00:45:43 -!- casmith789 has joined ##crawl-dev 00:46:17 -!- Rejuxst has quit [Remote host closed the connection] 01:29:44 -!- dpeg has joined ##crawl-dev 01:33:45 -!- upsy has joined ##crawl-dev 01:47:07 moin 01:52:25 -!- ophanim has quit [Read error: Connection reset by peer] 01:59:53 -!- hoody_ has quit [Remote host closed the connection] 02:15:13 -!- upsy has quit [Quit: Leaving] 02:36:32 -!- galehar has joined ##crawl-dev 02:36:51 hi 02:36:52 galehar: You have 2 messages. Use !messages to read them. 02:36:56 !messages 02:36:57 (1/2) dpeg said (10h 23m 51s ago): Salut! Some results of a discussion about Ash with elliptic: (1) reskilling xp penalty can go back to 10%, we think. (2) Fighting can be included in reskilling, with the recent nerf. 02:37:07 !messages 02:37:08 (1/1) dpeg said (10h 22m 23s ago): Lest we forget: elliptic thinks skill boost at low levels is still too good. Finally, what about making skill boost less dependent on full cursage (more/only on specific cursed gear)? See you! 02:40:47 FR: Combine Ash and TSO so we can have Cursades. 02:54:34 -!- hoody_ has joined ##crawl-dev 02:59:06 -!- casmith789 has quit [Read error: Connection reset by peer] 03:11:16 -!- hoody_ has quit [Remote host closed the connection] 03:16:59 -!- elliptic has quit [Ping timeout: 258 seconds] 03:19:22 -!- hoody_ has joined ##crawl-dev 03:39:24 Keskitalo: <3 03:58:11 03MarvinPA * re0b4a38d298a 10/crawl-ref/source/mon-pick.h: Remove some references to a non-existent branch 03:58:23 03MarvinPA * r124da3843d68 10/crawl-ref/source/newgame.cc: Allow octopodes to be displayed on the species selection menu 03:58:23 03MarvinPA * rfcbb59b81361 10/crawl-ref/source/ (branch-data.h mon-pick.cc mon-pick.h): Give the Vestibule of Hell its own monster set 04:02:57 -!- syllogism has joined ##crawl-dev 04:07:09 -!- monqy has quit [Quit: hello] 04:07:41 -!- Twinge has joined ##crawl-dev 04:09:38 -!- galehar has quit [Ping timeout: 252 seconds] 04:10:24 monsters get *1.5 damage while berserk, players 12 slaying 04:10:37 it might be a good idea to make this consistent 04:11:38 -!- Twinge_ has quit [Ping timeout: 250 seconds] 04:12:08 03MarvinPA * rd69fde2fec4e 10/crawl-ref/source/chardump.cc: Fix misaligned header in notes (#4119) 04:12:30 MarvinPA: what's the problem with the selection menu? It worked for me. 04:13:04 hm, it did? there wasn't enough space for octopodes, it seemed like 04:13:29 3 columns of 8, with felids at the end of the last column 04:14:06 I had four columns, both in console and in tiles 04:14:27 with only octopodes in the fourth one which looked ugly, but they're not for 0.9 anyway 04:15:52 is that with a wider terminal than default? i can't see that 4 columns would fit, i had 3 columns in tiles too 04:17:09 oh, that's possible 04:18:29 right, it's terminal size 04:18:33 -!- galehar has joined ##crawl-dev 04:35:37 -!- edlothiol has joined ##crawl-dev 04:43:33 -!- hoody_ has quit [Remote host closed the connection] 04:49:41 -!- evilmike has quit [] 04:54:07 03MarvinPA * r210540b74d9d 10/crawl-ref/source/dat/descript/branches.txt: Tweak branch description wording 05:09:08 -!- edlothiol has quit [Quit: edlothiol] 05:09:28 -!- edlothiol has joined ##crawl-dev 06:51:28 so, how about merging brendan's new abyss layout in trunk? 06:53:37 yes! 06:53:42 -!- petete has joined ##crawl-dev 07:06:58 galehar: from initfile.cc: // NOTE: Tiles relies on the IBM character set for evaluating glyphs of magic mapped dungeon cells. 07:07:01 is that still true? 07:07:30 I have no idea. 07:07:57 Let me take a look... 07:08:59 doesn't seem like it to me 07:12:21 tilepick.cc:524, seems this comment is obsolete 07:13:15 kilobyte has done a lot of work with the charset, maybe is more knowledgeable about that 07:13:40 you can also change or remove it and see if it breaks anything 07:13:44 ah, good 07:13:48 yes, I'll try that 07:13:56 thanks 07:17:14 edlothiol: native tiles still can't show characters above 255 07:17:24 (255, not 126 like regular ASCII) 07:17:41 lots of the "IBM charset" is above that range though 07:18:16 to have proper support for the rest, we'd have to either use a massive font or use system fonts 07:18:50 the latter would require using a library like fontconfig... this was in my plans, but I won't make it for 0.9 07:19:52 not a big loss since it matters only in three cases: 1. character notes/inscriptions, 2. messages about files (like init.txt) being in directiories with non-ASCII paths, 3. translations 07:20:25 ok... well this is more for webtiles anyway -- if we want to output both tiles and curses at the same time, it would be bad to have the IBM charset hardcoded (as it currently is for tiles) 07:20:48 1. is not important, 2. is a bug but without much impact, 3. won't happen until 0.10 at the least 07:21:05 we could just plain out remove that thing 07:21:21 IBM is no better than Unicode here 07:25:17 ok, so how about we merge webtiles? It would be nice to have a trunk build for webtiles 07:26:04 and outputting both tiles and curses would allow us both way spectating, right? That would be great. 07:26:19 galehar: yeah, that's kind of the plan 07:26:43 just waiting for a certain person to ok it 07:28:25 for the merge, give me a week or so, then I'll be ready ;) 07:29:56 * kilobyte hoped for "in 7 seconds" ;) 07:30:26 all right, that sounds great :) 07:32:09 03kilobyte * r90d015526134 10/crawl-ref/source/ (monster.cc ouch.cc): Cap Injury Mirror damage to hp the victim had. 07:32:09 03kilobyte * r7ea3a18a2b1b 10/crawl-ref/source/main.cc: Revert "Fix doubled berserk penalties." 07:32:10 03kilobyte * r4f6a1759538d 10/crawl-ref/source/ (defines.h misc.cc monster.cc player.cc): Reduce berserk speed-up to *1.5. 07:32:10 03kilobyte * r0398e8d17302 10/crawl-ref/source/newgame.cc: Make the species selection menu look better in no-octopode builds. 07:32:10 03kilobyte * r43a085464fbe 10/crawl-ref/source/ (makeitem.cc makeitem.h): Get rid of a "BCR acquirement hack", whatever it is. 07:32:11 03kilobyte * r55b460201728 10/crawl-ref/source/acquire.cc: Make Trog give only 1/3 as many artefacts as before, twice as much antimagic. 07:32:11 03kilobyte * r9b6252b8e1c6 10/crawl-ref/source/ (itemprop.cc player.cc player.h tag-version.h tags.cc): Record what misc items have been seen. 07:32:11 03kilobyte * r8f2c1472f777 10/crawl-ref/source/ (evoke.cc itemname.cc): Mark a few of misc items "useless" when appropriate. 07:32:12 03kilobyte * r9320de653f54 10/crawl-ref/source/ (acquire.cc itemprop-enum.h): Improve misc acquirement. 07:32:19 03kilobyte * r77a078e98b7d 10/crawl-ref/source/itemprop.cc: Count obvious but "unidentified" misc items as seen. 07:32:19 03kilobyte * r7638acb84bfd 10/crawl-ref/source/decks.cc: Use symbolic names for deck ranges. 07:32:20 03kilobyte * rb8177409a1a3 10/crawl-ref/source/ (19 files): Don't require ", c1" in COMPILE_CHECK(). 07:32:20 03kilobyte * rf1b5ad20f35e 10/crawl-ref/source/ (aptitudes.h misc.cc player.cc): Make felids lose a level on death, step down gain of lives, undo aptitude nerfs. 07:32:20 03kilobyte * r1ae53f3915ea 10/crawl-ref/source/initfile.cc: Get rid of hardcoding of CSET_IBM in tiles. 07:32:20 03kilobyte * r0ada8313a747 10/crawl-ref/source/ (dat/descript/gods.txt mon-stuff.cc religion.cc): Nerf TSO's and Makhleb's HP and MP on kills. 07:33:54 galehar: for Trog, I merely reduced chances of artefacts like you suggested instead of removing them at all; let's see if that's enough 07:34:20 I'm a bit afraid this might lead to longer scumming until you get something very good 07:38:57 thanks! 07:39:21 about felids, how about dropping inventory on death? 07:39:32 it strenghten the "new life" flavour 07:39:52 it shouldn't cripple them since they don't rely on inventory that much 07:40:26 no rune or orb stealing and die on 07:41:18 -!- ais523 has joined ##crawl-dev 07:42:06 03kilobyte * rb194622bb53a 10/crawl-ref/source/misc.cc: Kill drained and Borgnjored kitties instantly if maxhp <= 0. 07:43:23 problems: 1. Abyss (I have an idea for that), 2. portal vaults, 3. any place that requires levitation/etc (including portal vaults) 07:43:48 I cannot attack friendly monsters with ctrl+direction (https://crawl.develz.org/mantis/view.php?id=4120) by moekitten 07:47:51 hmm. right. 07:49:46 for 3, moving you to a place reachable from stairs would fix being trapped but not your gear being unreachable 07:50:06 03dolorous * rb018639871f9 10/crawl-ref/source/ouch.cc: Comment fix. 07:51:37 losing all food means it's random whether you'd starve before finding a meat ration (in non-living branches, especially Pan and the Abyss) 07:55:07 -!- Cryp71c has joined ##crawl-dev 07:56:19 -!- ais523 has quit [Remote host closed the connection] 07:57:37 -!- ais523 has joined ##crawl-dev 08:00:05 at the very least, we can drop runes and orb 08:02:10 -!- ais523_ has joined ##crawl-dev 08:02:28 -!- ais523 has quit [Disconnected by services] 08:02:30 -!- ais523_ is now known as ais523 08:03:39 still not in the Abyss :( 08:10:27 it probably wouldn't hurt Crawl at all if non-living branches had a few living monsters now and again too 08:10:33 to help out characters who depended on them 08:11:13 there've been plans for adding things with corpses to hell and pan for a while, yeah 08:11:36 it'd make sense for demonspawn to leave corpses, for instance 08:11:39 i vaguely started working on making monster demonspawn for pan a while ago but didn't get very far 08:11:43 indeed 08:11:46 due to not being true demons 08:31:36 Lots of commits recently 08:44:13 -!- edlothiol has quit [Ping timeout: 250 seconds] 08:46:08 03dolorous * r399be8c5bae0 10/crawl-ref/source/ (item_use.cc player.cc): Generalize a tentacles check, and clean up maximum mutation level checks. 08:49:17 hrm, old powerpc problems raise their ugly head. I happily removed a workaround since gcc-4.4 is no longer the compiler used on Debian, but now people complain about backports not building :( 08:49:46 -!- edlothiol has joined ##crawl-dev 08:51:05 kilobyte: you mentionned goodness in you mingw recently. Is that with msysgit 1.7.4? 08:55:19 no idea 09:01:17 -!- elliptic has joined ##crawl-dev 09:10:50 -!- edlothiol has quit [Ping timeout: 260 seconds] 09:14:09 03dolorous * r1bc6113cb6fd 10/crawl-ref/source/item_use.cc: Fix generalized tentacles check from 399be8c to account for transformations. 09:14:19 03dolorous * r0c8963f13f30 10/crawl-ref/source/item_use.cc: Don't use mprf() when mpr() will do. 09:14:19 03dolorous * re9ded139c2c3 10/crawl-ref/source/item_use.cc: Add spacing fixes. 09:30:48 -!- ophanim has joined ##crawl-dev 09:39:27 Recent commit breaks Tiles build (https://crawl.develz.org/mantis/view.php?id=4121) by rriegs 09:42:08 -!- ais523 has quit [Read error: Connection reset by peer] 09:44:05 03dolorous * rd7cf80a7743e 10/crawl-ref/source/item_use.cc: Fix apparent misuse of the "&&" operator. 09:47:04 -!- hoody_ has joined ##crawl-dev 09:51:17 kilobyte: putting a ring as an octopode results in an infinite loop 10:10:24 -!- LordSlothW has joined ##crawl-dev 10:10:25 -!- Twilight-1 has quit [Read error: Connection reset by peer] 10:10:30 -!- Twilight-1 has joined ##crawl-dev 10:14:50 kilobyte: btw, i did not add tentacles as an accessible-to-everyone mutation, whereas i did do that for camouflage. it could be one, or even a DS facet 10:15:20 it is pretty similar to claws 10:18:05 could also be extended to tentacles for legs... right now that's hardcoded for OP but maybe we want people who are only tentacles from the waist down 10:18:43 Should I submit the following as a bug? Didn't see it in mantis: 10:19:03 Amulet of stasis does not prevent usage of Passage of Golubria: v 0.8.0 10:20:02 sounds like a bug to me 10:24:26 Eronarn, more DS body facet mutations (especially ones with non-combat bonuses) are + for me. 10:24:42 Passage of Golubria ignores amulet of stasis (https://crawl.develz.org/mantis/view.php?id=4122) by LordSloth 10:28:06 03dolorous * r85db8d63144d 10/crawl-ref/source/dat/database/rand_all.txt: Add a few more artefact names. 10:29:07 i'd like them, yes 10:29:14 like wings (-cloak) 10:36:08 03kilobyte * r4770d0a99e08 10/crawl-ref/source/item_use.cc: Fix reversed logic causing an infinite loop. 10:40:37 yeah 10:40:50 although non-active ones are even better. 10:45:30 perhaps we could make wings a three level: 1 cfly, 2 activate for temp lev, 3 permafly 10:45:34 (and no cloak) 10:45:51 (but with kenku-style boosts, not just strictly worse than 2) 10:49:57 -!- MaryPoppins is now known as Adeon 11:00:02 -!- ais523 has joined ##crawl-dev 11:02:09 -!- galehar has quit [Quit: Page closed] 11:02:17 -!- LordSlothW has left ##crawl-dev 11:14:06 03dolorous * rf362b6cc33f6 10/crawl-ref/source/makeitem.cc: Simplify ego selection for gloves. 11:14:17 03dolorous * r51378ebc26b6 10/crawl-ref/source/makeitem.cc: Constify. 11:31:40 -!- blabber has joined ##crawl-dev 11:52:31 -!- upsy has joined ##crawl-dev 11:54:38 -!- Sonko has joined ##crawl-dev 11:54:40 hi 11:55:03 i'm possibly looking for some coding, can you fellas help me there? :) 11:57:21 clarify what you mean 11:57:48 well 11:58:17 first of all, where i can see the source 11:58:47 http://gitorious.org/crawl 11:58:50 thanks 11:59:03 im a programmer, so im interested in it 11:59:16 and it's a fun game, so why not contribute 12:00:28 sure, that's always welcome 12:01:12 i'm looking for c practice, too 12:01:24 used to code in java 12:01:30 crawl is written in C++ for the most part 12:01:33 with some things in lua 12:02:48 probably the best thing to do is find some part of crawl you enjoy a lot and figure out something you can code that will make it better 12:03:19 or just browse through the code, find the worst parts and try to figure out how to make them better 12:03:36 trying to scare him off? :P 12:03:44 no, but that's what I have done :P 12:03:51 and it worked to an extent 12:03:59 ofc code lurking is the first on the list 12:04:10 as i said im not very experienced in c 12:04:20 Eronarn, I'm too awful at the game to take your suggested approach :P 12:04:30 Zaba: the early game could use work too :) 12:04:42 not everything has to be about zigs 12:05:09 Eronarn, I also have virtually no sense of balance 12:05:49 I make vaults that murder players and then pretend that they've been done that way on purpose :P 12:05:56 murdering players is good 12:05:58 keep it up 12:06:07 yup, it goes well with the philosophy of the game :P 12:06:23 ossuary_zaba_flooded has killed a lot of people 12:06:37 because I figured, "bog mummy" sounds cool! 12:06:40 i have not decided what i will work on over the remainder of the summer - possibly swarms, possibly some new spells i want to add 12:07:48 spellcasting could use a refactor, I guess 12:07:57 spl-cast.cc:_do_cast is kinda awful 12:10:01 blood could use a refactor 12:10:05 for colored blood 12:10:51 what for :P 12:11:01 and, adding new stuff != refactoring 12:11:30 Zaba: i mean that the current blood code does not at all support the possibility; it would need to be refactored to allow it 12:11:41 also, for glaive of prune :) 12:11:44 you can refactor to make adding new stuff easier, but doing the two at the same time can get painful 12:11:45 and insects not bleeding red, etc. 12:11:57 and harder to track and to revert and such 12:12:13 contributing usually consists on implementing new features, or rather coding new monsters/areas/items/etc? 12:12:35 the stuff you listed on the right are things that count as new features 12:12:48 Sonko: more than that 12:12:51 Sonko, normally it consists of making vaults 12:13:00 Sonko: also interface improvements, new tiles, new vaults, new speech 12:13:29 i see 12:14:18 There is a tag "Implementables" on the Mantis tracker, and the interface ideas wiki page has a list of items marked as "patches welcome" (the Xom list too) 12:14:37 do we have 'patches unwelcome' yet? we should 12:16:36 I have the impression that development consists mostly of nerfing things that make things easier for the player and buffing things that make it harder. 12:17:11 Chousuke: as it should be! 12:17:14 sadly, there are regressions 12:17:27 true 12:18:42 as long as there are players who can get the orb, your work is not done? 12:19:18 You know it's not true: armour unnerfed etc. If you're serious about the complaints, please take it to ##crawl. 12:20:35 -!- Zaba has quit [Ping timeout: 276 seconds] 12:25:28 -!- Zaba has joined ##crawl-dev 12:26:36 -!- neunon has quit [Remote host closed the connection] 12:26:36 well, thats enormous code to be digested.. but im unemployed, so D: 12:34:27 the game generates the same dungeon for every class/species? 12:35:03 yes 12:35:15 why would it generate different ones? 12:36:05 for example, for casters, more spellbookz 12:36:14 no, no stuff like that 12:36:24 yeah 12:36:33 "no plate armor for you, dude" 12:36:35 :) 12:36:42 everybody gets exactly the same odds at troublesomeness. 12:37:24 nice 12:47:13 -!- upsy has quit [Quit: Leaving] 12:50:28 -!- hoody_ has quit [Remote host closed the connection] 12:51:24 -!- ophanim has quit [Quit: Leaving.] 12:52:07 -!- ophanim has joined ##crawl-dev 12:52:23 -!- ophanim has quit [Client Quit] 12:59:52 -!- Sonko has quit [] 13:01:40 -!- hoody has joined ##crawl-dev 13:05:17 -!- dpeg has quit [Quit: leaving] 13:20:24 -!- monqy has joined ##crawl-dev 13:22:11 -!- hoody has quit [Read error: Connection reset by peer] 13:23:26 -!- hoody has joined ##crawl-dev 13:28:18 -!- hoody_ has joined ##crawl-dev 13:28:20 -!- hoody has quit [Read error: Connection reset by peer] 13:30:53 -!- blabber has quit [Ping timeout: 252 seconds] 13:45:01 -!- Sonko has joined ##crawl-dev 13:50:22 -!- blabber has joined ##crawl-dev 13:53:50 -!- hoody_ has quit [Remote host closed the connection] 14:03:24 -!- upsy has joined ##crawl-dev 14:05:47 -!- blabber has quit [Quit: Lost terminal] 14:09:00 -!- Textmode has joined ##crawl-dev 14:25:09 -!- galehar has joined ##crawl-dev 14:41:26 -!- galehar has quit [Remote host closed the connection] 14:43:36 -!- galehar has joined ##crawl-dev 14:52:31 03galehar * rd93c957518e9 10/crawl-ref/source/skills2.cc: Allow reskilling to fighting. 14:52:32 03galehar * r6f61cdb7bd61 10/crawl-ref/source/dat/lua/stash.lua: Add "branded" as an alias for "ego" in search. 14:52:32 03galehar * r4237d65238d5 10/crawl-ref/source/skills2.cc: Bring back the reskilling XP penalty to 10%. 14:52:39 03rriegs * rd0e4d3d76398 10/crawl-ref/source/tilepick.cc: Fix broken COMPILE_CHECKs in tilepick.cc 14:59:31 -!- neunon has joined ##crawl-dev 14:59:36 -!- neunon has quit [Changing host] 14:59:36 -!- neunon has joined ##crawl-dev 15:22:09 -!- hoody has joined ##crawl-dev 15:27:08 -!- Cryp71c has quit [Quit: Leaving] 15:53:18 -!- Sonko has quit [] 15:56:38 -!- neunon has quit [Remote host closed the connection] 15:58:17 -!- evilmike has joined ##crawl-dev 16:30:36 -!- edlothiol has joined ##crawl-dev 16:36:18 -!- Textmode has quit [Quit: Ex-Chat] 16:36:29 03kilobyte * r7a3a21ddfb6d 10/crawl-ref/source/wiz-you.cc: Update maxhp on wizmode race changes. 16:36:30 03kilobyte * r1c94e3868d03 10/crawl-ref/source/ (files.cc syscalls.cc syscalls.h): alarm() for Windows. 16:36:30 03kilobyte * r62783aa6b3ef 10/crawl-ref/source/crash-u.cc: Hard-limit crash dump to 5 seconds. 16:36:40 03kilobyte * r63a492f51f97 10/crawl-ref/source/package.cc: Fix a warning on certain architectures. 16:37:11 kilobyte: quad damage in the regular game really? 16:40:21 elliptic: it was suggested as part of misc overhaul -- it's an equivalent of drinking a set of boost potions all together 16:41:10 the flavour really doesn't fit well in the regular game IMO 16:41:47 it also is completely different from any other misc item at the moment 16:42:02 I haven't seen the misc overhaul plans, though... 16:42:49 you'd have to burn a number of acquirement scrolls to get one, for a single fight, or be exceptionally lucky with an "acquire any" vault (some Lair ends have one) 16:43:23 no concrete write-up anywhere, and what write-ups are around are all wildly different 16:44:00 and changes that are wrong may cause negative feedback 16:44:07 like, say, just what you're doing :p 16:44:34 I'm not so worried about balance, yes... I mainly just think the item is way too silly for regular crawl 16:46:03 right, so I should proceed with clockworkdemons then? :p 16:46:36 clockworkdemons sound better to me than quad damage, sure :P 16:48:47 -!- syllogism has quit [] 16:48:51 rename it megadamage ;) 16:50:52 galehar: hey, I had a question I wanted to ask you about skill training 16:51:02 shoot 16:51:21 currently skill training is noncommutative: the order in which you train skills matters 16:51:50 yes, because of skill_cost_level 16:52:05 if you have lots of xp in your pool, then the order in which you victory dance skills with it matters 16:52:35 this effect is pretty small in practice, but it still bothers me (and other people have brought it up) 16:52:44 how hard do you think it would be to fix this? 16:52:58 remove victory dancing? :p 16:53:09 well, the problem isn't with victory dancing 16:53:12 really though, it doesn't seem to make a difference to me 16:53:18 kilobyte: it won't really fix this 16:53:48 training low level skills first and they level up fast while the cost is low 16:54:11 whether you'll obtain skills A and B paying 2 for A then 3 for B -- or 2 for B then 3 for A, your gain doesn't change 16:54:26 but as you gain skill points faster, it makes the cost raise faster too, which means specialising will be harder 16:54:37 kilobyte: that's how it should work, but it isn't actually true with the current system 16:54:44 the discrepancy is quite small as I said 16:54:50 ok, then that's a bug 16:55:44 elliptic: I think the current system has a benefit. It kinda forces you to choose between being a specialist or a generalist. 16:56:02 training a lot of skills will make raising one high slower 16:56:59 yeah, but the order should never matter 16:57:20 getting 1 A, 1 B, 10C should cost just as much as 10C, 1A, 1B 16:59:07 the current system with the skill cost has been written by Peter. His source code comment says he ran a simulation of gaining skills. 16:59:32 (the numbers before A, B and C are values of XP put into a skill, not the skill rank, of course) 17:00:00 Rewriting that while keeping the balance sounds hard. 17:01:35 how is that different from reskilling? 17:02:08 if costs differ later, it's abusable 17:02:22 reskilling just transfer skill points from one skill to another. It doesn't care about XP. 17:03:08 yeah, and that's why obtaining skill points shouldn't depend on order 17:03:33 well, yes, you're transfering points you didn't buy at the same cost. 17:03:53 first X skill points should cost a given amount no matter what you use them for 17:04:22 To achieve that, skill_cost should only depend on skill_cost_level, not skill level. 17:04:32 galehar: yes 17:04:48 that's actually how I always assumed it worked before I looked at the code carefully... 17:06:06 https://spreadsheets0.google.com/spreadsheet/pub?hl=en_US&hl=en_US&key=0AqvhLOPFHpiMdFV3UWg1RndxM2FtM3ZrOTNUQnBqRHc&output=html 17:06:15 that's the skill cost table 17:08:12 we can try using the diagonal (treat skill level as if it is equal to skill_cost_level 17:08:46 it would make training a new skill in the late game harder 17:09:11 well, in the mid-game it would be harder. It wouldn't change the late game 17:09:16 might not be a bad thing 17:10:55 why would it make it harder? It shouldn't change how many skill points you need to get skill rank X, just how many skill points you get per point of XP. 17:13:47 and you would get less. It would make the skill point cost more XP for low level skills. 17:16:48 the difference would be especially harsh in the early game. At skill_cost_level 4, 10 skill points for a L0 skill cost 1 XP. With the new system it would cost 5. 17:17:20 you could bump starting XP to compensate 17:19:07 or reduce the cost in skill points of the first skill levels. Especially the first one which is expensive (200 skill points). 17:20:51 yes, tweaking the skill point costs a bit would probably be a good idea with a change like this 17:21:59 since as I understand things, currently higher level skills effectively cost more xp than you would expect from the skill point costs alone because of the dependence of skill_cost on skill_level 17:22:31 and lower level skills cost less 17:22:45 "bump starting XP"... except that 25 free XP has no obvious way of being ported to instant assignment 17:22:58 -!- ophanim has joined ##crawl-dev 17:23:13 right, with instant assignment we'd want to get rid of that starting xp 17:23:39 elliptic: the difference gets lower as you progress. Until skill_cost_level 24 where all costs are the same at 250 17:23:48 really the starting xp doesn't serve much purpose even now 17:24:22 on a lot of casters you should just victory dance it before doing anything, and on melee chars it usually gets used up near-instantly 17:25:02 Fixing that sounds a good idea, but I don't know how to do it properly. Run simulations like Peter did? Have an expert tweak the numbers manually? Both? 17:26:17 record what people use it for? 17:26:34 (or just make a wild-ass guess, which is cheapest and about as good) 17:27:54 skill level gains are in the notes. We already have a lot of data we can use too. Maybe if we add skill_cost_level to the notes it can help to figure it out. 17:31:31 do you mean, for the first 25 XP or for further gains? 17:32:03 for further gains. I was thinking about how to fix the skill cost being dependent on skill level 17:33:41 if you want to analyze >6GB of morgues, tell me what to search for 17:35:02 I really don't know yet, I'll have to think about it. 17:36:38 but with the upcoming removal of victory dancing, it would be good to address that. With full control over skill training, some crazy people might try to optimise their training to abuse the variable cost. 17:39:13 about that removal, I think we should speed up, to have a modicum of time to playtest it... and it's a big change 17:39:36 true 17:39:39 I thought about some of the details, but for the real meat, I'm not sure I understand the skill system enough 17:40:09 we're going to keep victory dancing the first point of a skill, right? 17:40:24 yes, that's the idea 17:40:46 so you'll have to disable manual mode to gain new skills 17:40:53 if so, it'd need to borrow XP instead of using the non-existant pool 17:41:12 no, it has hardly any ties to autimatic/manual mode anyway 17:41:58 in manual mode, your actions are ignored, all the XP goes into the selected skills. You can't train new skills. 17:42:06 having skills train only with a full pool means we'd have to display the pool 17:42:37 automatic mode also makes all XP go into selected skills, it just makes choices for you 17:42:54 yes, so there's no pool, it's always empty. 17:44:17 you'll have to do some "pre-battle dance" 17:45:11 but you won't have to dance too much. As soon as the skill has some % of incoming XP, it will keep being trained. 17:45:32 if you disable others, then it can get most (or all) incoming XP easily. 17:46:01 it's not good -- you'd have to both do a pre-battle dance and to skip doing other things 17:46:28 even worse with reading scrolls which have an actual cost (few scrolls) 17:47:12 disabling other skills would help, but it's a lot of micro-management of the skill menu 17:47:39 I thought the idea was to eliminate victory dancing anyway... do we really need to keep it around for this? 17:48:19 borrowing XP like kilobyte suggested sounds like it could work, though 17:49:30 but how can you tell if a player is dancing to gain a new skill or not? It's no good if you select a skill to gain all the XP and it still goes to some you don't want. 17:49:51 it would only be for getting skill 1 17:50:28 I think it is fine that people will gain some skills regardless 17:50:46 otherwise people might never learn of the existence of certain skills :P 17:50:57 So maybe we can have say 5% of XP to go in the pool, which is cap very low, only used to gain first level and hidden 17:52:34 something like that sounds okay to me except possibly at the very beginning of the game 17:52:37 that'd just effectively give you -25 starting xp 17:52:58 since we don't want to make it too hard to gain L1 of a weapon skill at start 17:53:38 it'd steal some XP to fill that additional pool, and it'd be full 99.9% of the time 17:54:57 -!- hoody has quit [Read error: Connection reset by peer] 17:54:58 You only need 20XP to gain the first level of a skill in the early game 17:55:36 kilobyte: intentional that randart armour can have really low negative enchantments now? I just found a -10 chain mail... 17:55:37 about draining: should I add some logging of how much is actually lost? 17:55:38 the cap should probably depend on the skill_cost_level 17:55:51 -!- hoody has joined ##crawl-dev 17:56:26 elliptic: I had no idea what would be good values... if it can go high, it makes sense for it to go low as well, but if you have any feelings, please improve. 17:57:03 well, but the mere existence of that pool would bring victory dancing back 17:57:13 hm, how low can corroded armour go? 17:57:14 you'd have to read a scroll, kill something, read another scroll 17:57:16 kilobyte: I don't think it's necessary. It would help (again!) if we had a unique skill cost. 17:57:49 it shouldn't be possible for randarts to go much below extremely corroded armour, I'd say 17:57:50 instead of having same effect as reading both scrolls without to trying to game the system 17:57:54 but doesn't really matter of course 18:05:26 How about this: we keep track of all the exercise of L0 skills. As a debt. 5% of incoming XP is used to pay debts and train L0 skills. That way, no matter when and how you dance, skill will be trained. 18:06:48 no need to limit that to 5%, it could take care of draining as well 18:07:52 we can even use a similar system in automatic mode btw (instead of a queue as I suggested on the wiki). With a cap on the number of exercises. 18:10:28 03kilobyte * r86dade9f2026 10/crawl-ref/source/ouch.cc: Spam people's notes about XP/pool lost to draining. 18:12:40 -!- upsy has quit [Quit: Leaving] 18:23:48 -!- galehar has quit [Remote host closed the connection] 18:23:56 -!- hoody has quit [Ping timeout: 276 seconds] 18:25:24 -!- hoody has joined ##crawl-dev 18:34:42 with petrification added, transmutation miscasts should cause that instead of paralysis (which is what they do now) 18:42:27 good idea 19:28:22 -!- hoody has quit [Remote host closed the connection] 19:42:27 03kilobyte * r212b4e48ebda 10/crawl-ref/source/spl-miscast.cc: Use petrification rather than paralysis in Tmut miscasts. 19:42:38 03kilobyte * rcb6b05805603 10/crawl-ref/source/ (actor.h beam.cc monster.cc monster.h player.cc player.h): Don't pretend petrification has an argument. 20:49:41 -!- edlothiol has quit [Ping timeout: 240 seconds] 21:03:50 -!- hoody has joined ##crawl-dev 21:11:29 -!- Twilight-1 has quit [Ping timeout: 260 seconds] 21:19:40 -!- petete has quit [Remote host closed the connection] 21:20:04 -!- Twilight-1 has joined ##crawl-dev 21:20:23 -!- hoody has quit [Remote host closed the connection] 21:20:25 -!- petete has joined ##crawl-dev 21:43:21 Multiple sets of the same scale mutation? (https://crawl.develz.org/mantis/view.php?id=4123) by wolfechu 21:43:43 -!- elliptic has quit [Quit: Leaving] 21:43:58 -!- elliptic has joined ##crawl-dev 22:10:06 -!- Twilight-1 has quit [Read error: Connection reset by peer] 22:10:29 -!- Twilight-1 has joined ##crawl-dev 22:38:18 -!- Sequell has quit [Ping timeout: 255 seconds] 22:38:26 -!- hoody has joined ##crawl-dev 22:38:58 -!- Sequell has joined ##crawl-dev 22:52:59 -!- Twilight-1 has quit [Ping timeout: 260 seconds] 23:03:25 -!- Twilight has joined ##crawl-dev 23:03:47 -!- ophanim has quit [Read error: Connection reset by peer] 23:03:52 -!- Twilight is now known as Guest16446 23:21:36 -!- hoody has quit [Remote host closed the connection] 23:28:19 -!- hoody has joined ##crawl-dev 23:28:53 any mac devs here? 23:30:45 having lots of problems with getting the xcode project to build a tiles binary 23:30:53 could use a pointer 23:33:12 We don't have an active mac dev at the moment, so it tends to lag behind ... 23:33:39 ah 23:33:46 the xcode project is just not even building 23:33:53 is tht because no mac dev is on top of the source? 23:34:07 if so it will save me banging my head against the wall figuring out what I did wrong 23:41:58 My guess is that that'd would be the case, yeah 23:49:05 probably because there have been a lot of source code, etc, changes. 23:49:29 yeah 23:49:31 ill give up then 23:49:31 hoody: the makefile should still compine in theory; if you do get the xcode project working, we'd appreciate a patch 23:49:37 i would 23:49:43 but there's literally thousands of errors 23:49:51 loads and loads of "can't find x x x " 23:50:00 not got time to pick through it 23:50:16 by the time I did, source would be updated 23:56:19 hm 23:56:38 best suggestion would be to find the most recent changes to the xcode in git 23:56:46 that should give a hint as to what needs to be changed.