00:03:16 -!- ainsophyao has quit [Remote host closed the connection] 00:04:47 -!- elly has quit [Ping timeout: 248 seconds] 00:06:48 -!- elly has joined ##crawl-dev 00:14:52 -!- hoody_ has quit [Remote host closed the connection] 00:19:56 -!- hoody_ has joined ##crawl-dev 00:40:23 -!- hoody_ has quit [Remote host closed the connection] 00:45:56 -!- dtsund has quit [Quit: dtsund] 00:49:49 -!- Fail_Bot has quit [Ping timeout: 252 seconds] 00:56:32 -!- valrus has quit [Remote host closed the connection] 01:04:53 -!- greatzebu has quit [Quit: greatzebu] 01:57:28 -!- Textmode has quit [Quit: Ex-Chat] 02:12:47 -!- ixtli has quit [Quit: Computer has gone to sleep.] 02:17:47 -!- galehar has joined ##crawl-dev 02:21:41 hi 02:21:48 -!- evilmike has quit [Quit: something happened] 02:34:58 -!- dpeg has joined ##crawl-dev 02:35:12 kilobyte: great commits, especially reaping! 02:37:59 -!- Adeon has quit [Ping timeout: 258 seconds] 02:42:21 -!- MarvinPA has quit [Read error: No route to host] 02:42:41 -!- MarvinPA has joined ##crawl-dev 02:45:22 -!- Adeon has joined ##crawl-dev 02:46:53 Unstable branch on crawl.akrasiac.org updated to: 0.10-a0-692-g40c6b91 (32) 02:53:15 -!- dpeg has quit [Quit: leaving] 02:56:55 -!- valrus has joined ##crawl-dev 03:01:23 -!- valrus has quit [Ping timeout: 260 seconds] 03:08:05 -!- ainsophyao has joined ##crawl-dev 03:08:29 -!- Euph0riaX has quit [Quit: Recharging Brain] 03:23:28 here is my idea for variable abyss speed. It starts at 0 and tops out at 100. Every 20 turns, it has 1 chance in 2 to increase, and 1 chance in 5 to decrease. So on average, you reach max speed after 6666 turns spent there. 03:23:56 If you leave the abyss, the speed is saved and when you're banished again, it is divided by 2 03:25:30 first, morphing is scaled to it. Then tie monster generation to it too. Could also have a chance to put some malign gateways at high speed 03:38:20 -!- st_ has quit [Ping timeout: 276 seconds] 03:44:28 Unstable branch on crawl.develz.org updated to: 0.10-a0-692-g40c6b91 (32) 03:44:36 you rarely ever get to choose whether to get banished or not, so I wouldn't go overboard with scaling monster generation too much... but morphing sounds like a good idea 03:44:54 it's morphin time 03:49:03 kilobyte: ok. Some scaling would be good I think, but you're right about being careful not to go overboard 03:49:22 we can increase exit vaults frequency too (if they are guarded) 03:50:07 kilobyte: I remember you talked about a nasty abyss bug. Saving and reloading causing a shift. I can't reproduce that. Did I misunderstood? 03:58:27 hmm most exit vaults have no monster or very few. 04:09:23 do you have any pending edits to abyss.cc? 04:21:09 -!- ainsophyao has quit [Remote host closed the connection] 04:28:56 nothing significant, why? 04:29:53 !seen cript7c 04:29:53 Sorry galehar, I haven't seen cript7c. 04:31:13 !seen Cryp71c 04:31:13 I last saw Cryp71c at Mon Sep 12 15:09:07 2011 UTC (18h 22m 6s ago) parting ##crawl-dev with message chanpart. 04:31:52 is he still working on fighting rewrite? Been a while since we heard about it. 04:33:16 kilobyte: do you seriously consider laying the foundations of translation for 0.10? 04:35:47 at least foundation, yeah... but I dread merge conflicts, as mprf() and friends were overhauled in the fight rewrite branch 04:36:28 and you're right that it would have to be done soon 04:36:56 that "18h ago" sounds interesting though... it used to be "X weeks ago" 04:37:36 -!- Fyren has quit [*.net *.split] 04:37:49 !tell Cryp71c are there any chances you could do at least a partial merge of combat soon? 04:37:49 kilobyte: OK, I'll let Cryp71c know. 04:38:58 -!- Fyren has joined ##crawl-dev 04:39:08 for abyss.cc, I'm reluctant to #ifdef changes anymore, it's not like we're going back to the old Abyss, so I pondered cleaning up the old version so no one tries to mess with it 04:43:23 kilobyte: good idea (cleaning up abyss.cc) 04:43:50 the code is still there in the repository if needed 04:56:17 -!- MarvinPA_ has joined ##crawl-dev 04:57:24 -!- valrus has joined ##crawl-dev 04:59:22 -!- MarvinPA has quit [Ping timeout: 260 seconds] 05:01:44 -!- valrus has quit [Ping timeout: 258 seconds] 05:02:43 03kilobyte * rc090c7372eeb 10/crawl-ref/source/ (abyss.cc tag-version.h tags.cc): Save the state of the Worley noise Abyss. 05:18:07 03kilobyte * rbae41ef08151 10/crawl-ref/source/acquire.cc: On non-divine acquirement, pay far more heed to bardings. 05:18:08 I've updated the planning 05:18:10 https://crawl.develz.org/wiki/doku.php?id=dcss:planning:planning 05:18:45 please, put your name in front of features you're planning to implement. And add them if what you're planning to do isn't listed. 05:30:02 -!- elliptic has joined ##crawl-dev 05:32:49 elliptic: I thought about your comment on the fact that the abyss exit isn't generated by the morphing 05:33:11 even if it were, it would still be optimal to go in the same direction to see more terrain 05:34:13 but I think it's better to prevent morphing exits and altars for theme. Those things shouldn't pop out of nowhere like any random feature. 05:34:40 probably reasonable, yes... I don't know much about the new abyss :) 05:35:33 -!- Fail_Bot has joined ##crawl-dev 05:37:02 In practice, I imagine turning will still prove common, given monsters being threatening as they should 05:37:55 Aside, I do not see what is wrong with making it optimal to explore the abyss. 05:38:28 -!- Fail_Bot has left ##crawl-dev 05:38:49 unless I misinterpreted something somewhere and responded to a false point 05:46:37 monqy: well, my only concern is abyss becoming more spoily... how are people supposed to know that they have 0 chance of finding an exit if they go in a circle, even though the entire abyss seems to have changed each time they go around the circle? but I don't really understand how the new abyss works myself and I don't have much experience in it, so this may not be an issue 05:48:20 ah 05:55:45 actually going in circles in the abyss might not be so easy 05:56:06 elliptic: each turn, there's a pass that goes over the whole map and each cell has a chance (which depends on abyss speed) to be turned into another feature. 05:57:51 the feature is chosen with the worley algorithm which I don't understand at all. It's an algorithm used for procedural generation. 05:59:54 sadly, there's little Worley left, it's mostly some bad randomness again 06:04:15 you mean since I've smoothen out the morphing? 06:08:18 hmm... what exactly did you do? 06:08:19 Worley noise is supposed to hold the whole information, you can recreate the whole grid without prior knowledge 06:08:19 using the old values for anything but at most tile flavours means Worley is not used 06:08:19 instead of morphing the whole map every 20 turns, I changed it so it morphs every turn, but only a few cells. 06:08:19 uhm... that's the problem 06:08:42 so instead of a Worley layout it's pretty much just the old Abyss with some random changes over time 06:09:39 but the changes aren't random. They use the worley algorithm to choose to features. 06:10:54 Worley is noise, so it's hashed randomness where the hash can't repeat 06:11:08 -!- ivan has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 06:11:08 -!- djinni has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 06:11:46 it's similar to Perlin... do you know that? It's far more widely known. 06:12:08 -!- Pingas has joined ##crawl-dev 06:12:19 -!- djinni has joined ##crawl-dev 06:12:19 -!- ivan has joined ##crawl-dev 06:13:02 I know the basics, but I won't say I understand the algorithm 06:16:50 your change did fix the flipping between feature types, but pretty much reverted to the old Abyss. I wondered where the layout went... 06:16:51 do you know when bmh returns? 06:16:55 he was around last week I think, but he's very busy 06:16:55 he's moving 06:16:55 my change was 84410e7c2 06:16:55 -!- Ashenzari has quit [Ping timeout: 260 seconds] 06:17:47 -!- Ashenzari has joined ##crawl-dev 06:17:47 -!- The topic of ##crawl-dev is: DCSS Development | Logs: http://tozt.net/crawl | People with +v have commit access. | Please keep general crawl-related chit-chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Long stuff to pastebin.ca please | Immortal Warwalrus and Crazy Yiuf forever. 06:19:08 abyssal_state is constant between each morph. It's only updated when the abyss shifts. 06:20:20 anyway, if you have an idea on how to get back the worley layout while keeping the smooth morphing, go ahead. You seem to understand this better than I do. 06:20:38 heh, after undoing that commit, the layout is back... but so is flipping 06:21:20 -!- ais523 has joined ##crawl-dev 06:22:39 bmh wanted a fourth dimension of the noise, to determine the features 06:27:57 I see... the feature comes from the lowest bit of the noise, ie, is not persisent at all 06:27:58 the core of my change is just to call morphing every turn but not on every cell: if (!one_chance_in(you.abyss_speed) && morph) continue; 06:27:58 uhm... you're supposed to increase the rate of going through the third dimension, if you want a higher speed 06:27:58 otherwise it totally falls apart into randomness again 06:27:58 I just morph more cells when speed increase... 06:27:58 you can experiment speed change with &^S 06:27:58 also, it should never use the random generator 06:27:58 you're supposed to morph them all 06:27:58 it doesn't matter much if they get morphed every 20 turns or 1, the result should be nearly the same if the code works right 06:27:59 -!- Ashenzari has quit [Ping timeout: 240 seconds] 06:28:44 -!- Ashenzari has joined ##crawl-dev 06:28:44 -!- The topic of ##crawl-dev is: DCSS Development | Logs: http://tozt.net/crawl | People with +v have commit access. | Please keep general crawl-related chit-chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Long stuff to pastebin.ca please | Immortal Warwalrus and Crazy Yiuf forever. 06:29:32 yeah, but the changes are supposed to be small; it's a matter of CPU use. But you're right here, it's not that bad. 06:29:46 so if you morph them all each turn but with only a little change in the third dimension, then only few of them will actually change, right? 06:30:00 and we'll have a smooth morphing and a nice layout 06:30:27 anyway, have to go now. 06:30:35 look forward to seeing you fix this :P 06:31:37 -!- RichardHawk has quit [Ping timeout: 276 seconds] 06:37:31 it fails to pass time_taken, too 06:37:32 -!- syllogism has joined ##crawl-dev 06:37:32 -!- Ashenzari has quit [Ping timeout: 276 seconds] 06:38:11 -!- Ashenzari has joined ##crawl-dev 06:38:11 -!- The topic of ##crawl-dev is: DCSS Development | Logs: http://tozt.net/crawl | People with +v have commit access. | Please keep general crawl-related chit-chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Long stuff to pastebin.ca please | Immortal Warwalrus and Crazy Yiuf forever. 06:41:01 -!- elliptic has quit [Ping timeout: 258 seconds] 06:54:56 -!- Ashenzari has quit [Ping timeout: 260 seconds] 06:55:50 -!- Ashenzari has joined ##crawl-dev 06:55:50 -!- The topic of ##crawl-dev is: DCSS Development | Logs: http://tozt.net/crawl | People with +v have commit access. | Please keep general crawl-related chit-chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Long stuff to pastebin.ca please | Immortal Warwalrus and Crazy Yiuf forever. 07:04:21 -!- Mu_ has joined ##crawl-dev 07:04:25 -!- Ashenzari has quit [Ping timeout: 260 seconds] 07:05:23 -!- Ashenzari has joined ##crawl-dev 07:05:23 -!- The topic of ##crawl-dev is: DCSS Development | Logs: http://tozt.net/crawl | People with +v have commit access. | Please keep general crawl-related chit-chat to ##crawl. | Dev wiki: http://crawl.develz.org/wiki | Long stuff to pastebin.ca please | Immortal Warwalrus and Crazy Yiuf forever. 07:05:49 there are two options: 1. bmh's Worley noise, and 2. old Abyss with your purely random changes (mostly current trunk) 07:06:13 if the latter, all the Worley code can be thrown away as it doesn't contribute anything at all 07:13:25 -!- Pingas has quit [Read error: Connection reset by peer] 07:13:39 -!- Pingas has joined ##crawl-dev 07:14:05 -!- ais523 has quit [] 07:14:47 -!- ais523 has joined ##crawl-dev 07:14:55 -!- ais523 has quit [Changing host] 07:14:55 -!- ais523 has joined ##crawl-dev 07:19:25 -!- dtsund has joined ##crawl-dev 07:21:16 -!- RichardHawk has joined ##crawl-dev 07:23:59 kilobyte: I think your reaping commit broke allies dying. 07:24:09 -!- ainsophyao has joined ##crawl-dev 07:24:26 But I don't immediately see why. 07:28:45 Fyren: fixed it literally ten seconds ago, testing the fix 07:29:36 not just ally, any monster fighting another can cause this 07:31:57 spent too long time trying to analyze saves... this mindex/mgrd corruption doesn't survive a save/restore 07:32:37 03kilobyte * rb60e43113b9b 10/crawl-ref/source/fight.cc: Fix mgrd corruption on monster-vs-monster combat. 07:33:48 galehar: could you trigger an update on CAO? This bug is quite nasty. 07:36:19 Unstable branch on crawl.develz.org updated to: 0.10-a0-695-gb60e431 (32) 07:38:42 -!- monqy has quit [Quit: hello] 07:47:19 haah... first jewellery shop after CDO update, and the ring of Vitality in it :) 07:48:23 give it to me, i'm in need of some vitality right now! 07:57:54 -!- dtsund has quit [Quit: dtsund] 08:00:01 galehar: Good work on the planning page 08:06:42 -!- ivan has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 08:06:42 -!- djinni has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 08:07:38 -!- djinni has joined ##crawl-dev 08:09:40 -!- ivan`` has joined ##crawl-dev 08:22:38 -!- Guest39776 has quit [Read error: Connection reset by peer] 08:22:52 -!- Guest39776 has joined ##crawl-dev 08:35:14 -!- monqy has joined ##crawl-dev 08:35:26 -!- elliptic has joined ##crawl-dev 08:39:15 -!- ZChris13 has joined ##crawl-dev 08:42:50 kilobyte: I don't have rights on CAO 08:46:45 I'm going to change job soon. Less time to crawl. 08:56:34 kilobyte: I just got gloorx's level a second time in the same game 08:56:46 I assume this is caused by whatever you did with ignacio? 08:57:25 I thought that was fixed? I guess not 08:57:51 elliptic: uhm, isn't this done by a vault tag? 08:58:11 -!- valrus has joined ##crawl-dev 08:58:13 kilobyte: I don't know 08:59:06 ghrmblah, no. It checks whether the Pan lord is still alive. 09:01:18 adding a check to exclude them from "not dead if left in a Pan level" logic would be the easy solution 09:02:16 I did not add it in the first place to make it easier to allow for dpeg's ideas of allowing the likes of Gloorx to join the orb run fun, though 09:02:23 (besides having the code simpler) 09:02:45 -!- valrus has quit [Ping timeout: 260 seconds] 09:03:38 btw, what do you guys think of this idea? (giving Pan/Hell lords you've stolen from a chance to spawn during the orb run) 09:03:57 galehar: this is what I meant :) 09:04:36 yeah I know. Just wanted to clarify 09:10:31 -!- MarvinPA_ has quit [Read error: Connection reset by peer] 09:10:53 -!- MarvinPA_ has joined ##crawl-dev 09:18:13 elliptic: good work with the meph/evap nerf 09:18:29 should we also scale the cloud effect with the power? 09:21:02 not sure 09:25:25 hyperbolic (L27 HESk) ASSERT(vault) in 'dungeon.cc' at line 2430 failed. (Pan) 09:26:09 -!- hoody_ has joined ##crawl-dev 09:26:49 elliptic: ok, we'll test it like that and see later if this is needed 09:27:29 hyperbolic (L27 HESk) ASSERT(vault) in 'dungeon.cc' at line 2430 failed. (Pan) 09:27:58 okay, I guess this character isn't playing more until that bug is fixed... 09:30:26 elliptic: in the meantime, you can have a look at the updated planning: https://crawl.develz.org/wiki/doku.php?id=dcss:planning:planning 09:31:04 anything you're volunteering to do? 09:31:22 (don't have to be already there, we can add it) 09:31:33 -!- Fail_Bot has joined ##crawl-dev 09:32:40 Same question to anyone who is active btw (kilobyte? due? MarvinPA_? Keskitalo? Anyone else?) 09:32:48 Zaba? 09:34:07 If I'm designing a portal vault, I can design seperate "mini vaults" for the entire vault and then hack them all together as needed? 09:34:17 I'm looking at the ice cave as an example 09:34:40 galehar: currently thinking about tornado balance (which doesn't seem far off to me, though a small buff might be reasonable), not sure yet what I'll do after that 09:34:55 Fail_Bot: subvaults? yes. 09:35:30 Fail_Bot: sprints ii and iii make extensive use of subvaults and many other features, if you want huge examples 09:35:34 ??tornado[9] 09:35:34 tornado[9/10]: Nerfed to oblivion in .9 09:35:36 ??tornado[10] 09:35:36 tornado[10/10]: < rwbarton> isn't tornado supposed to hurt things 09:35:55 I was trying tornado in sprint II and it seemed lackluster. 09:36:05 Fail_Bot: you could also try serial vaults but I don't think that's what you want 09:36:11 kilobyte: it has been very nice in my current game 09:36:31 kilobyte: "nerfed to oblivion" sounds like it's in jest, to me 09:36:51 It could last a little longer? 09:36:52 elliptic: I'm still not sure I can talk about it without using a lot of profanity. In July, I had a choice of going into a revert/shout fest or take a break to calm down. 09:37:09 and the rwbarton quote may have had context substantially changing its meaning 09:38:39 elliptic: in a vacuum, perhaps. It does compare terribly even to a pathethic (relatively!) spell like Shatter. 09:39:26 the absolute power might be acceptable if we want to nerf high-level spellcasting in general, but if that's the case, other spells should be nerfed thoroghly as well 09:40:01 I'd disagree that it compares terribly to shatter 09:40:56 for a single cast, only somewhat weaker as it needs time to do the damage 09:41:00 I wouldn't mind nerfing the storms a bit more, though 09:41:17 for a situation that calls for repeats, it is weaker by a factor of TEN 09:41:51 you seem to think that players aren't allowed to cast any other spells while tornado is in effect 09:42:01 this is sort of not true 09:42:02 of course, against greater nagas you'll cast it 1-2 times so in that case it is not that much weaker 09:42:27 elliptic: no, you forget that you are allowed to cast other spells while shatter is "in effect" too 09:42:32 uh, what? 09:42:40 (if for some reason you decide to pause 9 turns after casting) 09:43:13 right, but your "factor of TEN" was for if you cast shatter every turn compared to casting tornado once every ten turns and never casting anything else, no? 09:43:54 hmm, you're right, it's less than ten 09:44:40 no decent area spells though so you're not going to do much damage. Depends on the opposition, of course. 09:45:11 this char has mainly been using the tornado + airstrike combination 09:45:51 for the target audience CBL is not really going to do anything, though... monsters without rElec don't tend to need lev9 spells 09:46:04 also, note that shatter does 1/3 damage against fliers 09:46:05 galehar: should demigods be on the planning page ? 09:46:13 yes, of course, you can win using it 09:46:21 mumra: it is now 09:46:36 oh, just saw 09:46:41 if you try to use shatter to clear v:8... it is sort of nasty, you get swarmed by dragons of all colors 09:46:43 since you're actively working on it, it's planned for 0.10 09:47:01 elliptic: the set of monsters I did most of calculations on had >2/3 fliers 09:47:42 it also depends where you're using it 09:48:20 I haven't really played conjurers enough (just one three-rune win with Ice Storm), but nerfing the Storms sounds good to me. (: 09:48:34 tornado succeeds in giving AE a viable way of dealing with rElec monsters... in that context it looks pretty good 09:48:35 there's a big difference between charging head-on in a Ziggurat vs typical Pan lord vaults 09:48:40 there's still disagreement on where it should end up, but at least we all agree what direction in needs to go right now. Buff. 09:49:38 Now, the quesiton is how. when I nerfed the 0.8 tornado, I increased the denominator, when kilobyte buffed it, he raised the number of dices, thus reducing variance. 09:49:57 so, more dices or smaller divider? 09:49:59 kilobyte: sure, it isn't the best spell in pan... but it was great for clearing v8 and vestibule, and useful in pan for killing fiends with rElec 09:50:44 I think a 20%ish increase in damage would probably be reasonable, but I really don't think it is so far off as is 09:51:30 kilobyte: tornado + airstrike dealt with ignacio without real trouble, for instance 09:51:32 I really like how it sounds that Tornado has good uses and bad uses. What I dislike about the Storms is that they deal so much irresistible damage. 09:52:06 keskitalo: the percent of irresistible damage was decreased in 0.9, actually... but they still might be too good 09:52:08 elliptic: I disagree, if you cast Firestorm once and let them soak in the clouds, you still do 4-5 times more damage 09:53:02 would it be reasonable to make Torando more useful in other ways than damage? From the looks of it, moving monsters around isn't very useful, for one 09:53:04 kilobyte: well, certainly firestorm is better than tornado... this is desired behavior. and firestorm won't be so clearly better against all the rF+++ enemies 09:53:15 btw, I plan to improve the AI to make it smarter regarding clouds. Maybe some monsters will be better at getting out of storms' cloud and tornado's range. 09:53:16 monqy: it is useful for ninjaing the orb! 09:53:23 elliptic: so storms are not only faster and more versatile, but even in the "cast once and wait" mode they win not by tens of percents but by that 4-5 times 09:53:33 elliptic: oh, true! 09:53:46 kilobyte: they certainly don't win by 4-5 times against all enemies 09:53:55 ..and that's why Fire Storm should be nerfed. ;) 09:54:03 Keskitalo: amen! 09:54:13 -!- hoody_ has quit [Remote host closed the connection] 09:56:11 I'm thinking let's start by raising tornado from 7d back to 9d (where it was for a short time in july). We'll reevaluate once the AI changes are in. 09:56:17 an interesting tidbit: I'd expect Firestorm vs Icestorm should be proportional to the number of targets. Yet somehow Firestorm's total damage rises quite a big deal faster when there's a crowd. 09:56:38 kilobyte: maybe vortices are more likely to hit stuff? 09:56:56 galehar: 9d/10 and then look up, right? 09:57:18 galehar: 9d/15 is a bad joke looking at how much it has to gain 09:58:12 we won't get anywhere if we double the damage, then nerf it again, then boost it.... 09:58:16 let's take small steps 09:58:19 elliptic: yeah but I'd expect that to be mostly proportional 09:59:00 monqy: moving monsters around is also great for getting a crowd away from melee range 09:59:09 galehar: yes, doubling is a small step, as it'd still be a good deal weaker than the other spells 09:59:17 mumra: I thought it didn't do much for that. apparently I'm mistaken 09:59:35 mumra: perhaps a buff could be seeked in those areas? 09:59:45 kilobyte: as I suggested above 10:00:28 you'd typically want stuff to get into range rather than out of it, though 10:00:38 oh? 10:00:40 kilobyte: seriously, I know you've done a lot of simulations, but the spell is not so weak as is... doubling the damage is ridiculous 10:01:03 kilobyte: why would you want monsters to move closer to you when they are already in the radius? 10:01:45 I was quite satisfied with new tornado in taking out zot 5 10:01:51 Especially combined with chain lightning 10:03:16 galehar: and what about this: only that unnoticeably small damage boost you suggest, but double the duration? It'd still be weaker than competition, but we'd stress what makes it different. 10:03:38 28.5% damage boost is unnoticeable? 10:03:45 really 10:03:54 ...double the duration? That makes it ridiculous 10:04:08 The current duration is enough to take out an orb of fire caught in the radius 10:04:13 elliptic: when the competition is 400-500%, it is 10:04:19 kilobyte: no it isn't 10:04:29 you are an air mage 10:04:32 you are not a firestormer 10:04:34 RichardHawk: ... 10:04:52 kilobyte: Why the ellipses? 10:05:57 RichardHawk: not even halved its hp, and got four bad mutations in the meantime 10:06:14 what's your spell power? 10:06:32 106 10:07:02 here's another idea to boost it and differentiate it more than the storms: reduce the noise 10:07:23 RichardHawk: do you have stats on your character such as to calculate your tornado power? 10:07:25 it's currently at 25. Storms are at 20 or 25 depending on radius. 10:07:29 I'd like to compare the spell powers 10:07:35 A moment 10:07:36 kilobyte: I just tested at 137 spell power (which is quite reasonable) and the orb of fire went to "almost destroyed" 10:07:54 It should be possible , in the late game to get a 200 power tornado right? 10:07:57 http://pastebin.com/bm6yec5E 10:08:00 I don't know how much power affects tornado, but I think it would be interesting to see what an actual character has, vs. kilobyte's test 10:08:08 spell power affect both damage and duration, so it has a very significant effect on total damage output 10:08:19 with good skills and a staff, you can reach power 150 10:08:28 got to go, sorry 10:08:39 FaMott: 200 isn't very plausible, but 150 is certainly doable 10:08:48 and 130 is pretty standard IME 10:08:59 elliptic, isn't 200 like, 2 enhancers and 27 skill in all the relevant skills? 10:09:05 no 10:09:09 well, yes and 50 int 10:09:20 having max # is 150 power, not 200 10:09:28 3 enhancers then 10:09:35 you can't get 3 enhancers for air 10:09:39 oh right 10:09:41 Hm, so I was at over 150 power then 10:09:47 only 2, robe of archmagi and Staff 10:09:55 but just one enhancer and 30 int and 27 stats is 137 power 10:09:57 I was thinking ring of ice robe and staff of ice :X 10:10:05 Which is doable 10:10:09 with 27 skills, 72 int and a staff power is 196. 10:10:24 2 rings of ice? 10:10:31 So how much would it be for the three runer I posted, exactly? 10:10:31 galehar: you can use a robe of archmagi also 10:10:34 galehar: Robe of Archmagi 10:10:39 with a more realistic int of 40, power is 156 10:10:51 oh right 10:10:54 200 power just wrecks crap doesn't it? 10:11:08 RichardHawk: 161 10:11:54 35 int, max skills and 2 enhancers: 172 10:12:03 TAB auto-attack ignoring agate snail recovering in shell (https://crawl.develz.org/mantis/view.php?id=4536) by XuaXua 10:12:03 galehar: possibly tornado damage should scale less with spell power 10:12:20 Duration should scale less, damage should scale massively 10:12:23 galehar: I guess I'll work on the tutorial for 0.10, I'll edit the plan and put up a page later. 10:12:32 most other spells just scale linearly with spell power, tornado is more like quadratic I think? 10:12:35 Keskitalo: great! 10:12:38 is the bondage bot? 10:12:47 this might account for part of why kilobyte's simulations don't seem totally in accordance with reality 10:12:53 is the bug bot. I think 10:13:13 it's the bug tracker bot, to be precise 10:13:14 since the difference between power 100 firestorm and power 100 tornado should be larger than between the same spells at power 140 10:13:23 the god was named after the bot, rather than vice versa 10:13:23 and why it is so hard to make it viable in the midgame without being totally overpowered in Zigs 10:13:36 galehar: My plan is to address the Dungeon Olm usabilitity testing findings, and get the changes tested in yet another usability course next spring. 10:13:43 Tornado in midgame what? 10:14:26 ZChris13: yes, now that it's single school, you can cast it in mid or mid/late game 10:14:27 ais523: Actually, Ashenzari the god was named long before it was implemented, and the bot was introduced in the mean time. 10:14:27 galehar: well, it is definitely viable in midgame currently :) my KeAE who got it at L15 or so liked it too... 10:14:49 Where exactly do people define midgame? 10:15:00 midgame is very long 10:15:14 from end of lair to zot? 10:15:22 starts at or soon after lair/orc and continues until V8, IMO 10:15:27 Keskitalo: my favourite naming loop is on "Enterprise"; the space shuttle was named after the Star Trek spaceship, after which Star Trek canon was changed to make the spaceship named after the space shuttle 10:15:31 zot is definitely endgame 10:15:38 in fact, probably endgame starts at the third rune 10:15:41 or maybe even earlier 10:15:57 Keskitalo: can add a section to the planning for the tutorial and the changes you're planning to do? 10:16:00 anyway, have to go to 10:16:01 a fifteenruner is normally referred to as mostly being in "extended endgame" 10:16:03 'later 10:16:18 endgame starts as soon as you have a reasonable try at the first rune. 10:16:19 ? 10:16:25 no 10:16:39 -!- galehar has quit [] 10:16:40 then endgame would start in lair for many characters 10:16:54 Okay 10:17:11 anyway, I just checked and a single cast of power 161 tornado does seem to kill orbs of fire 10:17:15 what a useless spell 10:17:21 Midgame is anything up to vaults 7 10:17:31 I forgot that Tornado acts as 2 negative power enhancers 10:17:35 it doesn't 10:17:36 when you're in it 10:17:37 no 10:17:37 Oh 10:17:39 not anymore 10:17:39 I thought it did 10:17:42 okay 10:17:45 For a while it did 10:17:47 so you can cast other things while you have it 10:17:52 like, airstrike 10:17:53 It's wonderful for shooting chain lightning out of 10:18:13 yes, like airstrike on the stuff that tornado is conveniently levitating for you... 10:19:46 \o/ 10:20:58 If anything Conjure Ball Lightning is the AE spell I can't seem to find any proper use for 10:21:50 ball lightning is situational; but great crowd control at the right moment (assuming you have rElec) 10:22:18 two trials of power 161 firestorm vs orb of fire both needed 5 casts to kill it 10:22:38 CBL could use a buff probably, yes 10:23:52 elliptic: Convenient levitation sounds fun! 10:24:18 !lm mikee class=ae uniq=the royal jelly -4 10:24:18 2. [2009-09-04] mikee the Spry (L25 KoAE) killed the royal jelly on turn 112162. (Slime:6) 10:24:29 that's mikee killing trj with a few casts of cbl 10:24:41 @??the royal jelly 10:24:42 royal jelly (08J) | Speed: 14 | HD: 21 | Health: 230 | AC/EV: 8/4 | Damage: 5008(acid:7d3), 3008(acid:7d3) | Flags: sense invisible | Res: 06magic(196), 03poison, 08acid+++, asphyx | XP: 15000. 10:24:49 hm, somehow I thought it had rElec 10:25:10 anyway I killed it this game with one cast of tornado combined with airstrike 10:25:26 Still think the ball lightnings should at least reliably explode on vanishing 10:25:45 RichardHawk: when they time out, you mean? 10:25:52 elliptic: yes 10:26:12 that could be cool even if I'm not sure it would make the spell that much stronger 10:27:05 You can use magic dart to detonate them on command. 10:28:03 It's a lot of setup and turns used for a spell just one level below Chain Lightning 10:28:47 maybe CBL could be level 6 10:28:49 yeah, I have been thinking of reducing it to L6 10:29:08 possibly combined with some minor nerf 10:34:26 -!- ainsophyao has quit [Remote host closed the connection] 10:49:44 -!- hoody_ has joined ##crawl-dev 10:55:14 -!- aristid_ has quit [Remote host closed the connection] 11:06:50 -!- elliptic has quit [Read error: Operation timed out] 11:06:54 is it still cheap amusement to leave TrSu ghosts behind everywhere? :D 11:12:25 -!- hoody_ has quit [Remote host closed the connection] 11:18:02 What are these webs lying around everywhere? 11:29:33 how do i shot web lol 11:29:43 ... 11:30:11 No seriously, what are these things. I can't figure out what they are or where they come from. 11:30:20 -!- aristid_ has joined ##crawl-dev 11:30:43 ZChris13: I almost gave you a NetHack answer, then realised I was in ##crawl-dev and deleted it 11:30:52 I think they come with the spider branch, in Crawl 11:31:00 and they're similar to net traps, but not exactly 11:31:04 I found D:13 completely covered in webs 11:31:20 d:13 is always completely covered in webs i think 11:31:20 They are a lot like net traps only reusable, and not as strong. 11:31:23 to test them 11:31:50 vaults that the Crawl devs are testing are nearly always really nasty to the player 11:38:20 -!- elliptic has joined ##crawl-dev 12:06:19 -!- st_ has joined ##crawl-dev 12:07:06 -!- st__ has joined ##crawl-dev 12:07:19 -!- st_ has quit [Read error: Connection reset by peer] 12:19:04 -!- MarvinPA has joined ##crawl-dev 12:22:46 -!- MarvinPA_ has quit [Ping timeout: 252 seconds] 12:39:32 a new fixedart: a +0 club, casts radius 1 Shatter on hit (no digging, though) 12:39:49 I need a message for when it deals damage. 12:43:18 -!- galehar has joined ##crawl-dev 12:44:40 galehar: OK, I'll let dpeg know. 12:44:41 !tell dpeg default layout wasn't a typo. I do intend to improve slightly the default layout for big terminal/resolution (smarter default options). Will explain later if you want. 12:44:48 -!- galehar has quit [Client Quit] 12:50:47 03kilobyte * r5fe7dec5908c 10/crawl-ref/source/ (dat/des/branches/pan.des dungeon.cc): Fix unique Pan lord levels generating multiple times. 12:50:50 03kilobyte * r9cc3ee3f91df 10/crawl-ref/source/ (5 files in 2 dirs): A shillelagh. 12:50:57 -!- st__ is now known as st_ 12:52:36 Terrible lag on webtiles.... >.< 12:55:48 I think some demons should change colours 12:55:55 namely sun demons and cacodemons 12:56:02 why / why 12:56:39 though i actually like the idea of a demon that has colored glow, and the kind of color determines what standing in it does 12:56:39 because sun demons are now on the tormentor glyph and hellions have too much colour overlap with other 2s 12:56:59 oh hey you could make sun demons that and have the effect be debuffs, kryptonite style! :D 12:57:17 green sun = -rPois! 12:58:23 -!- dpeg has joined ##crawl-dev 12:58:38 well, i don't think sun demons need to change just because they're on the old tormentor glyph, necessarily? 12:58:45 did anything happen after the yred discussion earlier, btw? i am not reading this channel as often 12:58:53 !tell galehar Didn't want to step on your toes with the "skills screen". Sorry. 12:58:53 dpeg: You have 1 message. Use !messages to read it. 12:58:54 but it'd be good to do something with the 2s so there's less overlap with hellions, definitely 12:58:54 dpeg: OK, I'll let galehar know. 12:58:59 -!- valrus has joined ##crawl-dev 12:59:11 !messages 12:59:12 (1/1) galehar said (14m 32s ago): default layout wasn't a typo. I do intend to improve slightly the default layout for big terminal/resolution (smarter default options). Will explain later if you want. 13:00:08 kilobyte: how many randarts to come? :) 13:00:53 I've changed sun demons to red, cacodemons to cyan (picked at random mostly) and I put tormentors on lightcyan because lolpink 13:01:15 also I changed hellions to lightred, same as fiends 13:01:51 -!- Fail_Bot has quit [Read error: Connection reset by peer] 13:02:05 -!- claus has joined ##crawl-dev 13:03:00 dpeg: quite out of good ideas, I'm afraid. That short sword that goes sinusoidally from fire to cold was declared "meh" by folks here, so I'll probably skip it. 13:03:40 -!- valrus has quit [Ping timeout: 260 seconds] 13:04:18 going to implement Twisted Res having the corpses assemble themselves, and then it'd be time for hard stuff like merging messaging parts of Cryp71c's branch or attempting Hellspider design. 13:05:06 kilobyte: ring of the octopus king? :) 13:05:37 Eronarn: what would it do? 13:06:21 -!- ZChris13 has quit [Quit: ChatZilla 0.9.87 [Firefox 6.0.2/20110902133214]] 13:06:45 kilobyte: 1 AC, 1 EV, 1 Str, 1 Dex, 1 Int - can show up in a game up to 8 times, each with another property (rF+, rC+, etc.) 13:06:55 kilobyte: sounds cool! 13:07:28 good but not amazing for a human, wonderful if you're an octo with several of them 13:07:40 I believe we have enough cool content for 0.10, if we all get it in, that is. (Spider branch, octopode new species, constriction fighting move, DG changes) So pushing those sounds great! 13:07:52 kilobyte: but nothing is wrong with new, cool artefacts. 13:08:15 another new, cool, relatively simple artifact would be a bludgeoning weapon that forces you to trample 13:08:18 Eronarn: players love the "collect them all" minigames 13:08:22 i thought this might be cool on a GSC of speed 13:08:50 dpeg: yes, i want to add in something even harder than the plog challenge :) 13:09:38 of course, it's great flavor that a full set = +8 AC/EV/Str/Dex/Int 13:10:07 (magic the gathering made a legendary octopus card. casting cost 8, 8 power, 8 toughness, hits 8 targets when it attacks - very resonant) 13:10:17 collect them all, huh? 13:10:45 how about a weapon that gets better the more types/genuses of monsters you kill with it? 13:11:09 Zannick: TDHS in reverse? 13:11:23 kilobyte: tdhs? 13:11:31 Zannick: there's, of course, the WWII thing of painting kills on your plane 13:11:48 Zannick: Transdimensional Hellspider 13:11:53 {slicing, MP+9, Kills: AQTYbgh} 13:11:54 -!- claus is now known as Fail_Bot 13:12:03 MarvinPA: by the way, when I did pan yesterday I had a level that had nothing but a randlord vault and like 4 nexoqecs, was the generation changed? 13:12:14 i guess long inscriptions might still be a problem though? 13:12:18 kilobyte: oooh. i haven't seen that in action yet, but yeah, i guess 13:12:33 also, chaos spawns seem to be summoned by stuff with summon demon and spawn in randlord vaults 13:12:44 Zannick: I think there is a weapon like this already in 13:12:54 -!- dexap is now known as paxed 13:12:58 oh, that reminds me, another easy implementable - make chaos spawn breathe chaos clouds 13:13:08 currently they only drop them when they die, making them very avoidable 13:13:36 you will really never run into the clouds apart from xom / vaults, which is a shame 13:13:43 dpeg: cool 13:14:04 (a chaos cloud breath weapon facet would also be a cool DS thing) 13:14:05 Yeah, there is a power for kills weapon: Wyrmbane 13:14:21 oh, that happened? that's cool i guess 13:14:30 also quasits for summon imps 13:14:58 please collect the good ideas somewhere (can be a mail to me), no time to do much right now 13:15:01 why do we even have quasits, they're a straight up D&Dism 13:15:49 4.1a has cool quasits, they're like mini executioners than drain dex 13:15:55 ^ 13:17:11 it has glow imps right 13:17:14 we should have glow imps 13:17:24 st_: summoning quasits and chaos spawn is a result of the changes i think, yeah 13:17:32 it has a ton of different imps 13:17:35 no idea about the level with only 4 monsters though 13:17:51 str and dex drain should be more of a thing in general, almost everything targets int 13:17:57 MarvinPA: would you take care of the name changes we discusses the other day? (imp -> crimson imp, fiend -> brimstone fiend) 13:18:23 MarvinPA: and if really good impy ideas come up here, code or collect, please :) 13:18:35 white imp -> snow imp! 13:18:38 with snowball throwing action 13:18:45 that is not urgent 13:19:08 yes, still some months till christmas season 13:21:01 an abominable snowman unique 8? ;) 13:21:54 sure, as far as imps go there's probably room to remove maybe one of them... is there much of a worthwhile distinction between lemures and ufetubi? 13:22:17 ufetubi are fast and lemures don't exist 13:22:19 more or less 13:22:56 Compton (L11 TrBe) ASSERT(!n_users || CrawlIsCrashing) in 'package.cc' at line 180 failed. (D:13) 13:22:57 MarvinPA: no idea. Removing is okay (we can always bring back later when interesting ideas come up). But! Think carefully which imp to remove. The names, the names... Of course, +4 for the D&D variety :) 13:23:15 well, i have no idea which names are d&d-inspired and which aren't :P 13:23:29 ufetubi probably is not :) 13:24:36 lemures are in d&d *and* nethack apparently! 13:24:47 !lm Compton type=crash -log 13:24:47 1. Compton, XL11 TrBe, T:10570 (milestone): http://crawl.develz.org/morgues/trunk/Compton/crash-Compton-20110913-182252.txt 13:24:59 *and* ancient Rome! 13:25:21 sixfirhy isn't d&d inspired ;) 13:26:04 MarvinPA: well, lemures are ancient 13:26:41 fair enough, i'm sadly uneducated on names of minor demons :P 13:27:28 sixfirhies are Sat Oct 23 2010, 00:12 CET 13:27:54 lemures should be undead rather than demons, though 13:28:09 them being demons is a D&Dism 13:28:21 is there a point to midges existing 13:28:58 on a swamp especially 13:29:05 -!- Kek has joined ##crawl-dev 13:29:08 RichardHawk: from the point of view of the midges, definitely 13:29:17 -!- Leissi has joined ##crawl-dev 13:29:30 hi there, which git branch should I use to create a diff so I can send a patch? 13:29:51 master 13:29:52 Kek: A present! What is it? 13:30:06 patches for the patch god? 13:30:21 mumra: <3 13:30:22 isn't a punk a patch god? 13:30:24 autooffer option @ init.txt 13:30:33 nice 13:30:47 mumra: hi! Was really excited to read about your DG patch announcement 13:30:51 lo 13:30:58 i've got it working now, basically 13:31:06 -!- jle__ has joined ##crawl-dev 13:31:23 needs endless tweaking and customising the minions and items, but it's all done thru a text file 13:31:33 mumra: you rock! 13:31:47 ;) 13:32:07 I bet there'll be a CDO main page announcement "Quere up your demigods, ladies and sirs!" 13:32:15 *Queue 13:33:42 also, is there any preferre method for creating the patch file itself? 13:33:43 Eronarn: what Yrdy changes? 13:34:06 Kek: sorry, I cannot reply about patches. But someone else should be able to help. 13:36:33 would be nice. also: i thought about adding a key for deactivating autooffer (like ctrl+a). would alt+o be ok? 13:36:57 Kek: it's a lot better to have a git repository we could pull from (gitorious and github provide free hosting if you don't have a server yourself); for minor patches "git format-patch" is enough 13:37:14 It might be better to have a key that lists the in-game toggles than a separate key for each 13:37:49 Kek: alt+p sounds better, since offer = p 13:38:47 makes sense. i really am wasting more time than planned on this.. my idea was to send a patch for 0.7 (which is debian stable source), but folks here told me it was a bad here 13:39:09 so I'll just send a patch and add a new key to toggle, if its not a major problem 13:39:28 *separate key 13:39:39 *bad idea 13:40:28 Kek: always use trunk, please. Everything else means a lot more work to us. 13:40:57 yes, thats what they said. trunk=master? 13:41:08 mm 13:41:15 Kek: yes 13:44:47 -!- dpeg has quit [Quit: Lost terminal] 13:45:03 -!- dpeg has joined ##crawl-dev 13:46:51 k done cloning. thx for the attention and keeping the best game ever alive :) 13:46:59 back to code, bye 13:47:03 -!- Kek has quit [Quit: Page closed] 13:48:33 Character select forgets chosen weapon (https://crawl.develz.org/mantis/view.php?id=4537) by Tiber 13:51:07 Does crawl have a hook for capturing monster deaths that I can use to script things in Lua? 13:52:38 Fail_Bot: https://crawl.develz.org/wiki/doku.php?id=dcss:help:maps:lua:triggerables#examples_of_usage 13:52:45 See the "skele_death" trigger 13:57:09 dpeg: regarding minions spawning - how close to the player do you think they should appear? 14:08:42 -!- purge has joined ##crawl-dev 14:12:03 -!- dtsund has joined ##crawl-dev 14:13:33 dpeg: the yred stuff was about allies, there are some issues with the current formula 14:13:36 mumra: oh, all versions are okay: very close (adjacent cells), can fill whole LOS in cases; further away but still in LOS; just beyond LOS; anywhere on the level (no message) 14:13:43 Eronarn: ah, I recall 14:14:45 -!- ainsophyao has joined ##crawl-dev 14:15:22 i have some ideas for new undead which could fit yred well but before i do a writeup i'd want to know if yred gifting is going to change 14:15:46 Eronarn: yes, surely. If someone has a good plan, it'll go right in, I'd guess. 14:19:55 dpeg: ok; so far it's just adjacent 14:24:50 mumra: no problem, but it will be interesting to have variants later on. (A cheap one would be spawning 3-5 jellies across the level -- message or not is left to the code :) 14:32:15 -!- Zaba has quit [Ping timeout: 252 seconds] 14:33:35 dpeg: god could speak the first time you encounter the minion if it's out of los 14:33:48 Post-Tutorial crash (https://crawl.develz.org/mantis/view.php?id=4538) by Tiber 14:33:50 dpeg: although it raises problems with the 50-turn time limit 14:37:29 -!- syllogism has quit [] 14:37:30 mumra: just edited the triggerables wiki page 14:38:03 mumra: announcing that a minion is around/nearby has the advantage that the player can more easily avoid him. 14:40:39 hm, perhaps there will be enough ideas on the random gods forum thread that actually something like a design emerges 14:40:52 -!- Zaba has joined ##crawl-dev 14:42:46 -!- Letchik1 has joined ##crawl-dev 14:43:04 How do i turn off fullscreen in offline tiles? 14:43:40 dpeg: the apostrophe problem happens because ** is used for additional comments on those code examples 14:44:15 dpeg: changing those to -- should fix it, but then the original comments aren't differentiated from new ones 14:51:43 mumra: can't we use some other indicator? 14:53:19 hey, jpeg replied on the door gods, I hope she doesn't accidentally code those instead of the gold god :) 14:58:58 haha 14:59:04 well, door gods have an altar tile already :) 14:59:46 dpeg: could use "-- **" to indicate those comments ... i don't think there's another Lua comment syntax 15:00:12 mumra: yes, would be better 15:00:36 I could change a "don't" in the comments to "do not" but I was helpless with the executioner's axe. 15:00:47 Yeah, need to get me a Gozag altar tile. 15:05:30 -!- ainsophyao has quit [Remote host closed the connection] 15:06:56 -!- aristid_ has left ##crawl-dev 15:12:59 Tried to add a crawl save to a note in a bug report on mantis 15:12:59 gotr 15:13:01 got 15:13:01 PPLICATION ERROR #9This is a duplicate file. Please delete the file first. 15:13:06 wtf? 15:13:18 and why do devs ignore that report? 15:13:23 Letchik1: perhaps it is already uploaded? 15:13:29 Letchik1: which one is it? 15:13:45 no it's not 15:13:56 PPLICATION ERROR #9This is a duplicate file. Please delete the file first. 15:13:58 oh 15:14:07 wait 15:14:18 The interface can be a bit confusing at times. We've all been there. 15:14:19 https://crawl.develz.org/mantis/view.php?id=4332 15:14:32 maybe it thinks they are the same files because they have same names 15:14:43 it's just because i use 7zip-> add to saves.7z 15:14:56 yes, rename 15:15:42 almost 2 months from report 15:15:45 I think some targeting bugs were recently addressed. Keep pushing it (like you do with uploading another save, or adding a comment), it will be addressed 15:15:51 no comments, no fix, no nothing 15:15:59 Letchik1: freeware, unpaid developers, open source 15:16:23 All of us do this in their spare time. 15:16:29 dpeg: so your advice iis to bump the thread? 15:16:47 yes, upload the save, make a comment if the bug is still there 15:16:49 and be polite 15:16:58 developers hate impolite reports because of ^ 15:21:59 -!- dpeg has quit [Quit: zzz] 15:37:53 -!- MarvinPA has quit [Ping timeout: 276 seconds] 15:39:22 -!- MarvinPA has joined ##crawl-dev 15:54:42 -!- galehar has joined ##crawl-dev 15:55:11 shillelagh breaks tiles 15:55:12 galehar: You have 1 message. Use !messages to read it. 15:55:22 !messages 15:55:23 (1/1) dpeg said (2h 56m 29s ago): Didn't want to step on your toes with the "skills screen". Sorry. 15:55:56 galehar: Have you started on skill fuzzing? :) 15:56:19 no, I haven't 15:56:37 I've been messing around with the AI and the Abyss 15:56:40 Ok, just asking, I'm really looking forward to it. 15:56:52 Those are nice too :) 15:57:28 I think the Abyss worley's algorithm is beyond me, I'll just left it for someone who has some maths skills 15:57:52 but I do intend to do some serious work on the AI 15:57:58 then skills 15:58:06 I thought I had some maths skills, but this semester is killing me >.< 15:58:07 then permacast 15:58:13 that's my plan 15:58:35 Heh, no small fixes there :) 15:59:02 however, I'm back from restaurant, and I think I had too much wine to do any coding anyway ;P 15:59:02 Worley's algorithm? 15:59:21 the new abyss use worley's noise to generate layout 16:00:01 -!- ais523 has quit [Remote host closed the connection] 16:14:21 -!- greatzebu has joined ##crawl-dev 16:15:00 -!- ussdefiant has quit [Ping timeout: 260 seconds] 16:22:43 03kilobyte * r98edf438c360 10/crawl-ref/source/rltiles/ (2 files in 2 dirs): As usually, commit forgotten tile files. 16:41:45 Prono the Thanatomancer (L23 VpNe) ASSERT(!monster_at(p) || monster_at(p)->submerged() || fedhas_passthrough(monster_at(p))) in 'player.cc' at line 414 failed on turn 179170. (Coc:1) 16:42:35 Napkin: I think the osx builds should be updating correctly now, let me know if you see any more problems 16:42:53 roger, thank you, greatzebu :) 16:42:54 Napkin: You have 3 messages. Use !messages to read them. 16:44:15 great, sorry for the fuss :-) 16:47:08 don't worry 16:47:19 just stopped the cronjob for a while :) 16:49:03 !lm Prono crash -log 16:49:04 18. Prono, XL23 VpNe, T:179170 (milestone): http://crawl.akrasiac.org/rawdata/Prono/crash-Prono-20110913-214144.txt 16:50:19 OSX builds of master branch on crawl.develz.org updated to: 0.10-a0-692-g40c6b91 16:50:22 I guess that sort of thing might be caused by the mgrd corruption? 17:05:21 -!- MarvinPA_ has joined ##crawl-dev 17:09:00 -!- MarvinPA has quit [Ping timeout: 240 seconds] 17:13:09 -!- Letchik1 has quit [Quit: Leaving.] 17:14:02 -!- Letchik has joined ##crawl-dev 17:14:14 -!- Wensley has joined ##crawl-dev 17:20:30 -!- SamB has joined ##crawl-dev 17:20:44 I've got a phantom rat: https://crawl.akrasiac.org/saves/dumps/SamB-crawl-git-40c6b91f22-110913-1818.tar.bz2 17:20:53 right below the player 17:21:12 oh, wait, it's transient 17:21:13 drat 17:23:24 anyway, what happened was that targetting thought there was a rat in a place where there was no rat 17:23:38 -!- ussdefiant has joined ##crawl-dev 17:24:50 I feel like crawl's random name generator could be improved--it produces a lot of very strange names. Any opinions on this? 17:28:03 there are multiple name generators, i believe. which are you referring to? randart names? pan names? 17:28:10 pan lord names * 17:28:45 pan names actually don't seem as bad to me, mostly referring to un-id'd scroll and randart names 17:29:16 greatzebu: sure, it can be improved. you've got ideas? 17:29:18 Zannick: there's just one, it takes capitalization and length as a parameter 17:29:29 kilobyte: oh, huh 17:29:47 i don't even notice the scroll names when i play, usually 17:30:05 I was thinking about using a pronounceable password generator as a starting point 17:30:20 there's a thread on the tavern about name generator for minion names (demigods) 17:30:22 https://crawl.develz.org/tavern/viewtopic.php?f=8&t=2544 17:30:31 then you will tend to get syllables instead of just reasonable sequences of vowels and consonants 17:30:37 ah, ok. i'll have a look 17:32:14 with the interesting idea that if a name generator takes race, job and god as parameters, it could be fun to use it for player's name 17:33:04 yeah, if the generator had a set of phonemes you could tweak their likelihood to give names with different flavors 17:33:40 norse-ish, english-ish, chinese-ish, etc 17:34:28 not sure how good we could make the quality, but it might yield some interesting results 17:37:14 -!- MarvinPA has joined ##crawl-dev 17:38:08 i think dwarf fortress uses phonemes 17:38:15 i remember looking in the data files 17:39:48 interesting, i think they do a pretty good job with naming things 17:40:36 well, if we need a better name generator, using one from another open source game is certainly an option. No need to duplicate work. 17:40:42 -!- MarvinPA_ has quit [Ping timeout: 252 seconds] 17:40:46 true enough 17:41:07 it's a nice self-contained unit 17:41:14 just needs to be flexible enough to be adapted to our needs 17:41:18 so importing something shouldn't be that difficult 17:41:21 right 17:41:38 greatzebu: i've been doing that demigod work - there are certainly better ways that names could be generated 17:42:08 although, with the name generation database as i'm currently using it, it's nice having a certain amount of control - e.g. i can drop in a list of real names for humans 17:42:10 mumra: i like the idea of giving job specific surnames 17:42:11 df isn't open-source, though (is it?) 17:42:50 yeah, we should definitely keep the ability to choose from a pre-defined list 17:42:58 Wensley: no, i don't think so 17:43:08 discworldmud has a really phenomenal foreign language system, but I don't know how well that would translate into a random name generator 17:43:23 -!- Letchik has quit [Ping timeout: 260 seconds] 17:44:33 df is not open source 17:44:53 indeed 17:44:56 maybe we could entice toady to just lend us his naming code 17:45:22 lend would be insufficient. 17:45:32 that implies it could be taken back at some point 17:45:41 and i am utterly certain that would break our license. 17:45:44 "can i have my code back?" 17:45:53 or maybe there are other open source projects with great name generator 17:46:16 anyway 17:46:16 galehar: there would have to be, or at least academic articles out there onmethods for better generation 17:46:26 there's that one new roguelike that's using crawl's interface, or a facsimile thereof, that has randomly-generated uniques 17:46:29 forget what it's called 17:46:38 -!- galehar has quit [Quit: zzz] 17:47:19 what more do you want from a name generator? 17:47:39 greatzebu: i had an idea for writing an algorithm that would process a body of text, and analyse probabilities of particular letters showing up next to each other and in patterns; then using that data to spit random words back out 17:47:48 to be honest, I've never been dissatisfied with crawl's random names 17:48:04 so you could drop in, say, a list of chinese names, and get chinese-sounding names back out 17:48:18 mumra: i think that sounds very cool 17:48:36 Wensley: they're ok to a point but they're not used for that many things. i think if every god is sending out minions named using the same system, it'll just start to look like a whole bunch of random letters after a while 17:48:52 Wensley: currently, it's used for Pan Lord names, and unid'd scrolls 17:49:06 mumra: sounds like you want a markov model for names 17:49:11 Wensley: Beogh's blessed orc names use a predefined list 17:49:27 -!- evilmike has joined ##crawl-dev 17:49:31 algorithm-wise, I was thinking about a variation on this paper: http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA017676&Location=U2&doc=GetTRDoc.pdf 17:49:39 I think you might have that same problem with any reasonable name generator, even df's names gets pretty ridiculous once you start generating thousands and thousands of entities 17:49:54 Zannick: just looked markov up, looks interesting 17:49:57 where ideally we could adjust the digraph frequencies according to some text corpus like mumra suggests 17:50:08 markov is a good possibility too 17:50:52 Wensley: if you spend too long looking at an english phone book, you get kind of the same problem of course 17:51:44 so what you're saying is that you don't think that real life exhibits enough verisimilitude :P 17:52:06 but, over the course of a game of crawl where you'll only see a few dozen random names maximum, you just need an algorithm that ensures each name is suitably distinct, and that the names loosely fit the background of the character 17:52:08 beh, even the old three-part generator is usually good enough 17:52:10 very strange names seem good for pan lords. ok for scrolls. bad for followers 17:52:31 a couple of years ago I trained one on a bunch of names of (then) ##crawl-dev regulars 17:52:35 greatzebu: that's what i thought 17:52:48 the code: http://sprunge.us/JSCD (you can see how it works) 17:53:29 neat 17:53:35 wow, that's so cool 17:53:40 kilobyte: that's basically what i was going to do with monname.txt 17:53:53 more seriously, though: there are hundreds of good algorithms 17:53:57 should do that every year at tournament time, to refresh the list 17:54:21 kilobyte: in fact, i've already done that for Trog names; just invented short syllables instead of splitting up existing names ;) 17:54:22 some just ok, some damn good... it's a matter of picking one and tailoring its data 17:55:38 greatzebu: i really like the idea of using US military intelligence to generate monster names :) 17:56:02 :-) 17:57:16 mumra: US military intelligence? We can do better! :p 17:57:43 "British intelligence would have better names" 17:57:53 ;) 18:00:31 -!- ainsophyao has joined ##crawl-dev 18:00:46 not sure what you mean by "using US military intelligence to generate monster names" 18:01:17 oh, wow, they even include code 18:01:30 yeah, it's pretty detailed 18:01:33 ancient, though 18:01:54 translate it to lua and add 35-year old code to crawl, then? ;) 18:03:04 what language is it? 18:03:26 Wensley: that link greatzebu posted 18:03:41 it's a USAF document about generating random pronounceable passwords 18:04:12 greatzebu: the only thing is, if you look at the "2000 random words" appendix ... they look a lot like crawl's random names ;) 18:04:14 PL/I 18:04:21 Hm. 18:04:28 There's a chance it's public domain if it's a work of the US military. 18:04:32 i actually love crawl's name generator, even though it can be a bit weird and unpronouncable at times 18:04:32 evilmike: You have 1 message. Use !messages to read it. 18:04:36 it's great for pan lords, at least 18:05:43 well, we'd test it out and see if it works nicely 18:07:09 evilmike: i totally agree, the pan lords are fine ... they *should* be slightly unpronounceable :) 18:07:17 yes, agreed 18:07:46 at the very least, a phoneme-based generator should not be too hard to create 18:08:12 this usaf algorithm is generating syllables based on consonant and vowel phonemes 18:10:47 part of the reason i love pan lord names so much is because i once got one named "Fear" 18:11:22 that is pretty rad :-) 18:13:38 -!- greatzebu has quit [Quit: greatzebu] 18:15:14 evilmike: did you see that bug report, there was a pan lord called "Ring" and his description stated he was a piece of jewellery 18:15:28 hahahahahaha 18:15:48 haha, that's a cool bug 18:16:15 -!- Textmode has joined ##crawl-dev 18:16:32 has that been fixed yet? 18:17:49 yes https://crawl.develz.org/mantis/view.php?id=4503 18:22:12 -!- Leissi_ has joined ##crawl-dev 18:25:05 -!- Leissi has quit [Ping timeout: 260 seconds] 18:37:35 -!- greatzebu has joined ##crawl-dev 18:50:11 i almost got killed by Iced the panlord 18:51:53 Eronarn: ice, ice, baby! 19:05:13 -!- Galefury has joined ##crawl-dev 19:05:29 hello from canada :) 19:05:37 (i live in germany) 19:05:38 !lg * killer=stood 19:05:39 1. daftfad the Farming Annihilator (L27 DECj), worshipper of Vehumet, blasted by Stood (crystal spear) in Pandemonium on 2009-03-28, with 792545 points after 231612 turns and 31:47:32. 19:05:55 is there some overview of the demons branch changes somewhere? 19:06:06 like a wiki page or something? 19:06:50 03MarvinPA * r13bea137bcf0 10/crawl-ref/source/art-func.h: Don't hunger vampires and ghouls when wielding vampire's tooth 19:08:04 Galefury: what part of canada? just curious 19:08:09 bc 19:08:42 i live in vancouver myself 19:08:47 im pretty much in the middle of nowhere 19:08:50 next to quesnel lake 19:10:43 -!- dtsund is now known as dtsund-afk 19:11:07 heh yeah, not much out there besides trees 19:13:52 -!- MarvinPA_ has joined ##crawl-dev 19:14:57 damn i really cant find a demons branch overview 19:15:15 nothing on the wiki really seems to cover it 19:15:59 Galefury: why not look at the commit log of the branch 19:16:04 MarvinPA_ writes good commit messages 19:16:08 trunk was merged into it 19:16:16 i don't think there is one, but you could just @?? the demons or do what eronarn said 19:16:31 Galefury: http://gitorious.org/crawl/crawl/commits/demons 19:16:58 -!- MarvinPA has quit [Ping timeout: 252 seconds] 19:17:26 commit log was my first stop, but it seems to contain all the trunk changes too 19:17:32 and it has been around forever 19:17:34 there is a way to filter them 19:17:40 but you need to load it in git itself 19:17:48 gitorious web interface doesn't support it 19:17:50 hm, dont have git on this computer 19:18:43 alternately just look at MarvinPA_'s commits to demons here: https://gitorious.org/~marvinpa 19:19:12 doy did some earlier ones too 19:19:13 if you click the '1 commit' part, it loads it on the same page 19:19:23 with the initial split of abjuration, and re-tiering/colouring various demons 19:21:38 thanks 19:21:52 is the new abjuration called "mass abjuration"? 19:22:35 i really want there to be a hex/sum spell that curses something's ability to summon new creatures :( 19:22:48 all the creatures they summon are neutral 19:22:52 it would be like worshiping ely 19:23:00 wensley: yes, mass abjuration 19:23:23 Wensley: all the creatures they summon are butterflies! 19:23:45 or have very reduced durations. or show up already injured! 19:23:53 evilmike: need a late D vault that somehow has hostile butterflies 19:24:05 vampiric butterflies 19:24:17 Wensley: someone else can do that 19:24:26 i made a ridiculous butterfly vault already, that's enough for me 19:25:19 there should be a vault that is a 1x1 room and when you open the door to it butterflies just keep streaming out 19:26:36 the only demons commits that show up are the ones merged from trunk :( 19:26:44 -!- MarvinPA_ has quit [Read error: Connection reset by peer] 19:27:34 -!- MarvinPA has joined ##crawl-dev 19:29:12 just need to make a fog generator, but with butterfly clouds 19:29:29 I think one of the butterfly vaults was changed to bats? I preferred it butterflies. 19:29:37 Wensley: Eronarn's swarm patch has that already 19:30:25 yeah, butterfly clouds would probably work fine inside a fog generator, though i haven't tried 19:32:08 Eronarn: did you implement the spider swarms that eat corpses 19:34:24 Wensley: they don't eat corpses but that doesn't sound too hard to do 19:34:34 do miasma clouds instarot corpses? they should 19:34:54 (spider clouds can cling, though!) 19:35:03 did you just make it so that clouds can have arbitrary glyphs 19:35:21 maybe some explanation is necessary 19:35:37 I'd be willing to update your branch for the current codebase 19:35:54 but it sounds like you did more than the absolute minimum 19:36:46 Wensley: no... right now they all just use the normal cloud glyph. however, it would not be difficult to either put swarms on another glyph (like clouds on #, swarms on section since everyone hated that for clouds), or make swarms able to appear on any glyph 19:37:16 just curious how you made it so that spider clouds can't go over water (?) 19:37:28 -!- dtsund-afk is now known as dtsund 19:37:38 Wensley: there's a ground_level() function which checks whether you are lev/flying 19:37:39 -!- greatzebu has quit [Ping timeout: 258 seconds] 19:37:50 the 'spiders' monster, representing the swarm, can cling but can't fly 19:38:07 therefore, if it's above water or lava and isn't at a valid cling location, it drowns/burns 19:38:11 Eronarn: i had an idea how to implement swarms as monsters instead of clouds; while retaining the ability to swarm onto the player's tile 19:38:26 Eronarn: so it removes the problems of having to apply monster AI and movement to clouds 19:38:27 I think swarms should be able to move over other monsters as well 19:38:35 Wensley: they can 19:38:43 and I also think that swarms should not have very substantial AI 19:38:55 uh, this is sort of a bad bug here 19:39:06 mumra: this is totally doable; the issue then, though, is that stuff that affects clouds has to be rewritten to also affect swarms (this is harder than making clouds move as if they were monsters, imo) 19:39:07 Wensley: with eronarn's implementation as clouds, swarms could go on any tile of course; but it means that other clouds can't interact with swarms 19:39:13 -!- greatzebu has joined ##crawl-dev 19:39:21 mumra: the plan was to make it so that all other clouds will overwrite swarms 19:39:31 Eronarn: what do you mean? anything that affects clouds also affects monsters 19:39:36 nope 19:39:44 can you give examples? 19:39:49 for example, clouds getting moved in the abyss 19:39:54 or pushed around by tornado 19:39:57 Wensley: but then, e.g., fire clouds can't burn swarms 19:40:07 Eronarn: all monsters get pushed around by tornado 19:40:13 my swarm code handles this; you would have to add additional code if swarms were cloud-like monsters rather than clouds-with-monsters-attached 19:40:17 mumra: the swarm will stop existing when the fire cloud spawns in its square 19:40:22 Eronarn: and what do you mean about the abyss? 19:40:42 Wensley: it won't make sense to players if clouds affect swarms differently than other monsters 19:40:44 mumra: there's a bunch of cloud functions for things like swapping two clouds, or moving one around, or etc. - they are not guaranteed to be used in places that monster functions are 19:40:54 !tell galehar I don't know whether this is an AI bug or something else, but this kobold suddenly stopped moving and won't even attack me when I move next to it: http://crawl.develz.org/saves/dumps/hyperbolic-b60e431-110914-0039.tar.bz2 19:40:55 elliptic: OK, I'll let galehar know. 19:41:06 so you could end up with a situation where clouds and swarms operate differently 19:41:19 mumra: it does make sense, swarms are their own class of monsters which already have their own unique traits 19:41:28 Eronarn: well that's my point; swarms and clouds should be different things 19:41:33 weakness to clouds is perfectly fine as one of these traits 19:41:47 but, what about a cloud of fog? 19:41:52 kills them 19:41:54 hmm 19:41:57 mumra: people smoke out bees 19:42:08 also, it's not really a 'weakness' to clouds currently 19:42:19 any more than fire clouds are 'weak' to steam clouds since they can't overwrite them 19:42:25 -!- ixtli has joined ##crawl-dev 19:42:38 Eronarn: what's the current behavior for swarm/cloud interactions? 19:42:52 also, since we have monster stats, we can do things like: the chance of a meph cloud overwriting (killing) a swarm tile is based on power vs. swarm's HD 19:42:54 yeah, that's a weakness of the cloud system in general; but for most existing clouds it makes sense 19:43:14 alternately, you could make placing a cloud on a swarm cause diffusion 19:43:23 the swarm splits into several much lower HP/HD swarms 19:43:37 Eronarn: the way i'm thinking is, swarms can be ordinary monster tiles; but when they want to move over a monster or the player, you confer a "swarmed" enchantment onto that monster/player 19:43:38 which can then re-merge, if say, they're heading down a corridor at you 19:43:56 Eronarn: so you don't need all the special-casing for the cloud behaviour; rather, a single special case for monster behaviour 19:44:02 and you can use existing AI 19:44:23 (e.g. the Herd AI which is already a kind of swarm; you can tighten the variables) 19:44:28 mumra: that's not a horrible way to do it but it would introduce plenty of complications 19:44:39 like what happens if something with the swarm enchantment is banished 19:44:51 the swarm goes with you 19:44:57 just like if you blink 19:45:10 (see what's going to happen with nets blinking with you) 19:45:22 I think it's fine if tloc effects work on the whole square 19:45:41 okay so now you have to handle bringing the swarm monster into the abyss with you, and having the swarm enchantment still in effect properly :) 19:45:49 there would be lots of weird code bits like this to take care of 19:45:51 I think it would be cool if more tloc effects affected a larger region, in fact. louise banishes you and comes with you! 19:45:53 enchantments always stay on you when you're abyssed 19:45:58 it wouldn't require any extra code 19:46:06 bleh affected 19:46:07 yes it would - to make the *monster* part work properly 19:46:10 no 19:46:18 the monster wouldn't exist while you have the Swarmed enchantment 19:46:25 and then it respawns next to you when you lose it 19:46:47 that's a lot of properties to stuff in an enchantment, unless you want a swarm to e.g. lose confusion if it successfully attacks you 19:46:50 (actually, it would exist, just not on the grid - you can stored an mindex on the player properties) 19:47:08 yes that's what i mean, you need to handle that :) 19:47:12 a swarm couldn't attack you if it was confused 19:47:28 Eronarn: stored a single property on the player is much easier than inventing an entirely new monster grid :) 19:48:00 what is what you had to do to make the clouds work, right? 19:48:13 mumra: well, it's actually just a second mlist, not a new mgrid. you could achieve similar effects by having a 'this is invisible' parameter on monsters and storing them in mlist but not mgrid 19:48:32 do u mean menv ? 19:48:37 the advantage of a separate list is that you have less unexpected interaction effects 19:48:52 since nothing is looking at the one you made for 'not truly on level' monsters 19:49:15 uh, whatever it's called, i forget :) there is the one that is a list of monsters, and the one that is a grid 19:49:21 (why would swarms even be confusable) 19:49:41 Wensley: butterfly swarms are permaconfused, for starters :) 19:49:44 Wensley: if they're not poison-resitant and don't have clarity, why *wouldn't* they be confusable? 19:49:58 baby swarms are vulnerable to alistair's 19:50:00 because confuse spells target single creatures, and swarms are hundreds of creatures? 19:50:08 ah yes, alistair's 19:50:08 by which i mean, swarms of babies 19:50:09 Wensley: mephitic clouds? mass confusion? 19:50:18 meph cloud would just kill them 19:50:36 Wensley: well, that's just an assumption - for example, even with my swarm approach rather than mumra, let's say you have one tile of spiders 19:51:12 you could cast meph on top of it, subtract some amount of duration/HP from the spiders, then disperse that HP over several swarms, then run normal dissipation for duration, adn then if any spider clouds rae left they might be alive but confused 19:51:33 currently my code doesn't do this but there's no reason it couldn't really 19:51:52 trickier is stuff like swarms of sparks in a meph cloud 19:52:09 how do you handle that? it's like 'how do you handle a fire cloud and a meph cloud' conceptually 19:52:20 Eronarn: anyway; you can take a monster off mgrid, but leave it in menv; you don't need a whole new grid or list to have an off-grid monster 19:52:23 or another monster in a swarm of sparks in a meph cloud 19:52:46 mumra: right, that's what i meant about the interaction effects 19:52:59 Eronarn: exactly, i think swarms should just be monster tiles, with some special casing so they can share tiles (e.g. like Fedhas worshippers can share tiles with plants) 19:53:16 you could leave it in menv, but then anything looking at menv and accidentally picking up on swarms will have to be handled 19:53:36 to me, it seems cleaner to have 'menv' and 'not_on_level_menv' 19:53:49 but things look at mgrid rather than menv for interactions 19:53:52 but you could accomplish the same effect by adding a parameter to the monster 19:53:56 because mgrid has the coordinates 19:54:00 mumra: generally yes but not always 19:54:23 for example it doesn't iterate over mgrid to figure out turns 19:55:16 but again, this is kind of not material to the idea of having monsters which aren't in mgrid :) the main point being that mgrid is what prevents monsters from sharing a space 19:55:20 Eronarn: that's not a problem. you can use the swarm's turn to attack the player. 19:56:04 Eronarn: do you know how Fedhas worshippers get around it? (i started trying to find out but didn't yet) 19:57:01 mumra: that's just an example - the point is that i don't know what all checks menv, and so it was a lot cleaner to make a second copy of it (especially since i want to use the framework for other things: like, say, a monster that is banished from time and so ceases to exist for a while but then comes back) 19:57:10 but, it's not a huge deal to change it 19:57:48 figuring out whether to do swarms as clouds, monsters, or both is more important 19:58:04 and no, no idea about fedhas 19:58:06 man, this all sounds like such a pain in the ass. I just wanted clouds of spiders!! 19:58:41 Wensley: really it can be done pretty simply... but simply isn't necessarily the right answer 19:59:06 for example, simply means ozo's refrig doesn't affect swarms, or that a xom effect that moves clouds doesn't move them 19:59:14 !learn add Wensley Wensley> man, this all sounds like such a pain in the ass. I just wanted clouds of spiders!! 19:59:15 wensley[9/9]: Wensley> man, this all sounds like such a pain in the ass. I just wanted clouds of spiders!! 19:59:29 clouds of spiders are the new vampire mosquitos 19:59:30 I will give catlobes 2.0 spider breath 20:00:06 Wensley: mutated guard dogs in cigotuvi wizlab that bark bees at you 20:00:11 yes 20:00:29 oh man, spider swarms need to leave webs in their wake 20:00:57 Wensley: i think they need to *emerge* from webs at random :) 20:01:13 they emerge from egg sacs! and also spiders laden with young 20:01:32 also FR: egg sac trap 20:01:43 just a fog generator for spider swarms 20:01:59 or a summons trap 20:02:06 Wensley: demonic crawlers burst into spider swarms and pus EVERYWHERE when you kill them 20:02:11 yessss pus 20:02:18 will go so well with my disease hell 20:02:24 rivers of blood and pus and urine 20:02:33 also, hey, that bee thing makes me realize 20:02:39 can recolour the blood tiles to make pus 20:02:53 one reason bee swarms would be an even better replacement for early game poisoning is that you can make bee swarms dissipate whenever they successfully sting 20:03:04 but also they follow you 20:03:11 so staying in clouds becomes better, and running from them becomes worse 20:03:48 i hadn't thought of that previously but that would make breathe bees a pretty compelling replacement for evap 20:07:07 Eronarn: I'm still not quite sure what you're saying about effects that move clouds. e.g. what's the Xom effect you mentioned? 20:07:46 -!- Pingas has quit [Read error: Connection reset by peer] 20:16:47 -!- ixtli_ has joined ##crawl-dev 20:17:26 piranhas clouds! 20:17:39 -!- ixtli has quit [Ping timeout: 252 seconds] 20:17:39 -!- ixtli_ is now known as ixtli 20:22:38 -!- Guest39776 has quit [Read error: Connection reset by peer] 20:22:49 -!- Guest39776 has joined ##crawl-dev 20:29:26 mumra: check out cloud.cc - there's lots of functions for moving clouds 20:29:32 i have no idea where they are all used 20:30:47 there aren't that many; they're mostly to do with cloud spread 20:30:54 and swap_clouds is used only by tornado 20:32:31 but i don't see any reason why a spider swarm would have to move unless its AI told it to 20:32:40 or unless tornado 20:33:41 one advantage of using an enchantment for being swarmed, is that if you move then the enchantment will just move with you (i.e. spiders are crawling all over your body, you can't just walk away from them) 20:34:44 mumra: on the other hand, this means you can't readily outrun swarms 20:35:04 Eronarn: not once you're swarmed; you have to run away before that happens 20:35:21 Eronarn: on the other hand, you could have a Dex check every turn, to see if you manage to shake the swarm off you 20:35:32 i wonder if a hybrid approach would work? a swarm 'attacking' you gives you the swarmed status effect, which then processes its attacks; even if you leave the cloud, you are still swarmed for a while 20:35:59 the last few lingering bees, etc. 20:36:54 - this could work whether or not the swarms are clouds or monsters, btw, just the idea that there's a separation between 'standing in a swarm' (this is just terrain for some kinds) and 'covered by a swarm' (status effect) 20:38:24 this would have applications for swarms that themselves have status effects too... like walking through a sleeping swarm :) 20:38:43 or warding not preventing being surrounded, but preventing being swarmed 20:39:42 i think walking thru a sleeping swarm would be pretty likely to wake them up :) 20:39:58 well, there's magical sleep, too 20:40:10 ah yes 20:40:55 or you could have a vault that's just severed body parts everywhere, around an altar of Yred, but they don't do anything until you either convert to Yred (they join your side) or take an item 20:41:18 http://magiccards.info/dpa/en/33.html 20:41:22 what, the body parts? 20:42:02 well, not join your side as in follow you around, but suddenly all animate and start swarming out of the shrine 20:42:13 they'd dissipate soon after 20:42:20 but it'd be very flavorful 20:42:35 aren't body parts too big to properly constitute a swarm ? 20:42:54 i personally peg a hand as about the largest thing that could belong in a swarm 20:42:57 you realise, of course, that the twisted resurrection changes will have animated body parts :) 20:43:22 yes, you could use 'parts swarms' in the spell itself - have the swarms move to locations where aboms are forming, but attack stuff on the way 20:44:18 i definitely feel a melee character should be able to fight against body parts, which would be a problem if they were clouds 20:44:20 gotta give fish their own glyph, so that body parts can go on ; 20:45:06 mumra: the rationale is that if you have a pile of a hundred hands, you can cut them apart in a way you can't to flies, but not fast enough to make an appreciable difference in their number before they would've dissipated anyways 20:45:34 but: curse toes 20:45:50 curse toe swarms! :) 20:45:51 and how do body parts "dissipate" ? 20:46:13 no longer concentrated enough to be a swarm (a room with some bees in it vs. a bee swarm) 20:46:22 of course, dissipation can be tweaked a lot 20:46:33 so where did these hundred hands suddenly go so you can't see them anymore ?? 20:46:59 yes, it definitely works better for more insect-like ones :P 20:47:31 but if you have a pile of a hundred undead hands milling about in a large room, eventually they'll disperse enough to be about as threatening as mice one on one 20:48:10 i dunno, i'd find a single undead hand pretty threatening all by itself; depending on which of my body parts it had hold of, of course ;) 20:48:12 (probably a good approach for dispersal would be for swarms to each turn flood fill some % of themselves into adjacent squares, determined by some cohesion value) 20:49:59 sounds like slime creatures 20:50:45 swarms -> globs, "you see here a titanic glob of slime" 20:51:02 Wensley: another reason why the whole implementation works better if the swarms are just monster tiles; merging and splitting code is already there :) 20:56:11 mumra: what i'd say is: if you want to tweak it so it's based on monsters rather than clouds, go for it; i don't really have the time to rewrite it, but it shouldn't be too complicated 20:57:01 but that you might want to look to see how this will actually matter, in terms of swarms being monsters instead of clouds, because it could have lots of unpredictable effects 20:57:14 Eronarn: i wanted to have a look at it after demigods; would be good to have for spider's nest. 20:57:37 Eronarn: I agree; but conversely, i think having them as clouds will bring a whole load of other problems. On balance i think there are more issues with clouds. 20:58:01 demigod swarms 20:58:03 Eronarn: e.g. adding special casing to every single spell that could possibly affect them, to do additional checks for swarms 20:58:36 yeah, that would be an issue - right now spells dn't affect swarms and i was mainly planning to handle radius/area spells 20:59:46 with monsters, all the normal logic applies. the only think is the status effect. and the majority of spells you can say "well that spell doesn't affect anything ON your tile" 21:00:26 mumra: the one concern i'd have is with spells like magic dart - i don't feel like this should affect a swarm 21:01:22 Eronarn: well yeah, you can set the EV of the swarm very high so projectiles will miss (does EV affect that?) 21:01:36 magic dart autohits 21:01:57 also if the swarm were paralyzed it would have 0 EV 21:02:09 just give it absurdly high HP 21:02:11 also i think there's always going to be a small chance of a hit even with very high EV? 21:02:36 i don't see a problem with *sometimes* being able to damage a swarm 21:03:16 mumra: i think spells should be able to damage swarms but that single-target spells should do very minimal damage. in practice this would require specialcasing 21:04:15 basically i want people to be able to fireball swarms, but not to magic dart swarms (or if they cast magic dart it does inconsequential damage) 21:04:15 true; but special-casing a few exceptions, rather than special-casing everything, is preferable 21:05:14 that's one of the reasons i went with clouds - i could just slowly roll out spells affecting swarms 21:05:21 rather than by default having all of them affect them 21:05:34 but most damage-dealing spells *would* affect them; the list that won't is very small 21:05:39 is there a distinction between magic dart autohit and fireball engulf autohit? 21:06:03 -!- valrus has joined ##crawl-dev 21:06:30 mumra: i disagree... bolt of iron, stone arrow, dart, imb, etc. lots of stuff that is single target-ish and shouldn't really do much to swarms 21:06:48 that list covers most of then 21:06:53 ^them 21:07:05 but there are, what, 20 or so area spells that could affect them? 21:07:26 i am pretty sure there are more single target spells than area spells 21:07:33 remember, there's stuff like monster breath weapons 21:07:41 breath should definitely affect them 21:07:49 it's a spacial volume, not a narrow projectile 21:08:00 except that breath can be stuff like spitting poison 21:08:14 some breaths are areas, but others definitely aren't 21:09:37 anyways this isn't insurmountable or anything but it is something to look out for, there really are a lot of spell effects that probably shouldn't do anything to swarms (and then there's the issue of communicating which spells do/don't affect them, which i haven't really thought much about yet) 21:12:33 -!- Mu_ has quit [Quit: Defecator, may everything turn out okay so that you can leave this place.] 21:12:57 could give each spell a new property, area_effect (or something) 21:13:22 that's actually not a bad idea... though there are non-spell effects to cover, too 21:13:30 god abilities, evokables 21:13:48 then again it's arguable that stuff that isn't currently a spell should be made one 21:13:53 precisely because of stuff like this 21:16:07 spells don't actually have a lot of properties at all; it's *all* special cases handled in code. but, you could add a new spell flag. 21:16:29 on the other hand, it's probably better to look at beam types; most spells, evokables, etc. come down to a beam type at some stage 21:16:43 so you just need a function that checks whether specific beam types affect them 21:17:01 it's going to be a pretty big undertaking regardless of which way you go, really 21:17:13 there's just so many ways of causing effects in crawl 21:17:32 but that just means it's better to start sooner 21:17:46 anyways, gotta go do stuff. have like 250 pages to read in the next several hours 21:18:33 Eronarn: good luck heh 21:19:12 nice 21:19:54 just had a quick look at beams: high EV will make the majority of stuff miss them; can just special case Magic Dart beams (which cover all autohits) 21:20:04 so i don't think it's anything like as hard as you're making out :) 21:21:02 anyway; laters ;) 21:23:33 -!- ainsophyao has quit [Remote host closed the connection] 21:29:41 -!- mumra has quit [Quit: HydraIRC -> http://www.hydrairc.com <- IRC with a difference] 22:02:18 -!- elliptic has quit [Ping timeout: 259 seconds] 22:07:24 -!- Pacra has joined ##crawl-dev 22:39:27 -!- elly has quit [Ping timeout: 248 seconds] 22:39:40 -!- elly has joined ##crawl-dev 22:40:31 -!- ixtli_ has joined ##crawl-dev 22:41:23 -!- ixtli has quit [Ping timeout: 252 seconds] 22:41:23 -!- ixtli_ is now known as ixtli 22:43:01 -!- Galefury has quit [Read error: Connection reset by peer] 22:50:54 -!- Galefury has joined ##crawl-dev 23:05:23 -!- Wensley has quit [Ping timeout: 260 seconds] 23:09:10 -!- Galefury has quit [Quit: Stay sane inside insanity!] 23:12:17 -!- Fail_Bot has quit [Ping timeout: 260 seconds] 23:43:36 -!- ixtli has quit [Quit: Computer has gone to sleep.] 23:59:35 -!- purge_ has joined ##crawl-dev