to this specific spot" and "Travel to this level >1 level away" 22:16:45 reaver: okay that is definitely weird 22:16:48 -!- MaxFrost has quit [Ping timeout: 240 seconds] 22:17:05 |amethyst: right, that's what I was trying to remember earlier I guess 22:19:25 I guess G> and G< just using the nearest unexcluded stairs regardless of used status wouldn't be crazy and would make SamB happy 22:20:07 I'd use G> even less than I do currently, but it would be simpler and G< would be useful for what SamB described 22:20:15 it seems like sometimes the upstairs get marked in some way as going to the floor above and sometimes they don't 22:21:08 I don't know that I want G> to do that; it would probably be better if it tried any unexcluded, known staircases first ... 22:21:31 it seems weird for G> and G< to behave differently though 22:21:48 there was some idea in the past of marking staircases as tried if you've seen both ends of them 22:21:53 without actually needing to go up them 22:22:11 I forget whether there was something wrong with that idea or whether it just never got implemented 22:23:04 I think the latter 22:23:28 or possibly ontoclasm didn't want to draw all those pairs of stairs ;-P 22:23:56 G> fails to go down staircase. by Reaver 22:23:59 SamB: well, it shouldn't involve new tiles 22:24:09 was a joke 22:25:30 idea being: how would you know which staircase was which if the three pairs of staircases don't each have distinctive appearances 22:25:54 <|amethyst> hm 22:25:57 You can see up them. 22:26:06 Just look at them and see the level above. 22:26:16 so no spirals then? 22:26:28 !send grunt_deathspiral SamB 22:26:28 The tiles aren't spiral. 22:26:31 Sending SamB to grunt_deathspiral. 22:26:38 okay then 22:26:42 that's good enough for me 22:27:03 oh, except how in the world could they really be straight? 22:28:07 SamB: Why couldn't they be straight? 22:28:08 the power of zot has warped space in the dungeon around it to make the staircases match up 22:28:25 when you leave with the orb, the dungeon falls apart 22:28:50 The power of Zot is invoked against the dungeon! Space warps horribly around the dungeon! 22:28:53 -!- pelotron has quit [Ping timeout: 272 seconds] 22:31:18 As you leave the dungeon, it begins to crumble beneath your feet. Not being a centaur or a spriggan, you are unable to escape the cave-in and die. GAME OVER 22:32:29 Random Stone of Tremor effects would spice up the Orb Run a bit though, particular if the disincentivized corridors. 22:34:17 -!- Somefellow has joined ##crawl-dev 22:39:53 -!- Taraiph has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] 22:44:13 -!- ussdefiant has joined ##crawl-dev 22:47:44 Stable (0.13) branch on crawl.lantea.net updated to: 0.13.0-15-g9204fa6 22:49:42 -!- reaver has quit [] 22:56:35 -!- N78291 has quit [Quit: null] 22:57:15 -!- pelotr0n has quit [Ping timeout: 248 seconds] 22:59:36 -!- myrmidette1 has quit [Read error: Connection reset by peer] 22:59:54 -!- myrmidette1 has quit [Client Quit] 23:02:54 -!- mizu_no_oto has quit [Quit: Computer has gone to sleep.] 23:06:20 -!- jeffro has joined ##crawl-dev 23:06:21 -!- jeffro has quit [Changing host] 23:06:21 -!- jeffro has joined ##crawl-dev 23:06:29 -!- bh has joined ##crawl-dev 23:06:42 If you spectate Jazzimus on cszo, you'll get a good glimpse of a bullshit death 23:07:13 He went down the stairs to Slime:3 with full HP 23:07:52 Immediately paralyzed and beaten to death 23:10:56 he knew all that stuff was there and then decided to go back downstairs into it? 23:11:20 that seems questionable 23:12:56 I don't think so? 23:13:09 He took the stairs and got hit by an eye and that was the end of it 23:13:23 he took the stairs and then went back up and then took the stairs again 23:13:37 oh. In that case, he had it coming 23:14:02 if that had been the first visit to slime:3 the monsters would not have acted immediately 23:14:12 giving him time to go back upstairs 23:14:22 This is why I suggest reviewing assertions about stupid deaths carefully; usually they're less stupid than they first sound like. 23:14:35 (Well, less "RNG fuck you" and more "stupid tactical decision".) 23:14:47 Grunt: his death is almost book good 23:15:05 the current slime death is a random teleport next to a wall, plus an eye :) 23:15:20 Sounds like quite the acid trip. 23:15:31 -!- Fhqwhgads_ has quit [Remote host closed the connection] 23:16:15 groan. 23:16:27 1learn add grunt groan. 23:16:36 <|amethyst> the eye wasn't in view the first time he went down 23:16:53 <|amethyst> the other stuff was, though 23:17:50 -!- magistern has quit [Quit: This computer has gone to sleep] 23:18:17 <|amethyst> kilobyte: #7648 can be closed, right? 23:19:22 Wop the Executioner (L13 LOBe) ASSERT(!feat_is_solid(grd(ctarget))) in 'cloud.cc' at line 567 failed on turn 17133. (Lair:5) 23:19:31 probably the main mistake there is "being in slime at all" anyway, followed by the stairdipping to let monsters get free moves 23:19:36 -!- raskol has quit [Quit: Konversation terminated!] 23:20:59 -!- orionstein_away is now known as orionstein 23:21:51 -!- Kalir has quit [Changing host] 23:22:20 -!- jeffro has quit [Read error: Connection reset by peer] 23:24:39 -!- wheals has quit [Ping timeout: 248 seconds] 23:29:42 Gargoyle description in the manual still mentions them having rC+ by nicolae 23:29:57 -!- Somefellow has quit [Ping timeout: 272 seconds] 23:30:35 mmm they have so much stuff that i forgot all the stuff that they have 23:30:51 MarvinPA: should we give them rHellfire? 23:31:06 we should remove rElec 23:31:16 elliptic: definitely 23:31:50 -!- Somefellow has joined ##crawl-dev 23:31:58 I mean, I get the flavor for it, but... they have a lot of mutations 23:32:14 -!- Blazinghand has joined ##crawl-dev 23:32:19 plus rElec is typically a hard to find resistance 23:32:36 -!- orionstein is now known as orionstein_away 23:33:00 hhehehh 23:33:01 * Grunt remembers having vague notions of introducing an insulation cloak ego at some point, not that he necessarily thinks this is a good idea at the moment. 23:33:11 also typically not a very useful resistance for things other than being less annoyed by eels 23:33:34 rElec on gargoyles is god damn amazing. 23:33:57 -!- arcline has quit [Quit: Lost terminal] 23:34:00 it's better on gargoyles than on the average character probably because they have low HP and high AC 23:34:19 It completely negates the need to depend on a very limited resource that can be popped by anything with frost darts. 23:34:36 Or on complete luck with finding an artifact with rElec. 23:34:50 <|amethyst> huh? 23:35:04 <|amethyst> frost darts cancel resistance? 23:35:09 no, no 23:35:10 potions of resistance 23:35:11 |amethyst: he's thinking of potions. 23:35:14 frost darts can break potions 23:35:16 the usual source of rElec 23:35:17 -!- ystael has quit [Ping timeout: 248 seconds] 23:35:22 <|amethyst> oh, that kind of popped 23:35:31 item destruction. meh. 23:35:50 Once the tourney is over with, I'm going to get back to rejiggering the abyss monster set. 23:36:05 I like dpeg's idea of picking a branch and populating some of the abyss spawns from it 23:36:17 <|amethyst> FR: remove all item destruction, including the destruction of scrolls when you read them and potions when you drink them 23:36:33 <|amethyst> as compensation, only one of each scroll/potion generate 23:36:33 <|amethyst> s 23:36:48 |amethyst: you've described what could be a very interesting game that is not crawl 23:36:48 fr: remove all scrolls, potions, and food. 23:37:00 sounds like desktop dungeons 23:37:13 Grunt: then remove Spriggans 'cause of slow digestion 23:37:31 Remove all species that use food. 23:37:40 mummydjinns 23:37:46 then remove Djinn 23:37:52 mummyscrawl! 23:37:53 <_< 23:37:55 dungeon crawl 23:38:14 for the times when you feel like making a mummy crawl across the dungeon 23:38:20 buddysprint could be awesome -- sprint where you have a unique as an ally 23:38:35 03MarvinPA02 07* 0.14-a0-315-gce4801b: Update Gargoyle manual entry (nicolae) 10(3 minutes ago, 1 file, 7+ 7-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=ce4801bb51d7 23:38:36 03MarvinPA02 07[stone_soup-0.13] * 0.13.0-16-g31db371: Update Gargoyle manual entry (nicolae) 10(3 minutes ago, 1 file, 4+ 4-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=31db37174b0b 23:38:37 in search of eternal rest 23:38:38 Do you get to pick the unique? 23:38:54 i suppose the text version of the manual is still out of date in 0.13, but i have no clue how to build it from reST 23:39:05 Grunt: how about... Mennas? 23:39:15 MarvinPA: I thought there was a rule 23:39:17 well, still out of date in every version, not just 0.13 23:39:54 fr: Mennasprint 23:40:05 <|amethyst> MarvinPA: that 0.14 change, did you make it online? 23:40:17 yeah 23:40:37 on the wiki that is 23:41:23 <|amethyst> as for the text version, make has a target for ../docs/crawl_manual.txt 23:41:36 <|amethyst> or you can use util/unrest.pl manually 23:41:42 apparently i just racked in a huge amount of tournament points 23:41:45 i guess that's pretty cool 23:41:52 MarvinPA: "make docs" 23:42:02 -!- Blazinghand has quit [Quit: Leaving] 23:42:03 or, well, use your usual flags too for best results 23:42:05 <|amethyst> oh, even better 23:42:44 actually is that even in the repository 23:43:28 <|amethyst> crawl_manual.txt? Yes 23:43:29 oh i guess it's not 23:43:32 MarvinPA: looks like "all" already depends on that so nothing is needed 23:43:44 |amethyst: it's in gitignore 23:44:02 <|amethyst> oh, I see 23:44:10 <|amethyst> %git ffa7a02e 23:44:10 07kilobyte02 * 0.9-a1-770-gffa7a02: Build the manual from reST source. 10(2 years, 4 months ago, 6 files, 3327+ 3256-) 13http://s-z.org/neil/git/?p=crawl.git;a=commitdiff;h=ffa7a02ef0a3 23:44:36 Unstable branch on crawl.lantea.net updated to: 0.14-a0-314-g3e3f616 (34) 23:45:33 -!- Blazinghand has joined ##crawl-dev 23:46:55 -!- Silent_Samurai has joined ##crawl-dev 23:48:05 -!- Somefellow has quit [Ping timeout: 265 seconds] 23:50:45 -!- xFleury has quit [Ping timeout: 248 seconds] 23:56:33 -!- Pisano has quit [Ping timeout: 272 seconds] 23:57:20 -!- eith has quit [Ping timeout: 240 seconds] 23:59:48 -!- ZebTM has quit [Ping timeout: 240 seconds]