00:04:02 Catoblepas lair ending (https://crawl.develz.org/mantis/view.php?id=4469) by evilmike 00:09:18 Unstable branch on crawl.akrasiac.org updated to: 0.10-a0-404-ga93dd5f (32) 00:14:34 -!- Twilight13 has joined ##crawl-dev 00:14:34 -!- Guest42342 has quit [Read error: Connection reset by peer] 00:15:54 -!- valrus has quit [Remote host closed the connection] 00:41:37 -!- dtsund has quit [Quit: dtsund] 00:57:44 -!- Textmode has quit [Ping timeout: 240 seconds] 01:40:50 -!- ussdefiant has quit [Ping timeout: 260 seconds] 01:49:41 -!- ainsophyao has quit [Remote host closed the connection] 01:57:08 -!- GreatZebu has joined ##crawl-dev 02:02:44 -!- GreatZebu has quit [Quit: GreatZebu] 02:13:32 -!- hoody_ has joined ##crawl-dev 02:44:32 -!- ussdefiant has joined ##crawl-dev 02:48:07 -!- galehar has joined ##crawl-dev 02:50:26 hi 02:50:26 galehar: You have 1 message. Use !messages to read it. 02:50:30 !messages 02:50:30 (1/1) dpeg said (12h 15m 45s ago): I've stopped replying to the forum Spider thread. The eggs idea is very good. If you want to, we can discuss and flesh it out. It'll rock. 02:54:59 -!- st__ has joined ##crawl-dev 03:02:11 -!- zpmorgan has quit [Ping timeout: 240 seconds] 03:14:12 -!- zpmorgan has joined ##crawl-dev 03:34:44 -!- purge has quit [Quit: !] 03:42:41 -!- Zaba has quit [Ping timeout: 240 seconds] 03:43:11 03akbecker * rc65901aada9c 10/crawl-ref/source/ (files.cc files.h mapdef.cc mapdef.h maps.cc): Prevent des cache collisions by using path to file. 03:43:50 -!- Zaba has joined ##crawl-dev 03:45:42 moin 03:46:33 Napkin: meow! 03:46:39 as I see "akbecker" - should I add some mappings to the cia-hooks? 03:46:51 moin Adam :) 03:47:21 how's it going? 03:48:20 lots of work, shouldn't mess with Crawl... but then, a drunkard can't say no to booze 03:49:13 * due waves tiredly. 03:49:57 hehe 03:49:59 !coffee due 03:50:00 * Henzell hands due a mug of black coffee, brewed by Crazy Yiuf. 03:50:08 moin due :) 03:51:19 who is akbecker? 03:52:10 -!- galehar has quit [Ping timeout: 252 seconds] 03:52:16 * due doesn't have time for /anything/ 03:53:25 lemme gather the list. I got akbecker, but not several others. 03:55:33 was the hook in the repository updated? i could just copy it over 03:55:53 -!- galehar has joined ##crawl-dev 03:56:14 7032. 03:56:26 oh 03:56:31 I did wonder why that didn't work. 03:56:38 * due whacks focussed windows to death. 03:57:35 hehe 03:57:48 at least take your time to drink the coffee ;) 03:58:41 web traps look quite good in the screenshot, btw 04:01:02 yeah 04:01:36 it's a bit silly that a web attached to one side is thick on two (WE, NS) but that's noticeable only if you look closely 04:02:06 03kilobyte * r41fc7aa361ca 10/crawl-ref/git-hooks/crawl-ref-cia: A few updates to CIA aliases. 04:02:07 -!- Pacra has quit [Read error: Connection reset by peer] 04:02:12 Napkin: I'm tired of sleuthing, found only four names out of many 04:02:30 -!- Pacra has joined ##crawl-dev 04:02:38 that's great, thank you, kilobyte :) 04:02:40 IIRC you need to copy the hook over 04:02:47 -!- hoody_ has quit [Remote host closed the connection] 04:03:08 Napkin: if you can think of someone I missed, let me (or due, or galehar...) know 04:04:16 oh, don't worry - CIA will tell us ;) 04:04:47 :p 04:07:39 -!- ussdefiant has quit [Ping timeout: 252 seconds] 04:29:05 -!- mumra has joined ##crawl-dev 04:32:10 kilobyte: when i first did the webs, i forgot about ones attached on a single side, so i just used the ns / ew ones temporarily; will provide the missing ones later (and improve the others a bit tbh, i did them all very quickly) 04:32:44 but i have to do proper work today so it'll probably be this evening ;) 04:41:12 btw; let me know what you did to optimise the pngs - so i can do the same when i upload them again 04:43:45 mumra: optipng -o4 -i0 -fix image.png; advpng -z4 image.png 04:47:15 ok thanks 05:00:25 we should add this to docs/develop/tiles_creation.txt 05:01:28 yeah, good idea 05:02:23 -!- Pacra has quit [Quit: Leaving] 05:13:32 -!- ussdefiant has joined ##crawl-dev 05:13:42 -!- dave123 has joined ##crawl-dev 05:19:13 has the idea of making part of poison resistance a trainable skill been discussed in the past? 05:20:21 was just reading through the 90% poison resistance looking for something like that, might be more now that victory dancing is gone 05:24:32 Monsters will not wear more than one item of armour, unless they start that way (https://crawl.develz.org/mantis/view.php?id=4470) by ion_frigate 05:26:06 03kilobyte * r7b8ff1816e3d 10/crawl-ref/source/dat/des/builder/layout.des: Sometimes add water to Spider maps. 05:26:07 03kilobyte * r46a2ee91edd6 10/crawl-ref/docs/develop/tiles_creation.txt: Mention how to optimize PNGs. 05:26:07 03kilobyte * r6ce96cd4f694 10/crawl-ref/source/maps.cc: Fix map cache not being updated when a timestamp changes. 05:30:11 I think it's only Mithridates(?) and a few others that trained poison resistance... too much work, too dangerous and you'd have to immunize yourself against too many different agents 05:33:35 well if you're getting poisoned anyway while you're adventuring, surely some tolerance would build up (albeit slowly) ;-) 05:48:54 I'm not rabidly keen on the idea, but it seems like it could be interesting way to deal with at least some of the objections to "binary poison resistance" 05:49:14 anyhow, back to ##crawl for me 05:52:56 -!- monqy has quit [Quit: hello] 05:54:55 -!- dpeg has joined ##crawl-dev 05:57:30 -!- dave123 has left ##crawl-dev 05:58:33 -!- stabwound has quit [Ping timeout: 258 seconds] 06:05:21 -!- ussdefiant has quit [Ping timeout: 258 seconds] 06:06:18 -!- Ashenzari has quit [Ping timeout: 258 seconds] 06:07:08 -!- Ashenzari has joined ##crawl-dev 06:07:08 -!- 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. 06:11:42 -!- Wensley has joined ##crawl-dev 06:22:09 mumra: spider sense with range of 3 that works only out of LOS is kind of useless 06:22:36 did you mean ~10? 06:25:23 -!- galehar_ has joined ##crawl-dev 06:36:19 why only out of LOS? 06:36:28 spiders in LOS should also be alerted 06:36:39 that is the whole point of webs, no? 06:44:44 right, the out of LOS check is only for players 06:45:12 range of sqrt(10) -- ie, 3, stands though 06:46:13 mumra uses a pre-squared distance there but people falling in webs calls the function with argument of 10 06:46:44 he's AFW, I'll make it 100 06:52:06 03pete * r0519e133364d 10/crawl-ref/source/ (enum.h mon-data.h mon-flags.h shout.cc shout.h traps.cc): Spiders sense when trapped in web 06:52:16 03kilobyte * r4305f1e5298f 10/crawl-ref/source/files.cc: Fix a harmful use of an uninitialized variable. 06:54:46 -!- galehar_ has quit [Remote host closed the connection] 07:00:43 -!- evilmike has quit [Quit: something happened] 07:05:39 -!- syllogism has joined ##crawl-dev 07:10:29 -!- ussdefiant has joined ##crawl-dev 07:13:49 kilobyte: I can imagine that water should be added by hand to Spider, i.e. vaults 07:15:46 probably yeah, sounds like a better idea 07:16:17 kilobyte? 07:16:18 very hard to get it right algorithmically, I can imagine 07:16:28 could you please apply this patch? http://pastebin.de/18554 07:16:40 oh.. +- is wrong 07:17:25 updated: http://pastebin.de/18555 07:18:30 -!- voker57 has quit [Read error: Connection reset by peer] 07:21:37 -!- voker57 has joined ##crawl-dev 07:21:37 -!- voker57 has quit [Changing host] 07:21:37 -!- voker57 has joined ##crawl-dev 07:21:54 also, i upgraded the crawl-ref-email hook: http://crawl.develz.org/crawl-ref-email.patch 07:22:07 03mumra * rc3c860bcac3e 10/crawl-ref/source/dungeon.cc: Generating webs on dungeon build 07:22:07 03kilobyte * ra03587e102aa 10/crawl-ref/source/dungeon.cc: Don't place that many webs in the Crypt, way too spammy even for a test. 07:22:08 03kilobyte * r54c1e52da850 10/crawl-ref/git-hooks/crawl-ref-cia: Napkin's changes to CIA. 07:22:09 * Napkin pokes busy kilobyte * 07:22:45 gee, secretly working in the background! 07:25:11 * kilobyte sniffs Napkin. 07:26:10 03kilobyte * r81c39c2b8244 10/crawl-ref/git-hooks/crawl-ref-email: Napkin's updates to CIA-email. 07:29:50 \o/ 07:29:53 thanks ;) 07:30:02 kilobyte: Regarding watery spider maps, it would be mostly useful next to walls (for clinging). So maybe weight water a bit more when close to wall. 07:30:09 kilobyte: I don't know much about dungeon generation so this may be hard/impossible 07:30:16 dpeg: creating new webs is problematic if they never go away 07:30:46 perhaps there should be a chance a web disappears when someone disentangles? 07:31:42 kilobyte: i can't see where the range is getting square rooted ? 07:31:51 galehar: my current attempt does that; it does often create nooks you can't explore into unless levitating/clinging/wwalking though 07:31:58 it looked to me like vampire max blood smell range was 144 (12*12) 07:32:36 mumra: for vampire blood smell it's ok, it's stuff falling into webs that was called with a static value of 10 07:33:06 that would still be a range of 10 unless i missed where it gets rooted? 07:33:43 wasn't quite sure what kind of range to use anyway, so it was only a placeholder 07:35:05 also: yes there should definitely be a chance to destroy the web, but wiki didn't mention anything specific so i left it 07:35:21 -!- Mu_ has joined ##crawl-dev 07:38:06 oh, i see now - circle_def does the sqrt. that makes a lot more sense :) 07:38:36 back 07:39:10 kilobyte: oh, I always assumed that disentangling from a web destroys it. 07:39:25 sorry I failed to spell this out on the wiki -- it was so obvious to me 07:39:47 yep, destroying the trap when evading it makes sense 07:40:09 03kilobyte * rfb4d66978b11 10/crawl-ref/source/ (dungeon.cc effects.cc traps.cc traps.h): Grow more webs as time passes. 07:40:10 always or with a chance? 07:40:14 galehar: regarding the water, we're might be best off to create watery ambush vaults, and to sprinkle some water over the level so it feels natural (to have those watery vaults) 07:40:19 that way, it can even have a chance of spawning new spiders instead of just alerting nearby ones (or does it already do that?) 07:40:24 kilobyte: always, I'd say 07:40:54 kilobyte: does web growth depend on branch depth? 07:41:05 for water, maybe weight it more to shallow than deep 07:41:25 mumra: another thing, pretty important: to have a chance to get caught in a web, you need to be a troll with 2 dex... 07:41:33 hehe 07:41:53 I think mumra mentioned this yesterday, it's an artefact of how nets work 07:42:18 not entirely - you can get caught quite easily with 0 t&d and 0 dodging - but you always escape on the very next turn 07:42:21 hey, new lair ending, cateblepas 07:42:29 mumra: ah, sorry 07:42:49 because there's no dummy net item for the escape calculations 07:43:05 which is an artefact of escape_from_net (from memory) 07:43:38 btw, we have to solve the Spider branch end issue 07:43:46 kilobyte and I are fine with giving the Hellspider a shot 07:43:52 any objections? galehar? 07:44:28 kilobyte: the (admittedly very basic) weighting algorithm i tried in dungeon.cc didn't quite satisfactorily work, ended up with a lot of nets in open space and not enough of a tendency to corners and corridors 07:44:36 although it almost worked :) 07:45:29 galehar: the wiki proposal stated 25% chance of a spider spawning next to you when caught in a net, it's the one bit i didn't implement 07:45:48 didn't have time to work out the best way of spawning a monster that acted as summoned but still left a corpse 07:46:16 anyone motivated to build some 0.10 packages for macosx? 07:47:04 -!- MarvinPA has joined ##crawl-dev 07:47:49 Napkin: maybe greatzebu. He posted some 0.9 binaries on the tavern 07:48:25 !tell greatzebu hey there - motivated to build some 0.10 packages for osx which we could host on CDO? 07:48:25 Napkin: OK, I'll let greatzebu know. 07:48:31 let's see :) 07:48:56 mumra: if the web is always destroyed when you escape from it, the spawned spiders can be real ones instead of summons (much better flavour) 07:49:33 -- 07:49:33 CXX dgn-delve.o 07:49:34 dgn-delve.cc: In function 'int ngb_groups(map_lines*, coord_def)': 07:49:34 dgn-delve.cc:156: warning: 'prev2' may be used uninitialized in this function 07:49:37 --- 07:50:29 dungeon.cc:3387: warning: 'void _place_door_mimics(int)' defined but not used 07:50:35 galehar: new webs are randomly generated tho; but maybe reduce random monster generation in spider to compensate for xp scumming ? 07:50:56 it's not different than monster spawning 07:51:00 can't be, prev2 is used only for cardinal directions and a diagonal is done first 07:51:23 galehar: it is a bit different in that webs keep coming and you could in principle abuse them for xp 07:51:30 the mimics warnings is known but I'm not silencing it since galehar is rewriting that code anyway 07:51:34 not practical, I know, but players get all sorts of silly ideas 07:51:51 instead of spawning spiders, sometimes you spawn a web which itself will spawn a spider when you stumble into it. 07:52:26 kilobyte: ^ on web growth depending on depth... It'd could be very low on Spider:1, to ramp up such that webs hurt on Spider:5, imo. 07:52:27 webs can stop spawning when monsters stop spawning too 07:52:35 galehar: true 07:52:43 good point 07:52:58 kilobyte: feature mimics are done. I'll merge it soon. Working on item mimics now. 07:53:05 btw: i noticed a practical limit of 400 traps on a level (there was a query on the wiki regarding this) 07:53:10 galehar: will they start to walk? 07:53:45 there used to be a monster limit too... I hit that when I made a foresty oklob vault 07:53:49 can those be relaxed? 07:53:52 item mimics? yes. I've removed the annoying teleport and change behaviour. 07:53:58 it's a #define somewhere 07:54:19 galehar: great 07:54:24 I've reduced speed to 8 to compensate so they are still easy to escape 07:54:38 and to differentiate a bit from feature mimics 07:55:04 -!- ainsophyao has joined ##crawl-dev 07:55:31 galehar: then we also need faster-than-normal mimic type for symmetry :) 07:55:53 orb of zot mimic? 07:55:56 :) 07:56:09 You know that NH has plastic imitations of the amulet? 07:56:45 yeah I do :) 07:56:56 So I think everyone is fine with the Hellspider =) 07:56:59 I've ascended NH a few times before crawling around :p 07:57:38 dpeg: I haven't looked it up yet, sorry. 07:57:39 let me 07:58:41 it has its own section on the Spider wiki page now 08:01:14 I see that sharing code between webs and nets causes nothing but problems 08:01:21 going to disentangle them... 08:01:38 Windows builds of master branch on crawl.develz.org updated to: 0.10-a0-416-gfb4d669 08:02:08 kilobyte: yeah, good move 08:02:15 kilobyte: that was my conclusion 08:02:48 the problem is, there are so many places that have to check for being caught, so they'll all need web-specific code 08:02:49 so we need formulas for getting caught and for escaping 08:02:50 Eronarn was calling for unification, which sounds good in principle, but also like a lot of work someone has to undertake 08:03:00 global search for ATTR_HELD, it comes up all over the place 08:03:02 I thought escaping quickly was intentional 08:03:35 kilobyte: can we unify the escape checks for nets and webs, though? 08:03:47 i think escape should be quicker than nets, otherwise it'll just be annoying - but 2-3 turns per web might be about right 08:03:49 the held status might also be used for constriction, so generalizing it is good 08:03:50 The basic ideas are just the same: give bonus for Str; count earlier attempts etc. 08:04:22 mumra: yes, ideally there would be a parameter (we could also use that for net types) 08:04:24 dpeg: The only thing I don't like about Hellspider is the upgrade based on how you killed it. it encourage swapping attack for the killing blow. I'd suggest just random upgrade. 08:04:32 but if it's best to seperate right now, just go ahead 08:04:44 galehar: that's the core idea 08:04:47 galehar: not killing blow... keep track of how damage was done 08:05:09 if 90% of damage was from summons, and then you use the axe... that's still a summons kill 08:05:22 kilobyte: just had a look at "Grow more webs as time passes" - for those webs they should be generated out of LOS (ideally) 08:05:32 yes 08:05:32 ok, might work then. Not easy to implement though 08:05:33 ok, trunk website updated as well 08:06:20 galehar: possible... might be adventurous, but we want unique uniques, eh? As indicated, there should be a bunch of damage categories, plus some Other, to keep track of what's going on. 08:06:23 is elliptic back from vacations yet? 08:06:31 no, doesn't seem like it 08:06:51 I wanted to evaluate the tournament with him... This reminds me: gotta write a wrap-up post for blog. 08:07:10 i think morgue links have to wait until then.. he added the proper logic to the tourney stats page already and i really don't want to fiddle with python at the moment ;) 08:09:11 what about summons which do a specific damage type? Do we care enough? 08:09:16 (never wrote anything in python... so it would end up very messy!) 08:09:22 kilobyte: maybe not for the first installment 08:09:31 we can refine later 08:09:34 probably not: flaming corpses, simulacra, Summon Dragon... too few of them 08:10:02 perhaps we'll also find out that in order for the later HS to be a threat, we need one random and one battle-derived upgrade. We'll see. 08:10:09 if you can cast Summon Dragon you're not really caring about the HS 08:10:39 the HS should retaliate with summoned Transdimensional Helldragons, obviously. 08:10:50 dpeg: I thought about it a bit, "one pip of rF" is way too small to count 08:11:02 Unstable branch on crawl.develz.org updated to: 0.10-a0-416-gfb4d669 (32) 08:11:05 kilobyte: yes, may easily be. Just wanted to indicate the direction. 08:11:52 quickly maxing resistances means it won't have space to grow further, though 08:12:09 There are questions: suppose I am a Trog axe man with no desire to think. Should I be able to axe murder the HS five times? How many piety/resources I am expected to spend on HS5? 08:12:25 mostly for partially resitable or non-resistable sources 08:12:27 kilobyte: well, it grows four times... 08:12:38 dpeg: an axe is non-resistable 08:12:59 yes, but I expect the HS to get a lot of AC and HP if you kill it four times in melee 08:13:00 boosts of AC/EV/hp, that's all 08:13:09 hmm right 08:13:21 it should also get a lot more attack damage on its own, too 08:13:26 not a toothless paper spider 08:13:52 not sure how to explain ranged attacks if it would get any 08:14:34 kilobyte: we could be defense-only on those fronts: warding, shielding/repel/deflect/reflect 08:14:37 I think it might be better to skip those 08:14:38 yeah 08:14:54 with stress on reflection :p 08:15:04 In my vision, a melee-only boosted HS should be a monstrous melee opponent. Something that a single BiA cannot cope with. 08:15:52 It could also get random upgrades among better speed/jumping (blink)/spitting webs etc. 08:18:20 -!- galefury has joined ##crawl-dev 08:18:52 was reading the log 08:19:15 i think making the first few encounters relevant fights is more important than super crazy upgrades 08:20:48 if HS1 to HS3 are just weaker or comparable to regular branch monsters, HS4 is hard, and HS5 brutal, what's the point of HS1-3? 08:21:22 galefury: they govern the HS's development 08:21:25 HS1 should still be hard, just without elemental resistances 08:22:18 dpeg: i think those fights being autowins where all tactics rely on spoilery knowledge of what facets to avoid is stupid 08:23:44 -!- ussdefiant has quit [Ping timeout: 240 seconds] 08:24:50 how about if it just adapts to your character's main strengths, completely ignoring how you actually kill it ? 08:25:07 -!- RichardHawk has joined ##crawl-dev 08:25:53 or maybe some combination or random choice between the two ... 08:26:51 just looking at skills might be easier than tracking all sorts of damage 08:27:17 but less cool 08:27:35 and less skill based 08:28:02 well it requires increasingly more skill as you figure out different ways to kill it 08:28:11 galefury: I don't understand what you mean 08:28:14 why "autowin"? 08:28:29 if HS1 is just as strong as a regular spider 08:28:31 rather than the skill being in the first few fights working out how *not* to screw yourself over for level 5 08:28:41 and you already killed 10 of those on Spider:1 08:28:47 i would call that an autowin 08:28:58 as kilobyte said, it would be tougher than a regular Spider:1 spider already 08:29:41 I mean, rather than a progression of L*4 it can have 10+L*2 08:30:17 yes 08:30:25 well, something that is already tough will be nearly unbeatable for an FE with three pips of fire resistance 08:30:40 and upgraded base stats 08:30:47 that's the point. Diversify or die. 08:31:03 i think mumra's suggestion is nicer though, it reverses the order of things 08:31:04 it would be an option just to make it stronger each time regardless of kill type (i.e. more HP/AC/EV/damage, perhaps random upgrades), but it'd also be much less interesting imo 08:31:18 with damage based facets you have to diversify during the first few fights to make the later ones easier 08:31:20 players are always so negative 08:31:29 galefury: yes, that is good 08:31:40 We will explain the mechanic to the player, don't worry. 08:31:45 with character based upgrades you need to diversify later, because no matter what you do the first times around, your main way of dealing with stuff wont work well on HS5 08:32:10 "The rune transforms into the Transdimensional Hellspider. It seems to be capable of adapting very well." 08:32:14 or some other nonsense 08:32:49 galefury: I disagree. As a Troggie, I'd try to use wands and ranged stuff early on, using BiA and berserk for the last one 08:33:04 dpeg: they have a point: you'd rather fight him with alternate means first so you can use your main stick in the final fight 08:33:11 You want to reserve your strongest tool for the last fight. 08:33:24 kilobyte: yes, but what's so bad about that? 08:33:34 dpeg: it's annoying and spoiler-y 08:33:48 As I said upthread, we need to think how far a pure fighter (say) can go. It should still be possible, but at what expense? 08:34:14 Eronarn: it is annoying because it is not your idea and it is spoilery if and only if we don't tell players about it. 08:34:30 I'm pondering mumra's idea... it would resist both your main way of dealing damage and your primary alternate. Too bad, there's a huge loss of theme and strategy.. 08:35:02 kilobyte: as I said, the cheap alternative would be to just have it grow each time, perhaps with a random gimmick for each level. 08:35:17 We can start with that, if we feel like it. 08:35:21 dpeg: no, it's annoying because in a roguelike you can't just pick up a different gun and start using it 08:35:42 it would amount to killing it with a different wand a few times, then using your main kill method against it the last time 08:35:55 it depends on the numbers, I have no idea how you can make clear-cut statements like this 08:36:39 There are many ways to acquire summons, there are many helpful god powers, it is not a trivial matter. 08:36:56 dpeg: it *could* end up making the last fight actually the easiest - of course if it grows in other ways as well, that can be averted 08:37:25 i still think forcing the player to adapt during later HS fights would be better than forcing the player to game the system so the final fight becomes as easy as possible 08:37:25 mumra: I doubt that :) 08:37:51 hey, the thing gains mass, HP and everything during each round 08:38:10 the last round is not the easiest one, it is the one where you may want to use your main gun with full power -- that does not mean easiest 08:38:28 dpeg: 'use your main gun with full power' == 'forcing the player to game the system so the final fight becomes as easy as possible' 08:38:55 and to be able to use your main gun effectively in the last round, it's important that you haven't used it in the earlier ones 08:39:10 I think it is okay if we expect a caster to diversify -- they have it easiest. 08:39:25 This is why I've been asking how a pure fighter should perform against the five HS. 08:39:44 -!- Wensley has quit [Ping timeout: 245 seconds] 08:40:13 And the answer is our choice. 08:40:20 maybe it should just gain defense when it grows. that way, killing it with your main gun in the first or the last round is the same. it only gets harder if you keep using the same gun. 08:40:41 galefury: also valid 08:40:41 pure fighter = physical melee, physical ranged, wands (5 elements), potions, god 08:40:52 yes 08:41:02 how is 'potions' going to kill it, exactly? 08:41:05 berserk 08:41:13 might, speed, hw? 08:41:23 = physical melee 08:41:25 those are buffs, not damage types 08:41:42 berserk is listed as a potential facet in the proposal 08:42:24 also getting dispel on hit might be fun if you fight it with a lot of buffs up 08:42:28 which may not be the best part of the proposal given how berserk monsters are dangerous 08:43:10 galefury: i like the idea of it being able to sense if you have buffs, and gaining dispel if so 08:43:26 but then you might question why it doesn't seek out your blank resistances, etc. 08:44:09 is it supposed to only gain defenses as it goes up? or can it gain useful anti-foo attacks too? 08:44:14 like knockback if you keep killing it in melee 08:44:19 hmm, for some reason i just got an image of neo vs agent smith, and the final fight where they've both grown into superhumans but utterly evenly matched 08:44:38 knockback plus faster than average speed would be hilarious (and should be avoided) 08:45:12 no wait 08:45:17 okay, so we start with regular increase of the standard parameters, plus gaining certain defences from attacks used (resistance, warding/abjuration, repel/deflect/reflect), and perhaps a random upgrade (elemental flavour to attack damage; summon spiderlings; blink etc.) 08:46:46 if it's permissible i think it would be good for it to switch between attack modes... so if you're ranged, it tries to be melee, and if you're melee, it tries to be ranged 08:47:21 this could then inform the abilities it gets 08:48:33 if it gets help against ranged attacks (see above) and a power like blink, then this wouldn't even be necessary 08:49:52 blink only makes melee monsters easier, and blink close/blink away are melee/ranged specific 08:50:18 by blink I mean "blink towards player" 08:50:28 (theme: jumping spider) 08:50:29 dpeg: it's not about being necessary - it's about making the fight last longer so that it has more time to show off its stuff 08:51:02 casters getting meleed by it will have to try to get away; meleers getting kited will have to try to get close 08:51:32 if it just runs at everyone regardless of their build, then it may end up a boring fight for meleers 08:51:35 better AI can also be done,s ure 08:53:15 strafing could be nice for vs. melee... 'the transdimensional hellspider skitters from side to side!' 08:54:18 another interesting power could be both HS and the player teleporting in tandem 08:54:37 so youre still in melee with the HS, but your summons got left behind, youre not in a corridor anymore, stuff like that 08:54:45 also its very transdimensional 08:57:50 i think one cool power would be: the hellspider transforms into eight tiny, FAST spiders which run elsewhere in the room 08:58:27 then it reforms once it's far enough away from you 08:59:43 hey, we want the HS as a branch sentinel for 0.10 09:00:10 it doesn't want too many powers - you could actually make about a dozen different equally interesting uniques out of some of the powers mentioned here :) 09:01:57 yes :) 09:02:27 we should probably start with a very basic version, which ignores the player completely, becomes stronger on level up and gains random powers 09:02:30 then we can test already 09:03:38 spider, for .10? i've been really unimpressed with the portal vault's design 09:03:43 are there new monsters for it or something? 09:05:15 i think it will work better as a branch than as a portal 09:06:09 difficulty varied too widely depending on poison resistance and whether you get one with a ghost moth 09:06:31 ghost moths were removed from the portals a while back 09:06:36 yes... but i don't see how making it a branch would change that, unless you mean it's going to have tons of ghost moths or something 09:06:51 i just dont know where ghost moths are currently at, they were changed so often... 09:06:56 i personally don't think ghost moths should be even in spider:5 (they'll get stuck in the webs!) 09:06:58 ghost moths are currently in zot, right? 09:07:09 and are they using their old stats or their new stats? 09:07:12 @??ghost moth 09:07:12 ghost moth (06y) | Speed: 12 | HD: 13 | Health: 55-94 | AC/EV: 16/10 | Damage: 1805(drain dexterity), 1805(drain strength), 1205(strong poison) | Flags: fly | Res: 06magic(104), 02cold, 03poison | Chunks: 06mutagenic | XP: 3460. 09:07:20 i think those are the new ones 09:07:20 ahaha 09:07:28 @?ghost moth 09:07:29 ghost moth (06y) | Speed: 12 | HD: 10 | Health: 39-70 | AC/EV: 10/10 | Damage: 805(drain strength), 8, 12 | Flags: fly | Res: 06magic(80), 02cold, 03poison | Chunks: 06mutagenic | XP: 971. 09:07:38 oh, there we go 09:07:43 indeed 09:07:46 although, ghost moths could get M_INSUBSTANTIAL, since they're ghosts after all; then they can pass thru webs 09:07:57 they went from @?? to @? because they were RIDICULOUS in lair, but now they're back to what's in @?? 09:08:14 yeah, but they were also removed from spider 09:08:20 mumra: they aren't actually ghosts though 09:08:22 but seem to be getting readded 09:08:27 they're weird alien moth things 09:08:45 i think having two kinds of weird alien moth things might be good 09:09:07 because a zot-level threat in a lair branch is about as ood as it gets 09:09:09 eronarn: they could act like ghosts for the purpose of webs, they'd need to in spider:5 since that will be the highest web density 09:09:28 nothing against them being insubstantial, just pointing out that they aren't actually ghosts 09:10:01 galefury: i think we could manipulate brain worms into being 'ghost moth larvae' 09:10:37 as in make them actually turn into ghost moths under certain conditions? 09:10:38 galefury: the threat of the ghost monster is not the monster itself (at least not only), in this regard it is similar to the eyes or the wrath moth 09:10:49 no - but they're already w glyph monsters that have an anti-spellcaster attack 09:11:00 dpeg: this is blatantly untrue for current ghost moths 09:11:05 dpeg: ghost moths are pretty terrifying even without drain 09:11:09 @??naga warrior 09:11:09 naga warrior (02N) | Speed: 8 (act: 80%) | HD: 10 | Health: 92-107 | AC/EV: 6/10 | Damage: 28 | Flags: see invisible, !sil | Res: 06magic(80), 03poison | Chunks: 09poisonous | XP: 748 | Sp: poison splash (d16). 09:11:10 @??ghost moth 09:11:10 ghost moth (06y) | Speed: 12 | HD: 13 | Health: 55-94 | AC/EV: 16/10 | Damage: 1805(drain dexterity), 1805(drain strength), 1205(strong poison) | Flags: fly | Res: 06magic(104), 02cold, 03poison | Chunks: 06mutagenic | XP: 3460. 09:11:30 note that naga warriors usually have a weapon 09:11:37 and armor 09:11:44 they usually have a robe, which is negligible 09:11:51 oh, right 09:11:54 and ghost moths have 16 AC 09:11:57 @??stone troll 09:11:57 unknown monster: "stone troll" 09:11:57 they often get shields though 09:12:02 @??rock troll 09:12:02 rock troll (15T) | Speed: 8 | HD: 11 | Health: 41-80 | AC/EV: 13/6 | Damage: 30, 20, 20 | Flags: regen | Res: 06magic(58) | Chunks: 07contaminated | XP: 375. 09:12:29 yeah, they're not directly comparable - ghost moths are at least in th same tier as them though 09:12:33 there is no reason why ghost moths have to be melee threats 09:12:43 but they're invis and fast, whereas nagas aren't 09:13:06 dpeg: i agree with this, i pushed for their initial nerf strongly 09:13:41 i think that if the moths are the ranged threats, and the spiders are the melee threats, that would be cool 09:13:56 -!- hoody_ has joined ##crawl-dev 09:14:02 spider will be full of webs though 09:14:20 moths not getting caught by them is hard to explain 09:14:22 galefury: what about making it have lots of cliffs? let me pastebin it 09:14:34 Eronarn: i think its a little late for that 09:15:20 galefury: two solutions -- either make ghost moths incorporeal; give all non-spider Spider monsters the ability to bypass/circumvent/ignore webs 09:15:53 having more types of moths could be nice though, should they all be incorporeal? 09:15:58 galefury: like: http://pastebin.com/TCG3cCCD 09:16:06 oh, did someone write a layout gen for it already? 09:16:22 mostly was done a while ago in a branch it seems 09:16:29 and recently merged into trunk 09:16:34 layout_delve or something 09:16:37 -!- mumra has quit [Read error: Connection reset by peer] 09:16:45 which makes wide corridors? 09:17:17 galefury: what about a 'lantern moth' which simply burns any webs that touch it? 09:17:47 they are drawn to each other :) 09:18:04 -!- mumra has joined ##crawl-dev 09:19:15 i think being incorporeal makes the "ghost" part of the name make more sense 09:19:16 ...it can be explained with some handwaving like "transdimensional alien moth" ;) 09:19:31 or a sonic moth which simply uses echolocation so it never goes into webs willingly 09:19:45 another thing about webs: there was some discussion about spawning spiders when the player is caught in a web, and whether those should give exp 09:19:55 imo a strong no to that, because it would be scummy 09:20:03 galefury: if someone is willing to finish up the swarm code, you could use spider swarms rarely + randomly 09:20:04 webs should be avoided, not farmed for exp 09:20:27 'You're trapped in a web! As you wriggle about, thousands of baby spiders crawl out to feast on you!' 09:20:34 galefury: galehar pointed out that web spawning would be like monster spawning 09:20:57 yeah, but much of spider will be covered in webs, and imo the player should be encouraged to avoid them 09:21:04 not to seek them out to get some extra exp 09:21:04 did we end up using the small corridor idea for spider? 09:21:37 no 09:21:44 but there is another go at it for Hive 09:21:52 optimal play being clearing the level avoiding webs, then triggering every single one to maximize exp seems like a bad idea to me 09:22:08 too bad, that would be great for this situation - you could spawn the spiders in tiny rooms adjacent to the crevices, then spawn webs near their exits 09:23:27 while i think about it: has anyone discussed the idea of traps & doors skill helping you detect trapdoor spiders ? 09:23:58 mumra: no, but it's a great idea 09:24:01 galefury: that's a good point - what if webs were placed such that you couldn't get downstairs easily without triggering some? 09:24:14 (or burning resources) 09:24:38 and perhaps have destroyed webs also trigger spiders to come out, some of the time 09:24:49 i think the plan is to spawn them next to walls and especially in corners 09:24:49 eronarn: best way to do that might be using stair vaults surrounded by webs 09:24:58 eronarn: triggered web traps already alert spiders 09:25:00 because that would drive the player to fight in open spaces and look really cool 09:26:04 having a 'messenger spider' could also be cool 09:26:08 the weighting could be modified to slightly favour stairs, that could also look good 09:26:20 say - the spiders hide until webs are triggered, so if you avoid triggering webs, you see almost no spiders 09:26:21 galefury: good point. No exp for web spider spawns. 09:26:37 (much of the above is already in the 0.10 trunk build already available) 09:26:41 but some specific kind of spider patrols, and if it spots you, it runs to a web and starts 'plucking' it to summon its friends 09:26:51 If T&D helps revealing trapdoor spiders, it could reveal mimics too. 09:27:56 galehar: i think no XP may be premature - there are probably ways of making it work - it just needs to not be easy to totally segregate individual spiders 09:28:02 if you have to fight, say, a room of them at a time that is okay 09:28:15 maybe stop spawning spiders from webs after a while 09:28:27 then they could give exp without encouraging scumming 09:28:45 something like no more than 5 web-spawns per level 09:29:19 possibly depending on depth, because deeper = more webs 09:29:30 but spawning webs without spiders is useless 09:29:33 what about holding respawns in reserve until there's a certain number of them, and only then having that group show up if you touch a web? 09:29:54 well, swarms would be great, but aren't finished 09:29:59 could lower the frequency of spiders spawning from a web, and maybe reduce the frequency the more have spawned 09:30:26 Eronarn: the problem is if you want XP, you'll buff up and run into a web to call them 09:30:33 galefury: well, the mechanical parts of swarms are reasonably good along actually 09:30:43 at one point i even had spider swarms properly using clinging 09:31:04 it's more the conceptual issues - like how can players respond to a swarm coming after them (currently, fireball would do nothing) 09:31:21 i think for spider swarms there doesn't need to be a way to kill them 09:31:42 step on a web, a multi-tile spider swarm is spawned on you, and you want to get out of the web and run 09:31:50 the swarm dissipates and is gone 09:32:08 also webs call nearby spiders 09:32:31 there can be different kinds of spider swarms, btw. the basic one is just a monster called 'spiders' 09:32:37 based on the generic 'spider' monster 09:32:50 -!- ussdefiant has joined ##crawl-dev 09:33:08 but you could have ones with different venoms, like other spiders do 09:34:07 (btw, one other conceptual problem - right now swarms just use the cloud glyph. perhaps they could use the section symbol, that had been used for clouds but abandoned?) 09:34:40 anyways, gotta go, I have an appointment 09:34:44 what's the point in having webs spawn spider again? 09:34:58 no magic spiders out of nothing = no problem 09:35:21 so just net+alarm? 09:36:01 then make them hard to escape so spiders have the time to come eat you while you're still entangled 09:36:49 I'm thinking of more shallow water 09:37:08 whats the plan for avoiding webs btw? if you go through a corridor filled with webs, will it take forever and be really annoying? 09:37:46 will T&D help? 09:37:51 galefury: that's why I would prefer a small slowdown, like currently, with no warning at all when you walk into one 09:39:29 kilobyte: since there are already trapdoor spiders, and since the webs are a spider alarm, i don't think the spawning is necessary 09:39:51 there could be some chance to generate trapdoor spiders next to webs to create the same effect 09:40:27 galefury: T&D and also (currently) EV will both help avoid webs, and also spider form provides total immunity 09:44:02 galehar: there? 09:44:35 yes 09:45:32 can't you cut through a discovered web with ctrl+dir? (functionally identical to disarming) 09:45:53 yes also that 09:46:06 with a chance to stumble into it by accident 09:46:58 galehar: I'm doing the monk starting piety. 09:47:30 And I was going to ask something but I forgot it... 09:48:02 I just added a line to crawl_manual.txt/reST, is this the correct way to do it? 09:49:07 -!- hoody_ has quit [Remote host closed the connection] 09:50:44 Ah well I dropped a first patch, nothing implemented for ^-screen, but we have darkgray **.... before first god. 09:50:58 -!- hoody_ has joined ##crawl-dev 09:50:59 Dunno how it interacts with Xom, just realized that was an issue. 09:52:47 dpeg: Uploaded a patch for monk piety. 09:56:41 ghallberg: never edit the manual in git, it will be overwritten on next wiki sync 09:56:52 ghallberg: which is the master copy 09:59:04 -!- stabwound has joined ##crawl-dev 10:06:53 ghallberg: thanks, will look into it 10:07:21 kilobyte: So where do I edit? 10:07:51 kilobyte: being trapped in a web is interesting, though - i think shallow water is a good starting point but it could have a chance to fully entangle you (say, more likely if you are flying or tiny) 10:08:44 ghallberg: the wiki 10:08:57 Ok, I'll add it there then. 10:09:36 oh, link to the manual section for octos? 10:09:38 i can do it now 10:09:42 appt was cancelled 10:10:31 Someone added a line to the wiki apperantly, but it doesn't really say anything about the functionality. 10:39:40 -!- hoody_ has quit [Remote host closed the connection] 10:40:40 -!- hoody_ has joined ##crawl-dev 10:45:14 -!- MarvinPA_ has joined ##crawl-dev 10:49:09 -!- MarvinPA has quit [Ping timeout: 252 seconds] 10:58:19 back from child play 10:58:42 one solution to the web abuse would be that no more spiders are spawned from webs once the level's HS is dead 10:59:15 ghallberg: <3 11:01:36 -!- galehar has quit [Quit: Page closed] 11:01:47 ghallberg: still around? 11:02:08 ye 11:02:13 Kind of 11:02:26 Do you indicate the Monk's preemptive piety? 11:02:45 dpeg: There is a grayed out piety bar when you have no god. 11:02:47 I wasn't sure, but we could print **.... in the piety display, preferably in darkgrey. 11:02:57 I didnt' understand the manual-wiki thing yet though. 11:03:03 ghallberg: awesome. Does ^ say something special, too? 11:03:09 ghallberg: I can do the manual. 11:03:13 Let me do it, if you preger. 11:03:18 Nope, didn't see that note on mantis until I posted tha patch and had to run. 11:03:18 prefer, even 11:04:03 I'd be happy to learn but I'm not sure where to edit :) 11:04:08 Is it the regular badwiki? 11:05:32 no no 11:05:40 the dev wiki has the manual 11:05:53 The advantage is that everyone can go there and fix a typo. 11:06:03 The drawback is that not everyone knows this. 11:07:25 dpeg: you could still theoretically stay in a small area of the map and harvest xp until the cows come home (unless the HS is wandering around the map). of course, that's what the OOD timer is for? 11:07:48 mumra: I don't think you can. The webs are spawned anywhere. 11:08:04 but some will get spawned near you 11:08:18 mumra: I wouldn't worry too much. If players actively try to abuse, then we make the spider summoned. 11:08:36 mumra: What is important to me is that T&D helps with web detection and disarming. 11:09:06 i don't personally see a problem anyway - just stop webs spawning after monsters stop spawning (i.e. tie to OOD timer) 11:09:18 of course 11:09:32 Definitely. Plus, we can also stop spawning webs once the HS is dead. 11:09:49 anyway, T&D already helps with detect / disarm, so you got your main wish :) 11:10:18 mumra: we don't have the Thief background anymore, but an early-ish branch where T&D is actually useful sounds pretty good to me. 11:10:46 (And there is still the vague hope for the Thief god, which would get us back the Thief background. A man can have dreams...) 11:10:56 hehe 11:11:27 T&D detecting trapdoor spiders would make it even more useful 11:11:29 It's not so easy, I have the barebones but it is hard to complete the design. 11:11:45 mumra: I'd say a patch is uncontroversial.... =) 11:12:01 :) 11:12:30 but yeah it's amazing how much work and thought has gone into the existing gods 11:13:53 nah, look at Sif, almost the same old boring bookworm that she ever was 11:14:22 affects gameplay in interesting ways though, so all is well 11:14:41 not all is well, it is good enough to be used, but a far cry from perfect 11:14:50 (this goes especially for the gifts, in my opinion) 11:14:52 well, yeah, but thats true of most things in the world 11:15:01 dpeg: Ok I can't think of anything snassy to write for the manual :P 11:15:21 sure, but the good thing about an open source, open ended game like Crawl is that you can address everything :) 11:15:22 gifting spells directly to memory might be interesting 11:15:37 galefury: yes, I agree 11:15:49 and occasionally a special ("Sif's ...") book as a gift 11:15:57 ghallberg: I'll do it once the patch is in, don't worry 11:16:15 dpeg: Ok 11:16:25 * dpeg is so curious whether/how ghallberg's change will affect Mo popularity. 11:16:45 i find it's too easy to eventually find pretty much all the spells anyway; at most i'm missing only 1 or 2 handy spells by say V:8, and definitely have enough to get by 11:17:08 dpeg: I think it makes monks pretty cool :) 11:17:17 mumra: too much book loot? 11:17:19 mumra: i disagree 11:17:21 ghallberg: I dearly hope so 11:17:31 About ^ screen, I can't really think of anything to write there either... 11:17:45 just in my games, i know others disagree 11:17:45 I'm too tired and unfocused to be creative :D 11:17:47 there are so many useful spells that i cant even learn them all, and without sif im usually still missing a few i want when starting extended endgame 11:18:14 and im talking universally useful utility spells here, not stuff for actually killing monsters 11:18:33 ghallberg: "As a Monk, you will take on your first god as a particularly pious follower." ? 11:18:41 (German speaker warning etc.) 11:18:43 i've only very rarely made it to extended endgame. but in my one 15-runer, i had literally every single spell before i even did the ziggurat :) 11:18:59 mumra: and neither S/V/K as gods? 11:19:14 dpeg: "As a monk you are especially good at ass kissing" 11:20:07 -!- [1]mumra has joined ##crawl-dev 11:20:17 I agree that there are simply so many spells gained during the course of a game that going Sif for a specific one seems pointless 11:20:20 ghallberg: that's the subtext, yes. 11:20:34 Especially as Vehu and Kiku give you an excellent guaranteed package 11:21:00 hmm, i agree that it is usually easy to fill your spell slots with useful spells 11:21:22 but i often miss things i really want even far into the game 11:21:40 might be related to me really wanting lots and lots of stuff 11:22:15 some more early book loot and less late book loot might be nice 11:22:40 Book loot is fine but spells should be more evenly distributed, I think 11:22:47 Some are in 1 book, some in 4(!) 11:23:04 -!- mumra has quit [Ping timeout: 245 seconds] 11:23:04 -!- [1]mumra is now known as mumra 11:23:06 im pretty sure thats intentional, and imo there are good reasons for that 11:24:07 also not all books are equally rare 11:26:51 -!- dtsund has joined ##crawl-dev 11:28:15 on the other hand, if i don't find certain spells, it forces me to play differently and try out different spells - sif is boring because you can decide in advance exactly how to build your repertoire 11:30:15 sif can hold some spells back awfully long 11:30:34 especially the ones only found in the capstone books 11:31:31 but i agree that sif fairly reliably giving you a library of every spell in the game is bad 11:31:55 maybe ill write something about gifting memorized spells on the forums 11:32:07 there are some problems with it, but i think those can be worked out 11:33:45 anyway, later 11:35:03 -!- galefury has quit [Quit: ChatZilla 0.9.85 [Iceape 2.0.11/20110701115916]] 11:35:12 i think it would be interesting with more randart spellbooks and less fixed ones in general 11:36:49 78291 (L27 MuCK) ASSERT(feat_is_solid(feat)) in 'beam.cc' at line 1114 failed. (Zig:14) 11:40:14 -!- Textmode has joined ##crawl-dev 11:49:10 make XP "Next:" a bar as well? 11:50:57 i always loved the "oh wow, that monster gave me 12345 xp!"-effect 11:51:04 which is gone now, right? 11:52:17 Napkin: Wow that monster gave me 25% xp! 11:52:53 hence -> bar with similar previous/now display as health has :) 11:59:30 -!- upsy has joined ##crawl-dev 12:08:20 Napkin: nah, that won't happen :) 12:09:01 * Napkin played too much of brogue lately * 12:09:04 bars rule! 12:10:34 true 12:10:55 -!- Letchik has left ##crawl-dev 12:14:11 -!- Twilight13 has quit [Read error: Connection reset by peer] 12:17:49 -!- Galefury has joined ##crawl-dev 12:22:18 brogue is great, i love their staves 12:31:47 -!- hoody_ has quit [Remote host closed the connection] 12:42:34 hey guys - is "<" and ">" used in normal mode? 12:43:14 how about making it travel to the next up/downstairs automatically, if you are not standing on one? 12:49:42 good idea 12:50:05 much better than X > . 12:52:00 Napkin: like all good ideas, it might be controversial 12:52:28 I'll take up the cudgels for it. 12:53:44 -!- ainsophyao has quit [Remote host closed the connection] 12:58:34 i've more than once hit the wrong key for taking the staircase i'm on 12:58:45 like, up instead of down 13:00:33 -!- Twilight-1 has joined ##crawl-dev 13:06:04 Zannick: but can any harm come out of it? 13:06:27 dpeg: if i run across the level to a staircase I didn't want to take? 13:06:34 It is natural that every key command improvement proposal is countered with "typo". Not sure what to do about that. 13:06:47 no, i agree. 13:07:15 and for "harm" i'd obviously have to concoct a scenario 13:07:42 Zannick: the worst cannot happen, since autotravel stops for monsters. 13:07:53 Of course, if food is scarce etc. 13:08:58 so, currently shift+dir are the only single-key autotravels 13:09:06 o! 13:09:15 -!- Zaba has quit [Ping timeout: 258 seconds] 13:09:31 Tab in a sense, too 13:09:50 o is explore, not travel. similar. but anyway 13:09:57 tab is one move per, i thought? 13:10:02 yes 13:10:22 yeah, i was not including that 13:10:22 Zannick: pressing o at an inopportune moment can cause great sorrow 13:11:03 i'm more apt to hit the wrong direction as a typo than the wrong stair 13:11:27 yes, that's most common, I think 13:11:31 I do it all the time. 13:11:38 but then i'm one square off only 13:11:43 well, at most 13:11:53 which can mean death :) 13:12:06 * Zannick spent about a minute a few games ago trying to hit a iron statue 13:12:31 (which was right next to an iron golem) 13:12:33 The problem I have with tab is that it's right next to `, repeat action. 13:13:02 Hitting the wrong one is... highly undesirable if you're a caster. 13:13:17 And casters use ` all the time. 13:14:44 -!- Zaba has joined ##crawl-dev 13:14:51 dtsund: you can deprecate Tab via macro 13:14:52 -!- Zaba has quit [Changing host] 13:14:52 -!- Zaba has joined ##crawl-dev 13:15:15 True, but doing that on a character-by-character basis is an irritation 13:16:55 but....okay. i had a level with mennas that i had to flee. teleported away and got to a stair. either o or the autostairtravel from the wrong key could have me wander back over to mennas. 13:17:23 i discovered | recently 13:17:25 ; doesn't seem to be used for anything, perhaps I'll remap it in Light 13:17:30 dtsund: you can save the macro 13:17:45 dtsund: ; is used 13:17:52 so i guess if hitting the wrong stair key does something i really don't want, i'll learn to use | and be more careful 13:17:59 It is? What for? 13:18:09 ; is "what's here?" 13:19:06 /, then, perhaps 13:21:02 dtsund, changing keys because of that is futile 13:21:22 because there are way more keyboard layouts being used than the one you use 13:22:28 I'm willing to bet the majority of the player base uses this one, though 13:22:52 Might be good to take a poll somewhere, though. 13:23:04 i don't use ` actually 13:24:12 -!- dtsund has quit [Quit: Changing OS...] 13:27:08 -!- dtsund has joined ##crawl-dev 13:31:11 I'm going to make a poll about keyboard layouts on the Tavern. Any suggestions for options? 13:31:43 dtsund: it's a hopeless endavour :) 13:32:04 And I managed to change a few keys... was a tough fight every time. 13:32:52 Yeah, well, for Light, I'm Belligerent Dictator For Life, so I can just unilaterally change things. 13:32:57 Of course, when designing stuff I do it with the US layout in mind (e.g. the new drop menu commands are . and ' -- this is awkward on my German layout but should be good on the US one) 13:33:01 dtsund: of course 13:33:13 :) 13:33:53 It is interesting how much players will defend their favourite options/commands with claws and teeth, when much bigger changes are shrugged off. 13:34:05 But if I find US keyboard layout usage is only ~50% or so, I won't bother 13:34:30 it's more 13:35:00 dtsund: jpeg and I made a survey last year. You should be able to find the results somewhere... US/Canadian users dominate with more than 50%. 13:35:10 Oh, the survey was already done? 13:35:19 yes, long time ago 13:35:28 we're planning a new one 13:35:31 507k USA versus 628k non-USA visitors in apache stats 13:35:44 Napkin: perhaps because there's a German server? 13:35:56 pardon? 13:36:16 germany is just second place with 80k 13:36:35 -!- ainsophyao has joined ##crawl-dev 13:36:39 and USA includes the per minute wget for all logfiles & milestones 13:38:05 or course, there is GB (78k) and Poland (23k), where most people have a similar keyboard to USA layout 13:38:12 in non-USA 13:38:56 Napkin: can you list the first five or so countries from the 628k share? 13:39:16 and we already know Germany, UK and Poland :) 13:39:57 I bet the Finns make up the lion's share! 13:40:36 germany 80k, great britain 78k, russia 60k, canada 52k, france 33k, finland 30k, sweden 24k, australia 24k, poland 23k, italy 15k, brazil 15k, new zealand 14k, japan 14k, netherlands, 12k, spain 9k, hungary 9k, south korea 9k, etc 13:41:36 that's determined by GeoIP 13:42:10 We beat the ausseis at least. 13:42:27 Yeah, well 13:42:48 An Australian connecting to CAO/CDO has to ping all the way around the world 13:43:01 it's not about game servers 13:43:16 it's about http(s)://crawl.develz.org 13:43:29 Aha. 13:46:09 Hm. Can't find a keyboard poll. Found a different poll from about a year ago with lots of other interesting stuff, but not on keyboard usage. 13:48:58 proposal for sif gifting spells to memory instead of books posted on the forum 13:49:04 https://crawl.develz.org/tavern/viewtopic.php?f=8&t=2460 13:49:12 comments are very welcome 13:49:41 Galefury: Isn't Sif a woman? :D 13:49:43 :S 13:49:56 Galefury: Do you consider Vehumet boring? 13:50:40 Galefury: I like the propsal, cool mechanic and idea 13:51:30 "Sif is currently almost equivalent to not having a god until at least mid-lair. That's boring." <--- If you think that, you're not Channeling often enough.. 13:51:33 maybe we collected US + Can + UK + Aus in the survey 13:52:19 I told you the Finns have way more share than they ought to, given the populations :) 13:52:54 Galefury: great, will comment on it later. Child bedtime story now. 13:53:23 I think it's interesting, but I also don't think I'd take that god on a pure caster. 13:54:18 sif slots are based on regular spell slots, so the spell gifts wouldnt be very useful to a fighter 13:54:59 but yeah, i guess hybrids would get the most out of it 13:54:59 That just means it would be bad on fighters too. 13:55:07 A big reason I often take Sif on pure casters is precisely to insulate a bit against the vagaries of the RNG and its book generation 13:55:28 also generalist casters who use lots of spell schools and can fit most spells into their playstyle 13:55:39 like wizards and transmuters 13:56:33 well, im not sure how useful it would be 13:56:40 which is why i posted it on the forums 13:56:46 Yeah. 13:56:48 people can discuss and flame :) 13:57:12 I think it's an interesting idea in principle, but casters will want some way to flag a spell as "I want to keep this" 13:57:52 Suddenly losing some great spell you've come to rely on will be extremely frustrating. 13:58:12 I'm speaking as someone who's played games with a very similar mechanic before... 13:58:31 (Specifically, Final Fantasy Legend 1 and 2.) 13:58:50 hm, promoting adaptation to the available spells is a big part of the proposal 13:59:12 rare book gifts weighted towards gifted spells that were used a lot could work 13:59:30 but this would force people to dance spells they want to keep 13:59:32 which is stupid 13:59:35 That has its own issu-yeah. 14:00:26 as a caster i find myself limited by available books early on and limited by spell slots later 14:00:49 this would help with both, so i think it could be useful 14:01:34 Late-game, I find I generally have all the slots I truly need, and find myself deciding between luxury spells 14:01:39 also the spell gifts would get a pretty big success enhancement and are weighted to be powerful (related to what you could usually cast) 14:01:47 so i think they could be a pretty big bonus 14:01:51 Yeah, I saw that bit 14:02:15 The problem is not being able to take a spell into a branch and rely on still having it later 14:02:36 i avoided this on purpose 14:02:36 You could start the Crypt and lose Dispel Undead halfway through. 14:02:54 yeah, but allowing players to scum for a spell they want for a branch is bad too 14:03:21 Yeah, but as-is players might put off branches to see if they get the spell they want for it 14:03:42 Though this kind of happens anyway. 14:05:02 I think a modified version would be really sweet for a hybrid-oriented god. 14:05:49 mhh, i thought of that too 14:06:02 but sif has some problems 14:06:25 Oh, another issue 14:06:52 Seems you could put on crystal plate in anticipation of learning spells, then take it off for serious success boost 14:07:10 from 90% to 95%? 14:07:37 also uncastable spells are not gifted, so you would block off the top spell levels by doing this 14:07:47 Is it that small? I can't keep track of the armor changes. 14:08:06 might go to 100%, but 90% to 100% is not huge either 14:08:22 theres a list, ill see if i can dig it up 14:08:26 It's pretty big, both in terms of requirements and gameplay effect 14:08:52 Gameplay-wise, having a spell castable *flawlessly* opens up tactical options that didn't exist before. 14:09:20 https://spreadsheets.google.com/spreadsheet/pub?hl=en&key=0AqvhLOPFHpiMdHFWbVpnU3F6M2VlQWV5Rk56UDNaNHc&hl=en&gid=2 14:10:05 maximum spellcasting penalty from armour seems to be 205 for gda at 0 skill, but this would cripple most characters 14:10:16 and block off a good deal of powerful spells 14:10:24 probably not worth the success increase 14:10:34 Requirement-wise, success is based on (essentially) a roll of 3d100 beating some target, which means you shoot right past 50% success at blazing speed as your skills increase but take forever to reach 100%, if you ever do 14:11:18 really? 14:11:22 Yeah. 14:11:42 my information about spell success is from the wiki, so might be wrong 14:11:45 I learned this when implementing the percentage-based success indicator for Light. 'twas a bit of a pain. 14:12:40 isnt success percentage listed in wizmode? 14:12:46 Nope. 14:13:20 right, thats power 14:13:22 Might show the to-beat number, but unless I misread the code, at no point does the game know the actual success percentage. 14:13:34 Oh, yeah, it shows power. 14:18:50 back 14:20:07 not sure if you're still talking about Sif, and old but pretty vague (so far) idea was some kind of library 14:20:13 so that you don't pile up the books 14:23:01 the merge with vehu idea was better imo 14:23:38 cannot go wrong with deicide 14:24:43 we need more ingame-killable gods 14:25:05 Eronarn: for some reasons, many players don't like the idea that Jiyva can die. 14:25:21 dpeg: probably because they don't like the idea of losing access to the god? 14:25:28 They even accused me of having it stolen from some kind of books I don't read. :) 14:25:37 Discworld? 14:25:43 yes, Pratchett 14:25:59 Eronarn: no no, they say this doesn't fit Crawl, which really only says how they see their Crawl 14:26:35 One of the Discworld books is about a god and his last worshipper, with the god afraid of the possibility of fading away 14:26:39 dtsund: The point is that seeing gods come and go, and measuring their power by number of worshippers is absolutely natural if you're an atheist who is interested in religions. What else could you do than see it as the life cycle of gods? 14:27:04 dtsund: this much I've heard, but it wasn't the reason why Jiyva can die :) 14:27:15 Heh 14:27:40 I suspect that may have been Pratchett's inspiration, too 14:28:10 I did read The Portable Atheist by Christopher Hitchens, for example (after Jiyva). In the introduction he hammers exactly this point home, but as I'm saying, it is not at all revolutionary. Standard fare, I'd say. 14:28:33 -!- purge has joined ##crawl-dev 14:28:35 yes, of course -- I am not saying that Pratchett did it wrong, I wonder why players are so short-sighted at times 14:28:52 Actual atheists are rare on the Disc, because the gods tend to break their windows. 14:28:54 Eronarn: the next natural candidate would be Beogh -- no orcs, no god 14:29:29 doesnt really make much sense though, hes only available to orcs 14:29:38 so theres going to be at least one left 14:29:39 so? 14:29:55 who's the last orc standing? 14:30:01 on the other hand, demigods should be worshippable ;) 14:30:21 also it would not be very relevant, beogh is useless without orcs anyway 14:30:43 so there is no gameplay effect of killing him when most enemy orcs are dead 14:30:55 so? 14:31:07 not everything is about mechanics 14:31:57 Zannick: there happens to be an Implementable on that =) 14:32:09 Galefury: cannot go wrong with deicide 14:32:09 theres millions of orcs outside the dungeon, most of which probably worship beogh 14:32:12 dpeg: that would explain why i think that :) 14:32:17 must have read that one 14:32:20 so i dont see it being good flavor either 14:32:21 dpeg: would be interesting to make lugonu killable. turns the abyss into a completely empty, itemless void except for portals 14:32:27 Would be fun if the morgue listed "killed gods" right after "killed monsters" 14:32:40 not my little Lugonu baby!! 14:35:13 Eronarn: That might make the Abyss slightly more boring, though. 14:35:53 i've thought about suggesting a different portal area like that. just an empty plane. 14:36:00 that reminds me, i need to check out the new abyss 14:36:03 it's in trunk, right? 14:36:13 How'd the Abyss change? 14:36:22 years ago, I made a rough plan for an after game (when you've won), called Twilight of the Idols, where the gods form factions and battle each other -- you're just a pawn in this game 14:37:16 dtsund: new map generation, morphing 14:39:07 there'll be more minor gods in the future, I hope ... definitely wanna see a spider god 14:39:32 since we get the branch for 0.10, there's nothing to stop us from having the god in 0.17 14:39:50 holy shit this is awesome 14:39:56 ? 14:39:59 new abyss 14:40:10 Banishment was never more fun! 14:40:14 Do we get wizmode on servers? 14:40:21 not sure 14:40:21 Or will I have to compile it myself? 14:40:34 It is all part of Lugonu's subscription plan. 14:41:00 I wish daftfad was still around. We made a pretty elaborate design for random/player gods. I am sure it could be finalised into something decent. 14:41:03 "For just $12 a year, you too can have wizmode!" 14:41:33 I heard that other roguelikes do something like this. If we do too, count me out. 14:41:58 -!- galehar has joined ##crawl-dev 14:42:09 dpeg: it would be cool to have minor gods, which are not as full fleshed, but also aren't bad wrath 14:42:12 there's a rather severe bug in the Abyss: you can force a shift by save+load, need either bmh or to learn his stuff 14:42:19 Fie, no wizmode. 14:42:20 and maybe push back the 'full' gods to later in the game 14:42:32 (like an exaggerated version of overflow) 14:42:44 Eronarn: what about pushing Demigods first? 14:42:56 Eronarn: well, that goes in the direction of having several gods at once (Powder-style), which is awesome but just not possible in Crawl. 14:42:59 kilobyte: i would like to see demigods made a bit more interesting, yes 14:43:07 dpeg: oh, i mean so you can swap gods, but cannot take more than one 14:43:12 ah 14:43:20 so you choose a minor god first, and a major one later 14:43:28 yes (if you want). or just don't bother with a minor god 14:43:30 that conflicts with making the game smaller, though :) 14:43:39 very tiny gods 14:43:42 kilobyte: maybe saving abyssal_state would fix this. I plan to do this when I introduce variable speed 14:43:50 kilobyte: could you do me the favour and revise ghallberg's Monk patch. 14:43:55 it would be a good place for testing to see whether a god could reach 'full' status, actually 14:44:00 like stealth god, or gold god, or ... 14:44:06 Oh, I guess I could explore the Abyss with an AK. Derp. 14:44:10 I might forget to update the manual otherwise (and I want to do it only when/if the patch is in). 14:44:28 Eronarn: ah, like portal vaults as testing grounds for branches 14:44:32 yeah 14:44:39 'shrine' rather than altar, say 14:45:02 kilobyte: and that was meant as a question, missed the ? 14:45:09 dpeg: in a bit, sorry 14:45:13 perhaps you find on average shrines before the ET, and altars after, though overlapping of course 14:45:16 Oh, mapping with maprot. 14:45:20 That's new. 14:45:58 item mimics native depth is 12. Feature mimics depth is 15. Since I'm completely changing generation, I need a hard limit for shallower depth. Is 7 for items and 10 for feature ok? 14:46:48 kilobyte: no sweat 14:47:07 galehar: yes 14:47:28 galehar: generation chance could depend on depth, too. 14:47:49 I can autoexplore Abyss now? 14:48:26 Galefury: replied 14:48:37 dtsund: THere is not wizmode on CAO, I cannot speak to CDO 14:48:37 dtsund: yes, an unexpected (but not undesirable) consequence of allowing mapping. 14:48:55 might still not be a good idea 14:49:04 rax: I tried both, no dice. 14:49:10 but that goes for all dangerous areas, the Abyss being no exception 14:49:20 Naturally. 14:49:22 dpeg: I'm not sure about increasing chance with depth. We don't want the lower dungeon to be full of mimics. 14:49:28 I always thought that we have wizmode on CDO. Perhaps only for devs? :) 14:49:34 it's enabled for devs, yeah 14:49:35 galehar: hey, I liked exactly that idea 14:49:49 Oh, _that's_ why there's that change in dgamelaunch :P 14:50:20 MarvinPA_: once the web traps are properly, I'll start talking about weapon traps. Perhpas we can make progress there. 14:50:43 Why don't the peons get wizmode? Could just mark those games as forfeits... 14:51:11 First and second class citizens, a time honoured and foolproof system. 14:51:57 Or have there be a wizrobin account 14:52:08 (no idea why, but there might be something... for example folks messing up with insanely large files?) 14:52:34 dpeg: I thought about the hive race. How about having eggs that hatch into larva? Should make it easier to control the timing of the race. 14:53:16 galehar: can do that... but controlling placement of larvae vs food would also do this 14:53:22 the egg timer could be randomly set based on distance to the entry. That way, vault design is much simpler. 14:53:35 sounds reasonable, yes 14:53:57 also, egg -> larva -> bee is much cooler, of course 14:54:27 galehar: if you feel like it, add a note to the Mantis item 14:54:28 if it's one big level, placing the larvas at increasing distance to the food the farther from the entry is cumbersome 14:54:35 ok, I'll add it 14:54:38 yes 14:54:42 thank you 14:55:03 galehar: during my sad tournament games, I made an observation regarding auto mode. Now or later? 14:55:17 shoot 14:55:42 As you know, I don't bother with manual mode, being the lazy snot that I am. 14:56:06 I do turn off skills, though. Now it often happened that some skills were turned on, but had a 0% rate. 14:57:04 In all my post-midgame adventures, this forced me to go manual. As I see it, if a skill is On in auto, then I want some training there. Of course, I don't care about precise numbers, because auto. What if On skills in auto get a minimal, guaranteed training rate like 5%? 14:57:52 Was the 0% skill a skill you never used? 14:57:54 (The 5% might actually be a function of number-of-active-skills.) 14:58:29 dtsund: for example Fighting on a pure caster. The 0% is absolutely correct, but the game can infer from me having the skill active that I want some training there. 14:58:30 oh, I noticed that too. The problem is that disabled skills are not put in the queue. So when you enable them back, they start over at 0 14:58:57 galehar: a guaranteed training rate would bypass this. 14:58:59 Ah, yeah. 14:59:27 I plan to fix it by having a separate queue for tracking disabled skills so when you enable them back they automatically get a training percentage based on recent usage. 14:59:42 galehar: but there was no usage for Fighting :) 15:00:10 you can use focus if you want to force a skill with low training. It get around 20% minimum/ 15:00:27 but it does not amplify 0% skills, does it? 15:00:31 maybe also train fighting by taking damage? because that is what it most often is wanted for? 15:00:36 dpeg: yes it does 15:00:43 ah, I missed that 15:00:45 phone 15:00:58 -!- syllogism has quit [] 15:01:03 Galefury: being attacked already train dodging/armour/shields 15:01:20 why does that matter? 15:01:51 hmm. I don't know :P 15:04:52 -!- hoody_ has joined ##crawl-dev 15:28:21 dtsund: Reasoning for not giving wizmode out in like 2007 or whatever was that it could lead to accidentally screwing up someone's config or files, or filling disk, or otherwise breaking the server. It probably wouldn't at this point, but I don't see a compelling reason to make it available still. 15:31:14 -!- greatzebu has joined ##crawl-dev 15:31:18 When you escape from the Abyss, you should return to where you came from. (https://crawl.develz.org/mantis/view.php?id=4471) by GreatZebu 15:31:51 woo 15:32:16 MarvinPA_: ? 15:32:23 that patch 15:32:27 :-) 15:32:28 greatzebu: You have 1 message. Use !messages to read it. 15:32:33 !messages 15:32:33 (1/1) Napkin said (7h 44m 8s ago): hey there - motivated to build some 0.10 packages for osx which we could host on CDO? 15:32:46 Any comments on the monk stuff? 15:33:15 Oh, I thought from the title you meant return to the exact place where you were banished. 15:34:10 possibly the increase in pan exit gate frequency doesn't need to be that high, since i added a chance of them showing up in pan lord vaults too 15:34:31 but increasing it a bit is probably still good 15:34:40 greatzebu: is it really necessary to save all this stuff (level description, abreviation,...) 15:34:51 can't it be regenerated when you come back? 15:34:54 yeah, very possibly a lower chance is somewhat better 15:34:55 Perhaps it could be initially smallish, but increase with extended stays? 15:35:27 also, the data isn't saved at all 15:35:45 galehar: very possible, but I didn't see a way of accomplishing that. someone who's more familiar with the codebase could probably clean it up 15:36:01 yeah, I'd go with a lower increase for the exit, but a higher chance for an exit in Pan vaults 15:37:46 one chance in 20-25 might be good 15:38:06 that would give an exit every 7-10 maps on average, plus vault exits 15:38:34 anywyay, thanks. It's a good start. 15:39:44 great, glad to help out 15:41:42 I found the stairs and map loading stuff is surprisingly complicated, so I am probably doing things in a suboptimal way, but it can always be improved now 15:45:54 -!- monqy has joined ##crawl-dev 15:46:46 indeed, the down_stairs function is insane. 15:51:34 is the thing to do for saving the data to add marshalling and unmarshalling calls in tags.cc? 15:54:10 Yep. 15:54:15 Breaks save compat, of course 15:54:44 And I'm pretty sure the marshall calls and corresponding unmarshall calls have to be in the same order. 15:55:01 yes they do 15:55:04 right, that makes sense 15:55:11 save compat can be handled. You need to bump the minor version 15:55:20 look in tags-version.h 15:55:31 is that all you have to do? very convenient 15:55:40 and docs/develop/save_compatibility.txt 15:55:46 ok, thanks 15:55:52 And yeah, I remember the horror that is down_stairs 15:56:23 haha, yeah it is a lot to wade through 15:57:01 TAG_MINOR_MONS_THREAT_LEVEL is a very simple example of how to add something to be saved while preserving save compatibility 15:58:37 ah, I see 15:58:44 ok, should be no problem 16:11:06 would it be better to keep the abyss data in the props hash table and avoid the save compatibility issue? 16:15:38 -!- ais523 has joined ##crawl-dev 16:17:23 -!- greatzebu has quit [Quit: greatzebu] 16:20:48 hmm 16:20:59 no matter how long you spend in the abyss, it always keeps the same tileset 16:21:13 it's chosen randomly at banishment and never changes 16:21:35 would be a lot more interesting if the tiles randomly changed as the abyss did 16:21:40 -!- greatzebu has joined ##crawl-dev 16:22:39 mumra: hmm, good idea. It could be changed when the abyss shifts 16:23:25 it's kind of the one thing that *isn't* random ;) 16:24:14 greatzebu: about keeping the abyss data in props, why not. It's not really a problem to save it, but then saving abyss state is useless when you're not there. 16:25:49 better to avoid the extra clutter in player.h and just keep in props, I agree 16:26:04 props is limited to 255 key/value pairs so we shouldn't put anything there either. 16:26:30 Basically, anything that is only temporary has a good reason to be in props. Abyss state is. 16:28:25 also, global variables don't need to be moved into player.h to be saved. They can be declared in tags.cc with extern, like stair_level and the others from dgn-overview.cc 16:28:42 although this is a bit ugly 16:28:58 would be nice to gather all of them into a Dungeon class 16:31:38 Auto travel exclusions in Abyss persist and wrap around map (https://crawl.develz.org/mantis/view.php?id=4472) by mumra 16:36:38 new abyss is fantastic, anyway 16:36:40 High CPU usage near Gnoll castle (https://crawl.develz.org/mantis/view.php?id=4473) by SamB 16:36:45 will be great with variable speed 16:40:47 03galehar 07stone_soup-0.9 * rb0cec6ae74bb 10/crawl-ref/ (10 files in 2 dirs): Add a new "show runes" command: "}" 16:40:57 03galehar 07mimics * r49ea4c671f2a 10/crawl-ref/source/dungeon.cc: Fix branch entries placed for mimicing at a depth too shallow for mimics. 16:40:57 03galehar 07mimics * rc8826eb8e39f 10/crawl-ref/source/dungeon.cc: Tweak the staircase mimic generation to be less predictable. 16:45:54 galehar: in 0.9? 16:46:46 -!- elliptic has joined ##crawl-dev 16:48:43 kilobyte: are you doing anything else on webs right now? i noticed a couple of things that should improve web generation 16:49:40 mumra: nothing this evening 16:49:42 mumra: the webs are very pretty, also 16:49:48 thanks! 16:50:02 kilobyte: i'll make a patch, i realised the main thing that's wrong with the weighting algorithm 16:50:22 orthogonals just need a higher weight than diagonals, should push things into corners much better 16:50:25 for example, it hardly ever makes that 200 traps... which is a good thing 16:51:18 kilobyte: yes it's a bit unpredictable, but i just played a few levels and it actually seems about right 16:51:18 Why does spl-clouds.cc spell beam 'beem'? 16:53:36 why does it spell spell 'spl'? 16:54:58 spl is understandable as an abbreviation 16:58:30 hey, does unarming traps actually still train the traps&doors skill once you got the skill itself? 16:59:16 in automatic mode only 16:59:30 oh? 17:00:10 i thought the difference between automatic and manual mode is the different %-Train only.. i guess that's wrong? 17:00:31 Napkin: I'd be happy to make some 0.10 mac binaries for cdo. We should really update the 0.8 stable binaries to 0.9 as well 17:00:51 automatic mode assigns the %-Train according to what you do -- like, in this case, disarming traps 17:01:34 so.. i have manual mode and 9% train on traps&doors 17:01:48 but that doesn't matter 17:02:31 so.. actually.. does it mean i should switch to automatic mode every time i disarm a trap, so the traps&doors skill gets raised. and then back to manual because i prefer it? 17:03:05 greatzebu: that would be awesome :) have you looked inside an old osx package to see the structure and what files are placed where? 17:03:53 yeah, I am familiar with putting bundles together so I think it'll be pretty simple. Maybe we should even make a nice .dmg 17:04:29 whatever you prefer and is more easily used and can be used by the most players 17:05:22 Napkin: there's no benefit in ever going to automatic if you want to micromanage things 17:06:08 kilobyte: but you just told me, that in automatic mode traps&doors is trained when i disarm traps - while it is not the case in manual mode - or did i just misunderstand you? 17:06:18 it's not trained 17:06:26 probably a .zip is best to start off with since it's easiest and it's been distributed that way in the past. dmg can probably provide a nicer experience for users though 17:06:26 the weighting is just increased 17:06:39 Napkin: in automatic mode the amount by which it's trained goes up; in manual mode, it stays being trained at whatever percentage you gave 17:06:41 but you can adjust the weighting yourself however you like in manual mode 17:06:43 MarvinPA_: yeah, which causes the skill to be trained 17:06:47 in each case, it only gets the skill points when you kill something 17:06:53 right but it's an important distinction :P 17:06:56 haha, kilobyte 17:07:40 greatzebu: you have created a package already, haven't you? could you link me please? 17:07:48 -!- hoody_ has quit [Remote host closed the connection] 17:08:34 I have, but I used a new SDK so people running Leopard can't use it. 17:08:36 it's at http://charm.cs.uiuc.edu/~abecker/crawl/ 17:08:54 I can install the older SDK for maximum compatibility though 17:09:01 that would be great 17:10:14 actually it would probably be pretty easy to write a little script that will produce a mac build every night from the current development head if you can scoop up the resulting binaries on cdo 17:10:49 greatzebu: what you should leave out of the package: docs/obsolete, docs/template 17:11:24 ok, easy enough 17:11:40 Napkin: "make install" should do everything 17:11:51 and if it doesn't, please let me know 17:12:09 it doesn't 17:12:22 look at current stone_soup-0.9.0-win32.zip 17:12:35 readme.txt is missing, manual is missing 17:13:03 i'm trying to find a release file which had all the right files, for a proper list 17:13:52 -!- Chousuke has quit [Ping timeout: 264 seconds] 17:14:25 -!- Chousuke has joined ##crawl-dev 17:14:30 main folder (beside the crawl binary) should contain: README.txt, LICENCE.txt, CREDITS.txt 17:15:31 the manual (automatically created from the wiki page) needs to be downloaded and put into docs-folder: http://crawl.develz.org/other/manual.txt 17:16:02 i think we refrained from including PDF versions, didn't we? 17:16:09 greatzebu: are you still reading this? 17:16:16 yes 17:16:26 no one seems to bother with PDFs... 17:16:35 great, makes the release smaller 17:16:49 alternatively we have html: http://crawl.develz.org/other/manual.html 17:16:57 Napkin: downloading the manual is wrong... it will give you an incorrect version 17:17:15 well, there is no other choice right now 17:17:24 when building, say, 0.9.1, you don't want octopodes or Spider Nest listed 17:17:28 unless you get a version of the wiki page from the release day 17:17:34 eh? It should work correctly. 17:17:43 is something wrong? 17:17:47 hmm? 17:17:51 what do you mean? 17:18:00 (unless your scripts overwrite the manual with something you wget) 17:18:08 pardon? 17:18:27 the manual in REST format is in the wiki 17:18:45 on release day this needs to be included in formated form in the release 17:18:49 there is a manual shipped in text format... needed for example for '?' in-game 17:18:51 or did that change again? 17:19:54 someone needs to manually type 'make rest' just before release -- and preferably from time to time, which syncs the rest version in git 17:19:58 when pressing "?" you get readme.txt first 17:20:17 -!- Pacra has joined ##crawl-dev 17:20:23 or is that ?V ? 17:21:02 well, if you add a rest target to makefile and don't tell me, i will still give out old information, of course :) 17:21:22 you are not supposed to ever run the rest target 17:21:32 and last i heard was that manual.txt was supposed to be totally deleted from git (to cause less confusion) 17:21:50 it is automatically built when compiling 17:21:54 -!- Wensley has joined ##crawl-dev 17:22:01 interesting, kilobyte :) 17:22:39 the problem with the copy on the wiki is that it's not versioned 17:22:46 anyways, those text-files i mentioned are missing from release zip 17:23:03 this means, after a release the stable branch diverges 17:23:05 don't tell me - i never liked the idea 17:23:14 but have to live with it ;) 17:23:41 greatzebu: so, remove the pdfs and htmls from docs/ as well, i guess 17:23:43 yeah... I'd personally prefer the master copy to be in git 17:24:18 it surely was no fun to create the haxy rest2text cronjob 17:24:45 -!- evilmike has joined ##crawl-dev 17:24:56 er, what? You did that? 17:24:58 and those 3 main info txt files beside the crawl-binary 17:25:01 i think that's it 17:25:15 if you did, nice duplicated work... 17:25:17 i had to, because all rest2text tools i was told to test failed 17:26:02 I wrote one too... 17:26:14 you have my sympathies ;) 17:26:32 i actually just run elinks on the html version - hence -> haxy ;) 17:26:42 after some grepping and sedding 17:26:52 another issue was that there's no python on most platforms by default, and it's not something that can be sanely shipped in contribs 17:27:25 greatzebu: i have some scp space you could move your packages too, which would then be moved via cronjob to the /trunk/ website - if that's ok with you? 17:27:39 haha @ python in contribs :D 17:28:27 but.. in general.. please don't do releases when i'm on vacations? :> 17:28:47 just too stressy to catch up afterwards ;) 17:29:24 hey kilobyte, since you just mentioned python 17:29:58 in case you are terribly bored.. could you add some logic to the scoring scripts about how to link the morgues for 0.9 of cdo? :> 17:30:41 Napkin: I'm not sure I follow what you mean by moving my packages 17:30:45 -!- hoody_ has joined ##crawl-dev 17:30:48 sorry, no :p 17:31:30 if version = 0.9 and game older then 19. August 20:20 CEST -> morges in trunk folder 17:31:36 napkin: should just be a matter of checking whether the start time was before or after the appropriate time... that's how I got the tourney scripts to handle it 17:31:37 yeah 17:31:43 let's say I barely know enough about python to insult it :p 17:31:43 if version = 0.9 and game younger than 19. August 20:20 CEST -> morges in 0.9 folder 17:31:47 elliptic! 17:32:06 hah! 17:32:07 i was looking for you already :) 17:32:23 python is great :-) 17:32:33 just sayin' 17:32:41 greatzebu: with moving i meant "uploading your packages to CDO" 17:33:15 ok, how would I go about uploading? 17:33:28 napkin: just got back to crawl after being busy with travel+conference+being stranded by a hurricane after conference 17:33:38 elliptic: yeah, i could add that check.. if it was perl or shell - but i don't intend to touch python until i'm 40 ;> 17:33:44 welcome back then! :) 17:34:02 i'll bug you again about that morgue logic later this week - agreed? 17:34:59 greatzebu: i need your public ssh key (per email to napkin@develz.org) and then you can scp or sftp the packages, once i sent the username and target directory via reply - if you are ok with that? 17:35:20 sure, I can send it your way shortly 17:35:39 napkin: yeah, I'll try to remember to take a look at it later :) I'm not really that familiar with the scoring scripts but it shouldn't be hard 17:36:07 greatzebu: don't hurry - 0:35 am here, about to hit the pillow soon 17:37:01 elliptic: awesome :) i managed to get cao's 0.9 games counted in the official stats, but that's about it i could do 17:38:05 napkin: yeah, I saw that got fixed :) especially nice that now http://crawl.akrasiac.org/scoring/per-day.html should be correct 17:38:25 yes, sir 17:38:35 that was quite an adventure, though :D 17:39:40 -!- hoody_ has quit [Remote host closed the connection] 17:42:46 g'night o/ 17:42:58 night! 17:43:10 'night 17:44:39 night 17:45:59 -!- Garhauk has joined ##crawl-dev 17:57:05 ellllliptic! 17:57:08 We had a tournament! 17:57:22 yes! 17:58:06 -!- dalaing has joined ##crawl-dev 17:58:12 -!- dalaing has left ##crawl-dev 17:58:26 -!- galehar has quit [Remote host closed the connection] 18:02:02 elliptic: welcome back 18:02:57 -!- upsy has quit [Quit: Leaving] 18:04:43 elliptic: I am itching to talk about tournament scoring rules, but maybe later when you're properly settled. 18:05:20 I'd be very interested in comments, yes... I only really saw how things were working in the first half of the tournament 18:05:35 elliptic: I think the rules worked out marvellously. 18:05:42 Awesome job. 18:06:03 there are a few banners that didn't work very well (nobody got them), but I think I'm pretty happy with the main rules on the whole 18:06:41 Of course, no-one, and definitely not I, can leave good alone, so there are more ideas. First, I think we really should have a god stock market like we did for species and background. And perhaps the weights can be increased a little. 18:06:42 I think in the past you were talking about ideas for making the tourney more interesting for players who don't get as far in the game? 18:06:51 yes, comes in a moment :) 18:07:12 It was very instructive to see a number of Kenku getting started when they were winless and giving >100 bonus points. 18:07:36 god stock market could be good but is a bit trickier to set up 18:07:43 because god isn't fixed over the course of the game 18:07:48 yes, I know 18:08:12 Would "taken as first god and brought to full piety" do? 18:08:40 The other obvious candidate would be "only god for the game", but that is less interesting. 18:08:42 it would be really hard/annoying to get jiyva that way 18:09:02 you could scum for an early altar or go athiest until finding one, i guess 18:09:02 -!- greatzebu_ has joined ##crawl-dev 18:09:02 evilmike: that is only one god 18:09:02 "first god brought to full piety" is what was being used for the god points in this tourney 18:09:09 -!- greatzebu has quit [Ping timeout: 252 seconds] 18:09:09 -!- greatzebu_ is now known as greatzebu 18:09:10 yes 18:09:29 it has issues with xom also... for xom, I insisted that players start as CK and never abandon xom 18:09:35 I don't think it's worth worrying about banners nobody gets; /dev/null traditionally had trophies that they didn't expect anyone to get (although someone did in the end), and junethack had several that nobody got 18:09:36 but in general something like that could work 18:09:48 elliptic: another one (to be used in addition to stock market, in my opinion): give every player (winning or not) the maximal piety/200 points for every god. I.e. up to 20 points for each god. 18:10:57 elliptic: of course, one can say it's our fault, but some gods are badly underplayed. And some of those actually have the power, they're just unpopular, e.g. Fedhas and Elyvilon. A stock market would see them get used for the appropriate builds (no challenge games required). 18:11:30 giving points for piety reached with each god even for non-winners would be good for players who aren't winning, certainly 18:11:37 The other piety/10 (piety/20 was a typo) rule means that non-winners can get some points from spreading worship as well. 18:11:50 yep 18:12:02 elliptic: by the way, I couldn't figure out who got the 100 points for winning game started last. 18:12:09 elynae did 18:12:17 started how much before deadline? 18:12:18 !lg * t won max=rstart 18:12:19 232. Elynae the Sabretooth (L25 FeBe), worshipper of Trog, escaped with the Orb and 3 runes on 2011-08-28, with 1348856 points after 100755 turns and 3:11:16. 18:12:22 !lg * t won max=rstart x=rstart 18:12:23 232. [rstart=2011-08-28 [20110728182022S]] Elynae the Sabretooth (L25 FeBe), worshipper of Trog, escaped with the Orb and 3 runes on 2011-08-28, with 1348856 points after 100755 turns and 3:11:16. 18:12:33 5h 40m, I think 18:12:39 awesome, I love it 18:13:37 The devnull tournament has some trophies for non-winners (I forget their names, but effectively for reaching branches etc.). jpeg suggested some "awards" specifically for non-winners. 18:13:57 !tell galehar thanks for the abyss patch feedback, I've added a patch to address the issues we discussed 18:13:57 greatzebu: OK, I'll let galehar know. 18:14:15 wensley suggested that each god should give two banners... one directed at non-winners and one at winners 18:14:29 Yes, for example. 18:14:46 devnull was {metal-type} star for reaching branch ends or doing the ritual or whatnot 18:14:47 some trophies for things like reaching lair, reaching vaults, etc could also be good 18:15:04 I wouldn't worry about unused banners, too much. If we get better ideas, we'll replace 'em, otherwise players have to man up. 18:15:09 elliptic: exactly 18:15:38 i like how each god has a banner. i also liked, in the may tourney, that each *branch* had its own banner and special goal 18:15:49 Perhaps some banner stunts should actually give points. Kikubaaqudgha's did, but 40 points is pretty lousy :) 18:16:29 Perhaps the first winner to take a Nemelex choice should get some bonus points, depending for how old the choice is. 18:16:42 One point per hour or so :) 18:17:21 maybe, yes... I didn't really want the nemelex choices to turn into speed events though 18:17:28 but a small bonus could be fun 18:17:36 elliptic: they were taken pretty quickly anyway 18:18:16 yes, they went a little more quickly than I expected even 18:18:20 But all of these apart from god stock exchange and more non-winner gimmicks is pretty minor. Did you see the usage curve? 18:18:32 usage curve? 18:18:47 do you mean the server activity page? 18:19:02 http://crawl.akrasiac.org/scoring/per-day.html 18:19:16 greatzebu: does you patch also address portal vaults? 18:19:36 dpeg: yes 18:19:47 although portal vaults in pan still don't work 18:19:47 cool 18:19:53 that is okay 18:19:56 they shouldn't 18:19:57 outside of zigs anyway 18:20:06 yeah, I'm pretty happy with how many players we got in the end... new record for number of distinct players winning in a tournament, for one thing 18:20:23 elliptic: and the tournament was shabbily advertised 18:20:33 almost as bad as the release :) 18:20:50 yes, and webtiles users couldn't play until nearly halfway through the tourney 18:21:30 any thoughts on the tourney length and/or timing for future tourneys? 18:21:32 After this 3 hour phone call, I have to decide between evilmike's vaults or a tournament summary note for the blog. 18:21:40 elliptic: I think it is good. 18:22:22 Can you guys help me by collecting the relevant data: total numbers of participants, wins, winning players, games, players of webtiles (if possible)? 18:22:59 sure, I'll collect data like that and e-mail it to you 18:23:04 awesome 18:23:15 if you can think of other interesting data, please add 18:23:32 and if you'd add the !lm commands, I will use them next time, promised! 18:23:36 except webtiles-specific data will take a little work 18:23:48 !lg * tiles 18:23:49 Malformed argument: tiles 18:24:21 webtiles milestones and games should be marked now, but sequell doesn't know about it 18:24:51 elliptic: I was looking at the jiyva banner and was fairly pleased that there were several players who got *only* that banner, which means that it succeeded on some small scale at motivating low-level players to try a lot of different characters. wish I knew how many low-level players went for it but never got it... 18:25:10 jiyva banner is definitely cool 18:25:41 -!- valrus has joined ##crawl-dev 18:25:46 The Sigmund banner was too technical, in my opinion. Did anyone catch it? 18:25:48 !lg * t xl>=9 x=cdist(crace),cdist(cls) 18:25:49 3971 games for * (t xl>=9): cdist(crace)=24; cdist(cls)=27 18:26:13 !lg * t xl>=9 s=name x=cdist(crace),cdist(cls) 18:26:13 3971 games for * (t xl>=9): 38x Elynae [18;19], 34x Neil [15;17], 21x theglow [15;16], 31x ionfrigate [13;13], 20x Kamina [13;12], 17x jeanjacques [12;13], 27x Lawman0 [12;12], 18x ophanim [12;11], 13x mikee [12;11], 15x Wolfechu [12;8], 27x Palladion [11;12], 12x patrick [11;11], 31x BirdoPrey [11;10], 15x Shador [11;10], 29x sstrickl [11;10], 13x swamp [10;12], 21x nmf [10;12], 16x heteroy [10;1... 18:26:30 I think a few people got it, yeah 18:26:30 dpeg: quite a few people did 18:26:34 dpeg: catch it how? as in recognize you needed to drag sigmund into the temple? 18:26:40 actually do it 18:26:44 !lm * t uniq=sigmund place=temple 18:26:50 38. [2011-08-27] Joehai the Digger (L6 TrEE) killed Sigmund on turn 3921. (Temple) 18:26:55 38 people :) 18:26:56 !lm * t uniq=sigmund place=temple 1 18:26:57 1. [2011-08-13] Thornet the Ruffian (L8 DrMo) killed Sigmund on turn 7234. (Temple) 18:26:59 !lm * t uniq=sigmund place=temple 2 18:26:59 2. [2011-08-13] jokeserver the Covered (L9 MiFi) killed Sigmund on turn 8330. (Temple) 18:27:00 okay okay, so everyone got it 18:27:24 there were people who got only that banner 18:27:55 I didn't say anything! 18:27:56 !lm * t uniq=sigmund place=temple xl>9 18:27:56 No milestones for * (t uniq=sigmund place=temple xl>9). 18:28:53 !lm * t uniq=sigmund place=temple s=name ?:N>1 18:28:53 38 milestones for * (t uniq=sigmund place=temple): 18:29:55 btw, another way to measure "went with god Goo" is to ask for "had god Goo while collecting N runes and got full piety" -- not sure that helps, though 18:30:38 to more formally propose my "wo banners per god" idea, to avoid cluttering player pages with tons of little banners (especially for milestone-ish banners that are awarded for doing normal, easy tasks), the high-level banner for each god should be a superset of the low-level banner for each god. so, say, tso's minor banner would be granted for getting one rune, and his major banner would be... 18:30:40 ...granted for getting all runes, and if you got both, on your player page it would only show the major banner, because it implies the minor banner 18:30:49 s/wo/two 18:31:02 wensley: that would be ideal, yes 18:31:12 Wensley: makes a lot of sense 18:31:30 could even have more than two, if we can come up with ideas 18:31:56 ...which would be easy for collecting runes 18:32:00 !lm * t uniq=boris s=gid gid!= ?: N>=5 18:32:01 385 milestones for * (t uniq=boris gid!=): 5x 1220729, 5x 1222941 18:32:29 banner idea: most boris kills in one game 18:33:40 grinding alert! 18:33:43 perhaps a banner for killing the most uniques in the shortest game 18:33:54 Wensley: what function to use? 18:34:06 I don't think banners with "most" in them are a good idea generally 18:34:08 total turns for a winning game / number of uniques killed 18:34:17 elliptic: yes, I think I agree 18:34:27 it shouldn't normally be possible for players to prevent others from getting a banner 18:34:35 except for beogh :P 18:34:35 ahem, Beogh :) 18:34:38 anyone who kills boris 2 or more times would be more in the spirit of a banner 18:34:40 beogh breaks all the rules! 18:34:52 saint was like that this tourney but that's sort of a special case, because it is within a clan 18:34:55 erm, not banner 18:34:57 * Zannick confused 18:34:57 and you choose your clanmates :P 18:34:59 points 18:35:06 recognition 18:35:09 etc. 18:35:16 I really like the saint banner 18:35:22 valrus: yes, well, that was a perfectly good banner in 2009 and 2010 and could certainly be used again :) 18:35:27 oh 18:35:29 from hearsay it worked quite well at encouraging intra-clan competition 18:35:30 the boris 2 or more times was a banner some time ago 18:35:32 I didn't recall 18:35:32 it certainly did in my clan 18:35:52 in my clan i was like "i'm not even going to pretend to try..." 18:36:00 evilmike: you're telling me 18:36:03 haha 18:36:04 the_glow managed to almost get 1 win every day 18:36:20 The Nemelex banner can be easily scaled. A first version would be "get an open Nemelex Choice character a rune" 18:36:23 the_glow was a crawl-playing machine :P 18:36:36 !lg * t s=name x=sum(dur) 18:36:37 * due doesn't even know that name :S 18:36:37 37593 games for * (t): 38x theglow [592659], 33x jeanjacques [471545], 423x Neil [433195], 189x nfogravity [414693], 119x Mussorgsky [413086], 284x Omrii [408544], 13x mikee [391699], 130x Tenaya [385988], 130x qtip [345915], 66x ophanim [336326], 293x Elynae [324982], 134x JohnAsti [318132], 94x IonFrigate [317970], 222x soul [312191], 189x Leibowitz [311208], 215x lerugray [310232], 117x OneEyed... 18:36:42 I've never really cared about the banners 18:36:47 yes... on SA he said "I look forward to not playing Crawl anymore." :) 18:36:51 I could have gotten semi-close to the saint banner if I had won my last game of the tournament without doing extended, but it seemed a shame to waste such a good musu :P 18:36:52 maybe if they were all worth points 18:37:10 on IRC he also said that he was done 18:37:12 and then got 4 more wins 18:37:17 haha 18:37:27 valrus: yes, I think they should... maybe that's against the original purpose, though 18:37:31 he averaged more than 10 hours of crawl-playing a day 18:37:37 I enjoy how the last three tournaments have all had different winners 18:37:42 He'll be back in a year or two. 18:37:56 i came very close to getting the slow and steady badge for a very rushed game on the last day that ended in a zot:5 splat 18:37:58 elliptic: but he used them pretty efficiently. 18:38:30 yep, he is a great player also... lots of streaking and good choice of races/classes 18:38:34 yes 18:38:44 zipcode pulled some amazing stunts, I'll mentioned them in the wrap-up 18:38:53 also kilobyte: I'm pretty sure python *is* included in most distros nowadays, at least all the ones I've heard of (i.e. newb-friendly ones) :P 18:39:00 the impressive thing is that he only beat mikee_ by a bit despite playing for 50% more duration 18:39:23 dpeg: yes, 78291's DDHe speedrun was great 18:39:58 mention how many combos were removed from oldwon thanks to nemelex's choice :) 18:40:14 wensley: 28 :) most people won't know what oldwon is, though 18:40:28 (also most of those weren't nemelex choice, of course) 18:41:39 dpeg: do you want the statistics from the 2010 tourney to compare with? 18:41:55 looking at the server activity page, what was the reason for that huge spike in early june? 18:42:28 oh wait, perhaps the axis is mislabeled 18:42:51 the spike is the may tourney 18:43:38 elliptic: yes, that'd be awesome 18:43:40 i wonder if there's a separate spike for when webtiles went online during the tournament 18:43:47 can't really tell from the graphs on cao 18:43:56 oh duh, webtiles is cdo, never mind 18:44:06 evilmike: it went online on 8/19 18:44:13 the CAO graphs count both servers 18:44:18 hmm, wonder what they're using to generate those graphs... 18:44:20 so not much of a spike 18:44:33 elliptic: but webtiles provided a fair number of players... 18:44:50 kilobyte: some patches for you (on mantis issue) 18:44:59 Wensley: Linux distros -- partially but yeah. FreeBSD, OS X, Windows (cygwin, msys) -- no. 18:45:29 dpeg: made webs generate out of LOS, increase with depth, improved edge placement 18:45:45 !lg * t rend>20110720 rend<20110721 x=cdist(name) 18:45:46 2102 games for * (t rend>20110720 rend<20110721): cdist(name)=395 18:45:55 !lg * rend>20110720 rend<20110721 x=cdist(name) 18:45:59 2237 games for * (rend>20110720 rend<20110721): cdist(name)=424 18:46:34 dpeg: oh, I figured it out... the webtiles spike on that page is hardly there at all because it includes non-tourney games 18:46:36 !lg * cv>=0.10 x=v 18:46:37 1225304. [v=0.9.0-a] TheWanderer the Sneak (L1 HuWn), got out of the dungeon alive. on 2011-08-31, with 50 points after 50 turns and 0:01:02. 18:46:42 :( 18:46:54 so before webtiles was updated to 0.9, it was still including people playing 0.8 webtiles 18:47:34 and there probably weren't nearly so many people who didn't play at all because of the 0.9 webtiles delay... instead they probably played 0.9 console or 0.8 webtiles 18:48:04 -!- greatzebu has quit [Quit: greatzebu] 18:48:05 !lg * t rend>20110717 rend<20110718 x=cdist(name) 18:48:06 1664 games for * (t rend>20110717 rend<20110718): cdist(name)=257 18:48:35 but we went from 257 distinct players playing tourney games on Aug 17 to 395 playing tourney games on Aug 20 18:50:32 mumra: <3 18:51:21 elliptic: so we can still gauge now many players switched from 0.8 webtiles to tournament on Aug 19? 18:51:29 ah, thanks :) 18:52:04 mumra: what are the current web generation chances on Spider:1 and Spider:5? 18:52:29 -!- hoody_ has joined ##crawl-dev 18:56:09 dpeg: it's total # of webs generated = ( 300 / (6-Depth) ) +/- 33% 18:56:32 approximately 18:58:07 this is number of webs at level creation? 18:58:17 yes 18:58:33 i forgot to adjust ones created during play, they'll always be at the same rate 18:58:36 and generation rate during play is constant? 18:58:40 :) 18:59:16 easy enough to do something similar there tho :) 19:00:13 sure 19:01:03 !tell galehar Let us ignore the comments on rP food from the forum, I don't think they help. We can go without them at first, but I think eggs are good. 19:01:03 dpeg: OK, I'll let galehar know. 19:01:33 eggs wat 19:03:44 certain spider eggs 19:04:03 hm 19:04:03 valrus: Swamp provides a way to get rP via swamp dragons. 19:04:41 what does that have to do with eggs? are people proposing that spider eggs give rP? 19:04:44 Perhaps we want something similar for Spider. It's not clear... but if so, one idea is to have a perishable item (hence the egg) whose consumption gives you rP for a number of turns, i.e. proactively in contrast with the potion. 19:04:59 that reminds me, think I will code up a Moth of Frailty with an rPois- gazing attack, for the spider branch 19:05:00 I don't think people do swamp for rP 19:05:13 elliptic: it is always mentioned as an option 19:05:23 elliptic: might be more common with the upcoming rune lock. 19:05:24 some people like swamp dragon armour, but not for the rP 19:05:33 dpeg: when the spider eggs perish, do they hatch into spiders? 19:05:46 galehar is obsessed with eggs, he also proposed to use eggs -> larva -> bee for new Hive. :) 19:05:58 you aren't even very likely to get swamp dragon armour before the swamp:5 vault 19:06:09 mumra: I don't think the player qualifies as a breeder. 19:06:10 he's just eggcited about them 19:06:14 Although: "Mommy!" 19:06:26 I'll... just go 19:06:36 spot the reference? 19:07:00 elliptic: there was a proposal to have more swamp dragons, or perhaps one guaranteed, on Swamp:1. 19:07:10 dpeg: Gleep? 19:07:18 bah guaranteed swamp dragon on swamp 1 19:07:24 kilobyte: yes? 19:07:32 dpeg: that might be a decent idea if there is actually going to be a rune lock, yes 19:07:33 presumably the eggs have already been fertilised ... could be a nice choice, rP or (possibly) friendly spiders? 19:07:35 make one guaranteed somewhere in swamp, but leave its location random 19:07:40 hide isn't guaranteed even then, nor enchant armour 19:07:46 Perhaps such an item is not necessary, but then again, it might be 19:07:47 dpeg: was it too obscure? :p 19:07:53 yes, only 1/3 chance of hide per dragon 19:07:57 kilobyte: yes! 19:08:04 mumra: i don't think this is a good choice, it amounts to 'rPois, or if you already have it, spiders' 19:08:12 you will basically never want to choose spiders if you don't have rPois 19:08:15 yeah probably 19:08:17 Having to force an ascension kit strikes me as terrible Nethacky design anyhow 19:08:24 Eronarn: yes 19:08:29 RichardHawk: huh? 19:08:36 tome actually has a spider egg item, which summons spiders when used 19:08:36 I'd rather see rPois be less important, really 19:08:40 so we've been beaten 19:08:42 elliptic: <3 for mail 19:08:55 dpeg: I mean, look three lines below "Mommy" in Nethack sources :p 19:08:57 elliptic: there is also the idea to have it provide only 90% resistance. 19:09:05 kilobyte: you're the master! 19:09:22 dpeg: The only reason to make sda guaranteed would be because the game balance desperately needs it - and in that case something is broken on a more fundamental level in my opinion 19:09:23 rPois already isn't a big deal in swamp if you have clarity or can kill swamp drakes quick 19:09:30 yes, I think rpois would be better if one level did not guarantee immunity 19:09:33 the main reason for rPois in swamp is swamp drakes, they can be a bit hard to evade 19:09:34 You can do the game without rPois fine as is 19:09:39 either add more levels, or remove immunity 19:09:46 what if their breath only followed a straight line, like catoblepas do? 19:09:47 Wensley: ah, the old idea of three pip rP 19:09:54 could be two 19:10:19 I'm just saying, I am dissatisfied with the status quo :) 19:10:22 technically you can also do the game fine without rf or rmut 19:10:36 rPois is overrated by most players currently because you can put off swamp/snake forever and it really isn't useful elsewhere 19:10:37 just that some important places get very painful 19:10:39 Zannick: well, if I can choose, I'd take the rP-less game. 19:10:45 elliptic: rune lock! 19:10:46 and making it less good will just make it even weaker elsewhere 19:11:12 rpois is a good find early game, for snakes, and it's good for melee characters in hive 19:11:24 but yeah, could be made more interesting 19:11:35 oh wait, I've been meaning to make a poison/disease hell, "malebolge" 19:11:52 stealth zin buff 19:12:03 !lg * t won god=zin 19:12:03 3. PigVomit the Axe Maniac (L27 HOFi), worshipper of Zin, escaped with the Orb and 15 runes on 2011-08-27, with 11193131 points after 146070 turns and 14:45:21. 19:12:10 dpeg: on further consideration, I don't think SDA is the solution at all... since making it more common would also make rPois more important for swamp 19:12:22 even in Zot, a single green drac can royally screw you up 19:12:25 kilobyte: no 19:12:28 that's quite rare, though 19:12:32 by then you'll have 480395290 !healing 19:12:37 kilobyte: less than any other single dracs can royally screw you up 19:12:48 since it can't item destroy or banish or summon death drakes or... 19:13:06 <|amethyst> but alistair's takes care of all the other draconians 19:13:06 -!- Mu_ has quit [Quit: Defecator, may everything turn out okay so that you can leave this place.] 19:13:13 elliptic: okay, so I've been misinformed 19:13:29 I have a feeling Alistair's is seriously broken. 19:13:32 it is 19:13:37 it is against the things it's useful against 19:13:43 elliptic: white dracs are there to take care of !healing :) 19:13:45 I didn't start using it until relatively recently actually, but it is really good 19:14:08 back in my day we used alistair's to clear elf 7 19:14:18 I was sad when I discovered that my mummy could not use alistair's to trivialize elven halls, so I had to use freeze instead and probably had a lot more fun 19:14:19 still works just as well on elf as ever 19:14:28 Should unleash MarvinPA_ on it. 19:14:38 s/freeze/refrigeration 19:14:42 out of Zot, though, rP doesn't matter once you're done with Snake/Hive 19:14:50 monsters should build up liquor tolerance so you can't reuse it on them 19:14:52 alistair's is incredibly good on zot:5 19:15:03 yeah it gets orb guardians nice and drunk 19:15:10 yes, even the berserk ones 19:15:16 well, now that we're aware of the problem, we will think about solutions and choose one, right? 19:15:26 rage > drunk 19:15:33 how often does alistair's cause int loss? 19:15:37 very rarely 19:15:50 more rarely than xxx or tentacle? 19:16:01 yes 19:16:02 without sustab you might lose a few points of int if you spam it on all 5 levels of elf 19:16:07 rarer than XXX, at least 19:16:10 and by a few points i mean not enough to remember 19:16:10 not sure about tentacle 19:16:20 xxx and tentacle are the same 19:16:23 gotta do evilmike's vaults and the announcement -- please file if there's good Alistair nerf ideas 19:16:24 best to nerf alistair's by making it non-guaranteed 19:16:40 or just nerf it in trunk :) 19:16:50 if it's not guaranteed, what the difference between it and mass confusion? 19:16:51 elliptic: in general i don't think we should have many guaranteed effects 19:16:55 Wensley: schools 19:16:56 eronarn: yes 19:17:05 it's fine to have spells that are similar except for schools 19:17:08 <|amethyst> maybe base the chance on HD rather than MR 19:17:15 pah, nerfing poison magic and transmutations 19:17:42 dpeg: let me know if you want any other tourney stats... if they aren't webtiles, I can probably produce them 19:17:57 I agree that it's fine to have spells from different schools be similar. what I'm asking is why a generalist would choose alistair's over mass confusion 19:18:07 they wouldn't? 19:18:08 dpeg: I'll poke at getting webtiles stats, though 19:18:13 what's wrong with alistair's being used by poisonmancers or transmuters 19:18:13 <|amethyst> lower level 19:18:24 alistair's is much better, although mass confusion is decent. it's hard to use now that it's a hex though 19:18:28 both of them have limited options for crowd control 19:18:29 Raise the level to make it harder to access for generalists 19:18:35 poison magic has meph cloud 19:18:59 yes, and that's all it gets 19:19:00 having it be resistible by HD or just random chance whether it succeeds or not sounds like a decent possible nerf 19:19:12 Meph isn't even really poison 19:19:15 just random chance IMO 19:19:16 You can float it with just Conj 19:19:26 i like random chance too, yeah 19:19:35 chance scaling with your spell power, perhaps 19:19:38 yes, it would be nice if the mechanics would be different, at least, even if the effects are similar 19:19:42 one issue with random chance is that peopl will still cast it against e.g. antaeus 19:19:45 elliptic: stuff like streaks, best clans etc. I get from the web. Can you tell me how many players got at least one rune? 19:20:03 i think making it feel more like a 'party' spell would be good 19:20:04 Oh, and I'd love a god list among winning games. 19:20:22 !lm * t11 type=rune x=cdist(name) 19:20:29 !lm * t10 type=rune x=cdist(name) 19:20:31 2370 milestones for * (t11 type=rune): cdist(name)=264 19:20:31 3072 milestones for * (t10 type=rune): cdist(name)=229 19:20:32 perhaps if a creature resists alistair's, it would get a reduced chance to resist it again, as the alcohol builds up in its brain 19:20:36 like - what about adding a drunk status effect, where stuff is neutral and partying, rather than confused? 19:20:42 elliptic: busy vaulting, will you send another mail, please? 19:20:48 dpeg: sure 19:20:57 -!- MarvinPA_ is now known as MarvinPA 19:21:02 Eronarn: interesting 19:21:03 the disco ball in the pan disco hall should confer the drunk status effect 19:21:06 yes 19:21:25 not sure what drunk would do to the player 19:21:33 maybe it's just confusion on the player? 19:21:38 Does anyone know who made the megacathedral vault? 19:21:53 dpeg: hi ;) 19:22:00 anyways, if you hit a drunk person, they should remember that they're supposed to be fighting you 19:22:00 Eronarn: You have 34 messages. Use !messages to read them. 19:22:02 due: I expected that much :) 19:22:14 dpeg: not my best vault, but the first i'd done in aaages 19:22:21 perhaps it does something similar to mislead, except instead of making a monster seem that it is a random monster, it makes them seem as though they are a monster of the same genus but of a lower HD 19:22:33 Wensley: nah, we don't need mislead 2 19:22:41 Wensley: we really don't need more interface fuckery :) 19:22:43 pandemonic boogaloo 19:23:12 Though, I would like a status effect that reversed all text! 19:23:14 I think the interface fuckery is lovely, if used sparingly 19:23:26 due: upside down and backwards! 19:23:46 (does xom flip levels by 180 deg currently? he totally should) 19:24:27 or even just reversing monster names. 19:24:29 or scrambling them! 19:24:34 Eronarn: an oversight in doing that caused one of the hardest-to-pin-down bugs in UnNetHack which lost people their games 19:24:38 "You see here a dlobok". 19:24:48 so be very careful when implementing a level-flip routine in a roguelike not designed for it 19:25:41 due: y s hr kbld 19:26:11 ais523: losing is fun! 19:26:28 could have different levels of severity, start it off with dysgraphia:: you see here a kodolb 19:26:37 Eronarn: losing due to game-breaking bugs isn't, though 19:26:43 unless the entire purpose of the game is to try to not trigger them 19:26:50 * dpeg could use a script that rotates text blocks by 90 degrees. 19:26:56 and even then, games designed around them (usually ROMhacks) are typically accused of not being fun 19:27:18 dpeg: do you need me to whip one up for you? it shouldn't be too hard, I hope 19:27:43 -!- GreatZebu has joined ##crawl-dev 19:27:46 dpeg: I had one of those. 19:28:02 -!- valrus has quit [Remote host closed the connection] 19:28:09 dpeg: But there's an ASCII art designer which lets you do that, too; I was using it to make vaults for a while but it wasn't as good as I'd hoped. 19:28:14 ais523: if you can.... My perl skills are too limited (perhaps not right tool). The reason is that for some reason, almost all vaults we get have height > width. This wastes lines and makes .des files less readable. 19:28:36 I was going to write it in Perl 19:29:01 I'd praise AceHack night and day, and tell everyone it's the Real Nethack! 19:29:02 note that a square vault will have height > width in almost all fonts... 19:29:16 evilmike: I mean letter count. 19:32:12 evilmike told me that coloured floor in console leads to traps of the same colour, instead of their original colour. Is there a quick fix to this, or should I report on Mantis? 19:32:24 wow, almost got it right first time 19:32:28 I was just missing a final newline 19:32:30 let me pastebin it 19:32:38 maybe it was fixed already, i based that on what i observed in 0.8 with the crypt vault i made 19:32:39 !won * t11 19:32:40 * (t11) has won 232 times in 37593 games (0.62%): 8xSpEn 7xMDFi 7xMiBe 6xDDNe 6xDEWz 6xKoBe 6xVpAE 5xCeWz 5xHOAr 5xHaVM 5xKoSk 5xMfIE 5xOgCj 5xTrAs 4xDECj 3xDsVM 3xGhMo 3xHECj 3xHEWz 3xMfGl 3xNaTm 3xSpDK 2xCeMo 2xDDEE 2xDDEn 2xDDHe 2xDDIE 2xDEHu 2xDrTm 2xHENe 2xHOBe 2xHOPr 2xHaHu 2xHuEE 2xKoGl 2xMDBe 2xMDDK 2xMuSu 2xMuWz 2xSEEE 2xSEGl 2xTrBe 1xCeFE 1xCeGl 1xCeSu 1xDDBe 1xDDHu 1xDEFE 1xDgAE 1xDgCj ... 19:32:41 ais523: many thanks 19:32:42 !won * t10 19:32:44 * (t10) has won 321 times in 58428 games (0.55%): 49xSpEn 23xDDNe 22xMfCr 18xGhFE 18xHaTm 17xSEAs 9xMfIE 7xHEWz 7xHaBe 6xHECr 5xDDCK 4xDECj 4xDSNe 4xHEIE 4xMfTm 4xSpVM 3xDEWz 3xDSAs 3xDSBe 3xDSVM 3xHEFE 3xKoCr 3xMDEn 3xSEIE 3xSpSt 2xDEFE 2xDEIE 2xDGRe 2xDSFi 2xDSIE 2xDSMo 2xDSWz 2xKoBe 2xMDFi 2xMfCK 2xMfFE 2xNaFE 2xNaIE 2xNaPr 2xNaWz 2xOgHu 2xTrAE 1xCeAM 1xCeHe 1xDDAr 1xDDBe 1xDDCr 1xDDHe 1xDESt 1... 19:33:30 http://pastie.org/2462945 19:33:40 also, it surprised me by rotating anticlockwise rather than clockwise 19:34:01 interesting to note the combo specialisation in t10 compared to t11 19:34:12 ais523: I apply it by running rotate.pl on a text file? 19:34:25 dpeg: it'll take from a file on the command line if there is one, or stdin 19:34:25 ais523: yes, I shall point it out 19:34:26 ais523: i got seriously messed up earlier when i was working with polar coordinates 19:34:27 and output to stdout 19:34:34 i forgot that 0 deg starts from the x axis :( 19:34:45 if you really need a clockwise rotation, just run three times 19:34:50 ais523: cool. You have no idea how often I've done this by hand. Want an entry in the credits? 19:34:57 you can use perl rotate.pl -i file.txt to rotate in place 19:34:58 ais523: I know this :) 19:35:11 dpeg: for Crawl? I think I already have one 19:35:12 I wrote two lines 19:35:21 (on 3*90=270, not on -i) 19:35:29 ais523: okay, but I spent hours rotating by hand 19:36:58 if there's a different credits entry, that's fine 19:37:13 I've already tested on a file with ragged line lengths, so I hope it handles that case correctly; if you find bugs, let me know 19:37:42 will do! 19:37:53 ais523: and we have only one credits file 19:39:19 err, may be perl -i rotate.pl file.txt 19:39:22 I hardly ever use that option 19:41:07 -!- ainsophyao has quit [Remote host closed the connection] 19:42:07 -!- hoody_ has quit [Remote host closed the connection] 19:42:46 urgh, perl is messy 19:42:58 * Zannick trying to fix a henzell bug 19:49:27 Perl can be messy, or clear; it gives people a lot of leeway to make it clear, and that gives equal leeway to make it confusing 19:50:18 -!- st__ has quit [Ping timeout: 240 seconds] 19:50:30 well, there are lines that go simply 19:50:35 s/something/something else/; 19:50:49 and i understand what that should do 19:51:23 it's just annoying to have to search for what string that is operating on 19:52:04 -!- Mu_ has joined ##crawl-dev 19:52:19 -!- Galefury has quit [Read error: Connection reset by peer] 20:00:42 i think the biggest challenge will be testing my changes to the function 20:01:13 i might just copy the relevant section and test that separately from the module itself 20:02:50 zannick: which bug are you trying to fix? 20:03:25 elliptic: ckiller issues. 20:03:37 ouch 20:03:48 after 0.8, kills were credited to eg. "sigmund the orange rat" instead of "sigmund, the orange rat" 20:04:02 and the uniques list was untouched 20:04:37 so every polyed unique, wiglaf, terpsichore, and beogh's named orcs all have kills credited as ckiller=a pandemonium lord 20:05:18 this is also silly 20:05:19 !lg * killer=~, x=ckiller 20:05:19 204. [ckiller=a rod of warding {god gift, !a}] xomscumming the Skirmisher (L4 MuCK), worshipper of Xom, slain by a rod of warding {god gift, !a} (summoned by Xom) on D:2 on 2011-08-26, with 161 points after 2243 turns and 0:02:00. 20:05:22 -!- hoody_ has joined ##crawl-dev 20:05:47 Did zipcode really kill only 64 monsters himself in this game? 20:05:49 http://crawl.develz.org/morgues/trunk/78291/morgue-78291-20110819-045810.txt 20:06:48 Or is it 427? I am confused. 20:07:02 dancing weapons should probably have their own ckiller, but i'm not going to touch that yet 20:09:41 dpeg: yes, 64 20:10:48 far from a record, but still cool 20:11:10 !log rob ddhe won 20:11:10 1. by, XL23 DDHe, T:92134: http://crawl.akrasiac.org/rawdata/by/morgue-by-20090919-165918.txt 20:12:22 -!- hoody_ has quit [Remote host closed the connection] 20:14:27 -!- hoody_ has joined ##crawl-dev 20:14:29 there's room for improvement there 20:17:38 is there any way to ensure spaces are not trimmed in html ? 20:18:26 dpeg: try a 
 tag
20:21:49 -!- Twinge has joined ##crawl-dev
20:21:53  works, thanks!
20:23:54  http://crawl.develz.org/wordpress/august-2011-tournament-summary
20:24:01  tell me if something crucial is missing
20:24:29  evilmike: vaults done, but pushing has to wait until tomorrow, 3.24 am
20:24:30  sorry
20:27:26 -!- zpmorgan has quit [Ping timeout: 240 seconds]
20:30:11 -!- ainsophyao has joined ##crawl-dev
20:30:16  Was there mac-specific stuff removed between 0.8.0 and 0.8.1?
20:31:25  dpeg: looks good
20:31:40  cool
20:32:11  not sure if this should be posted somewhere else, I will not post anything on r.g.r.misc anymore, that's just pointless
20:33:07  dpeg: I think it is funny how you corrected the non-standard spelling in "wucad mu-tang clan aint nuthin ta fuck wit"
20:33:08  :)
20:33:32  yeah, you should fix the spelling of that clan name
20:33:55  damn, my slang skills are not so good
20:35:41  it's a pretty long title for a silly pun
20:37:35  I'd expect nothing less from the SA clan :P
20:39:29  :)
20:39:40  bye
20:39:42 -!- dpeg has quit [Quit: zzz]
20:55:09 -!- Wensley has quit [Ping timeout: 245 seconds]
21:07:16  Is there a particular reason why you.friendly_pickup resets to default when entering a new level?
21:14:39  i always wondered about that actually, yeah
21:14:43  it's kinda weird
22:03:57  My best guess would be that it's done for characters who set it to all to hoover up the loot on a cleared floor for their followers but forget to reset it
22:04:26  But I think I might just take it out; it's unintuitive that it keeps resetting like that, and players may want their choice to persist
22:15:12 -!- Mu_ has quit [Quit: Defecator, may everything turn out okay so that you can leave this place.]
22:33:37 -!- hoody__ has joined ##crawl-dev
22:33:47 -!- hoody_ has quit [Read error: No route to host]
22:47:04  hm.
22:47:07  !lg * killer=koronagh x=ckiller
22:47:08  1. [ckiller=a pandemonium lord] tapicell the Believer (L8 HOPr), worshipper of Beogh, slain by Koronagh (a +0,+0 orcish halberd) on D:6 on 2009-06-05, with 729 points after 4560 turns and 0:16:45.
22:49:19  ah, 0.4.5
22:53:08  well, i think kills by named orcs include the monster class, so i've covered that case...
22:53:34  otherwise, henzell would need a list of known orc names if we wanted to not give ckiller pan lord credit...
23:12:53 -!- MarvinPA_ has joined ##crawl-dev
23:15:42 -!- MarvinPA has quit [Ping timeout: 250 seconds]
23:15:52 -!- MarvinPA_ has quit [Client Quit]
23:21:12 -!- hoody__ has quit [Read error: Connection reset by peer]
23:21:16 -!- hoody_ has joined ##crawl-dev
23:23:33  this was a lot easier than i thought
23:25:09  the hard part now is making a patch and a pull request
23:25:47 * dtsund shakes fist at git
23:35:34  !lm * type=god.maxpiety x=noun
23:35:35  2238. [2011-09-01] [noun=Okawaru] oxeimon the Grappler (L13 TrMo) became the Champion of Okawaru on turn 18379. (D:9)
23:48:04 -!- ussdefiant has quit [Ping timeout: 264 seconds]
23:50:52 -!- dtsund has quit [Quit: dtsund]