00:03:19 -!- evilmike has joined ##crawl-dev 00:03:23 Unstable branch on crawl.develz.org updated to: 0.10-a0-2485-ga1df136 (32) 00:05:06 03evilmike * r061dcfa635a0 10/crawl-ref/source/dat/des/branches/zot.des: Change permarock to stone in zot entry vaults. 00:05:16 03evilmike * rdd3d67d2346d 10/crawl-ref/source/dat/des/variable/float.des: Fix a missing wall square in abrahamwl_protected_by_tso_2. 00:18:54 Windows builds of master branch on crawl.develz.org updated to: 0.10-a0-2487-gdd3d67d 00:36:18 Pacified monsters become unpacified sometimes without angering Ely. (https://crawl.develz.org/mantis/view.php?id=4999) by elliptic 01:04:05 -!- Brannock has quit [Ping timeout: 252 seconds] 01:12:51 thanks for the feedback. that's something I fixed a while back locally, apparently pushing to the clone messed up somewhere 01:15:09 I've tried multiple constriction with an octopode, and have successfully killed two opponents with constriction alone while bashing on a 3rd with a weapon :D 01:35:03 -!- dtsund has quit [Quit: dtsund] 01:42:32 -!- petete has joined ##crawl-dev 01:44:52 -!- dtsund has joined ##crawl-dev 01:52:08 -!- Pingas has joined ##crawl-dev 02:07:06 ok, pushed a fix for the monster constricting player assert 02:10:50 -!- dtsund has quit [Quit: dtsund] 02:34:57 -!- monqy has quit [Quit: hello] 02:36:34 -!- syllogism has joined ##crawl-dev 03:05:05 -!- evilmike has quit [Quit: something happened] 03:18:30 -!- Cryp71c has quit [Quit: Leaving] 03:20:51 -!- MarvinPA has quit [Ping timeout: 252 seconds] 03:22:04 -!- alefury has joined ##crawl-dev 03:33:59 -!- st_ has quit [Ping timeout: 244 seconds] 03:54:12 -!- blabber has joined ##crawl-dev 04:49:08 -!- Textmode has joined ##crawl-dev 04:50:41 -!- Ragdoll has joined ##crawl-dev 05:04:24 -!- timecircuits has quit [Ping timeout: 244 seconds] 05:19:26 -!- jeanjacques has joined ##crawl-dev 05:38:46 -!- MarvinPA has joined ##crawl-dev 05:59:53 -!- blabber has quit [Quit: leaving] 06:07:05 -!- snoonan has joined ##crawl-dev 06:07:50 -!- snoonan has quit [Client Quit] 06:08:45 -!- neunon has joined ##crawl-dev 06:09:06 -!- neunon has quit [Changing host] 06:09:06 -!- neunon has joined ##crawl-dev 06:58:11 -!- Pacra has joined ##crawl-dev 07:19:22 -!- neunon has quit [Quit: Bye!] 07:30:15 -!- blabber has joined ##crawl-dev 07:55:19 -!- G-Flex has quit [Quit: Time for us to bug out.] 07:55:27 -!- edlothiol has joined ##crawl-dev 07:56:08 hmm... spiderform octos can breathe in deep water 07:56:09 can't decide whether this is good 07:57:12 Sounds bad to me. 08:03:15 spidermerfolk can't swim, can they? 08:03:17 MarvinPA: You have 1 message. Use !messages to read it. 08:03:21 -!- Brannock has joined ##crawl-dev 08:03:23 no idea 08:07:53 -!- timecircuits has joined ##crawl-dev 08:09:05 <|amethyst> they cannot 08:44:09 -!- ais523 has joined ##crawl-dev 08:44:28 !tell bluedave i posted a little more constriction feedback on the mantis item: https://crawl.develz.org/mantis/view.php?id=4868 08:44:29 MarvinPA: OK, I'll let bluedave know. 09:18:14 Reading a scroll identifies it, despite "Nothing appears to happen" (https://crawl.develz.org/mantis/view.php?id=5000) by moxian 09:20:59 -!- ainsophyao has quit [Remote host closed the connection] 09:24:51 -!- ainsophyao has joined ##crawl-dev 09:25:21 elynaeandxom (L11 MuCK) ASSERT(num_visits == levels_seen) in 'place-info.cc' at line 46 failed. (Lab) 09:28:14 -!- Elynae has joined ##crawl-dev 09:30:03 are there any plans to fix the labyrinth/abyss bug currently? my game is kinda screwed up due to that 09:48:32 -!- Elynae has quit [Remote host closed the connection] 10:03:28 Visible items list truncates entries early. (https://crawl.develz.org/mantis/view.php?id=5001) by elliptic 10:12:27 !tell evilmike The sacrificial humans in evilmike_hellhole interrupt resting and autoexplore for some reason, any idea why? 10:12:28 elliptic: OK, I'll let evilmike know. 10:23:15 -!- Wop has quit [Ping timeout: 252 seconds] 10:38:12 -!- Wop has joined ##crawl-dev 10:57:33 Unstable branch on crawl.akrasiac.org updated to: 0.10-a0-2487-gdd3d67d (32) 11:11:04 03dolorous * r9bbcde73b035 10/crawl-ref/source/food.cc: Remove the special cases for vampires' feeding on fresh human/elf corpses. 11:11:15 03dolorous * r558c8fad724f 10/crawl-ref/source/food.cc: Fix indentation. 11:27:02 what was the point of that rebuild ? 11:27:18 it doesn't look like anything changed besides a couple of vaults? 11:30:55 -!- Wop has quit [Ping timeout: 248 seconds] 11:41:23 samb: CAO updates daily automatically if that's what you mean? 11:41:36 oh, that was CAO 11:41:48 * SamB isn't used to it updating at this time of day 11:42:19 * SamB is used to 5:00 UTC 11:43:03 -!- Wop has joined ##crawl-dev 11:56:04 howdy 11:56:05 blueDave: You have 1 message. Use !messages to read it. 11:56:09 !messages 11:56:10 (1/1) MarvinPA said (3h 11m 42s ago): i posted a little more constriction feedback on the mantis item: https://crawl.develz.org/mantis/view.php?id=4868 11:58:58 good feedback, I hadn't thought to try the distinctions between controlled and uncontrolled transloc 12:00:27 -!- st_ has joined ##crawl-dev 12:01:01 -!- medgno has joined ##crawl-dev 12:06:07 -!- chukamok has joined ##crawl-dev 12:12:49 ...blink frogs constrict? 12:12:53 -!- alefury has quit [Ping timeout: 244 seconds] 12:14:23 no 12:14:51 apparently if you constrict them and they blink then they take you along, if I am interpreting the mantis comments correctly 12:15:10 oh, i see 12:15:48 -!- alefury has joined ##crawl-dev 12:21:02 that's interesting! 12:21:33 could be cool, if it, like, reduced their blink range ... 12:21:43 ... though there's water to consider 12:21:54 what exactly can constrict? 12:22:10 PC-wise 12:22:39 03dolorous * rb283bbdecfb3 10/crawl-ref/source/ng-setup.cc: Add comment. 12:29:31 someone should code an octo unique, btw 12:29:43 i've been thinking it would be cute to have one that's an octopode and a felid 12:29:47 eight arms, nine lives 12:29:57 octopussy 12:30:17 we could ask github for permission to use their logo for the tile ;-P 12:30:31 oh man 12:30:41 let's make them brigands 12:30:53 then the octopus can be named The Eight-Armed Bandit 12:35:32 octopodes and nagas can constrict 12:35:48 can they both swim ? 12:36:18 * SamB hasn't tried either much 12:36:35 hmm, don't know -- but it should veto the square if the function I used does what its name implies ;) 12:36:44 oh 12:37:13 okay, so only lava orcs and the like will get dragged over lava if their constrictee blinks there, etc.? 12:37:25 right 12:37:45 sounds cool 12:38:12 does it reduce the blink range at all ? 12:38:30 * SamB would imagine that more bodies == harder to blink 12:39:00 no, that might have been suggested on the wiki page but aiming for just getting it to work 12:39:12 fair enough 12:39:21 I'm not sure if it would be any fun, anyway 12:40:12 -!- elliptic_ has joined ##crawl-dev 12:40:18 -!- elliptic has quit [Read error: Connection reset by peer] 12:40:23 -!- elliptic_ is now known as elliptic 12:40:53 what to chanserv so long ? 12:41:04 (didn't you identify before you switched nicks?) 12:41:23 (yes, you can identify as your main nick even when you don't have it!) 12:41:24 if the underscore isn't registered it will wait till the nick change 12:41:42 unless it knows your real name? 12:41:44 (you just need to give the username before your password) 12:41:50 (iirc) 12:42:20 -!- alefury has quit [Ping timeout: 244 seconds] 12:42:28 I gather you have two computers and are logged in from both? 12:42:42 oh, yes, I do 12:43:09 three actually, but only two are connected to monitors and I haven't really taken to irssi yet 12:44:59 though SamB_XP itself is actually linked to SamB 12:46:09 -!- alefury has joined ##crawl-dev 12:46:15 so, what is the git:// url for your repo? 12:46:36 http://gitorious.org/~bluedave/crawl/constriction 12:46:51 hmm, that's http 12:46:58 I can figure it out 12:47:35 lets see, take git://gitorious.org/~neilmoore/crawl/neilmoore-crawl.git for an example, and ... 12:47:40 git://gitorious.org/~bluedave/crawl/constriction.git 12:47:59 ah, yeah, that was what I was going to come up with alright ;-) 12:48:29 I've been pulling master and rebasing into mine, which is getting very ugly 12:48:41 is it better to just leave that till merge? 12:48:55 afraid of dropping stuff, like the assert that popped up last night 12:49:07 for long term branches, it's fine to *merge* master into yours 12:49:57 for short-term-ish branches that nobody actually seems to be pulling from, you can rebase *them* onto master 12:51:18 under *no* circumstances should you rebase master onto yours, though cherry picks could be justifiable... 12:52:14 think ive been rebasing mine onto master -- pull master, switch to my branch, git rebase master 12:52:25 * SamB hates how gitorious defaults to copying all branches when you fork a repository ... 12:52:26 xomscumming (L1 VpCK) ASSERT(!did_msg) in 'spl-miscast.cc' at line 413 failed. (D:1) 12:52:39 blueDave: ah, yeah, that's not HORRIBLE 12:52:50 which should be putting master's commits under mine, but it's beginning to leave fallout 12:52:54 sigh. 12:52:56 though I'd suggest that you just pull master int yours 12:53:09 so I pull master, switch to mine, then merge master? 12:53:21 that'd work, yeah 12:53:33 are you using ccache ? 12:53:37 and master will go on top, then when mine gets merged it figures out the diffs and makes a new commit 12:53:58 not sur what ccache is, locally it's msysgit 12:55:00 it's a tool to help avoid recompiling the same thing 12:55:28 I think there's a script in msysgit for building it 12:55:49 ah, the make everything can be a pain, but at least it's only 3-4 minutes 12:56:06 * SamB uses plain msys + the prebuilt git for his Windows builds 12:56:16 xomscumming (L1 MuCK) ERROR: range check error (-1 / 80) (D (Sprint)) 12:56:18 yeah, I guess when you have a decent computer it's not too bad ;-) 12:56:32 everything is crashing on me, damn it all 12:56:37 * SamB 's main dev system is a 450 MHz PII 12:57:04 (Running Linux, or it wouldn't be my main dev system!) 12:58:02 -!- monqy has joined ##crawl-dev 12:59:00 !lm xomscumming sprint crash -log 12:59:00 my Windows machines have higher statistics, but also higher overheads for opening files and (especially) processes 12:59:00 2. xomscumming, XL1 MuCK, T:1799 (milestone): http://crawl.develz.org/morgues/trunk/xomscumming/crash-xomscumming-20111203-185606.txt 12:59:28 !lm xomscumming crash -log 12:59:28 21. xomscumming, XL1 VpCK, T:1803 (milestone): http://crawl.develz.org/morgues/trunk/xomscumming/crash-xomscumming-20111203-185216.txt 13:00:32 on that machine I can see needing to squeeze everything you can 13:01:28 (Also, Crawl's contribs have bastardized build systems that can't build DLLs, PLUS gold is ELF-only, so links take forever and tons of RAM.) 13:03:21 * SamB goes to try and file some sort of ticket about the stupid "copy all branches" behaviour of gitorious 13:03:49 -!- blabber has quit [Ping timeout: 240 seconds] 13:05:11 -!- blabber has joined ##crawl-dev 13:28:55 hmm, if I were to create a github "orginization" to hold mirrors of our repos, should I call it "dcss" or "crawl-ref" or what? ("crawl" seems taken.) 13:39:29 do items usually spawn in a wall in sprint? 13:44:08 probably not 13:44:15 that's almost certainly a bug in something 13:44:31 I've heard of such bugs before, certainly 13:46:13 -!- ainsophyao has quit [Remote host closed the connection] 13:47:00 so, anyone notice the change I made to "I"? 14:11:15 -!- edlothiol has quit [Quit: leaving] 14:12:46 -!- dtsund has joined ##crawl-dev 14:17:35 xomscumming (L1 MuCK) ASSERT(feat < NUM_FEATURES) in 'feature.cc' at line 24 failed. (Abyss (Sprint)) 14:24:25 xomscumming seems be producing a lot of unrelated crashes today 14:24:36 this is why we need to keep xom scummable :P 14:24:58 haha indeed 14:25:14 what is xomscumming? 14:25:22 mummy of xom 14:25:26 scumming 14:25:29 cant make head or tails of most crashes though 14:25:53 elly: the gist of it is you just do nothing while xom does do something, hoping he does something favorable 14:26:01 ah, I see 14:26:11 that sounds easily automated 14:26:15 indeed 14:26:52 my last two games i have been able to train shields in sprint while never having seen a shield 14:32:36 elly: xomscumming has quite a few entries in ??hilarious_deaths 14:33:19 oh :P 14:35:03 elly: hilarious deaths[63/68]: !lg elynae d:1 killer=pit fiend -tv 14:35:19 if you get a chance sometime, that's an inimitable one 14:39:05 SamB, blueDave: except for massive merges, "git rerere" can eliminate all merge commits but the final one 14:39:12 -!- petete has quit [Ping timeout: 244 seconds] 14:39:27 it can help, yeah 14:39:53 elliptic: question. xom always gives a staff of x on 'perhaps you should try this instead/have you considered using one of these/ how about this' is this intended 14:40:29 no time for a Xom overhaul, but perhaps we could at least eliminate ticks? 14:40:37 dont do it kilobyte 14:40:45 you would make me very sad 14:40:55 hm, looks unintended... godspeak.txt claims that this is for "Xom will choose a weapon of a different subtype from the one you're currently wielding." 14:41:14 yes, i looked it up, i picked axes instead of spears but staffs all the way 14:43:12 well, it looks intended judging from the actual code... godspeak.txt might just be wrong about "subtype" 14:43:36 assuming I know what subtype means in crawl 14:43:52 it looks like if you sit around wielding a staff, you should get weapons 14:44:04 ill try that out 14:44:12 also you should be getting misc items 10% of the time 14:44:42 xomscumming (L1 MuCK) ASSERT(feat < NUM_FEATURES) in 'feature.cc' at line 24 failed. (Abyss (Sprint)) 14:44:54 hm. 14:46:46 !lg . crash 14:46:46 Unknown selector: verb 14:46:55 !lg xomscumming -crash 14:46:55 Malformed argument: -crash 14:46:59 !lm xomscumming sprint crash -log 14:46:59 4. xomscumming, XL1 MuCK, T:6622 (milestone): http://crawl.develz.org/morgues/trunk/xomscumming/crash-xomscumming-20111203-204430.txt 14:48:17 xomscumming (L1 MuCK) ASSERT(feat < NUM_FEATURES) in 'feature.cc' at line 24 failed. (Abyss (Sprint)) 14:48:21 maaaan 14:48:31 _There is a staircase to the Lair here. 14:48:35 this is becoming a interesting abyss 14:49:28 Welcome back to the Lair of Beasts! Welcome to the Lair of Beasts! 14:50:42 i spawned on a known up-stairs 14:50:47 i took it and escaped the dungeon 14:50:51 this is brilliant 14:53:16 you esca[ed the dungeon from the abyss? nice 14:53:28 i went to lair:2 first. in sprint 14:53:55 i wish i explored lair:2 instead of going upstairs :( 14:54:05 -!- chukamok has quit [Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243]] 14:54:15 hey, would that let you get more than 10 runes in the 10 rune challenge? 14:54:29 you could collect the runes in the branches as well as the ones in the Sprint level 14:54:40 i dont think there was a way back to sprint 14:54:50 D:1, hello? 14:55:07 i took the upsstair case from lair:2 and escaped the dungeon alive 14:55:12 oh 14:55:23 so you're saying the whole thing went insane ? 14:55:29 very insane 14:55:41 i got abyssed and found a staircase to the lair there 14:55:51 * SamB blames evilmike 14:56:34 i like how my .txt file i made less than 30 minutes ago now have 8 bugs 15:16:58 03dolorous * r16f1acdcc707 10/crawl-ref/source/delay.cc: Fix Mantis 4987: Capitalize pronouns in monsters' weapon descriptions again. 15:23:34 -!- casmith789 has joined ##crawl-dev 15:32:25 03dolorous * rad8c685bac81 10/crawl-ref/source/dat/des/portals/ (icecave.des ossuary.des spider.des volcano.des): Remove doubled articles in some portal vault descriptions. 15:38:27 dolorous: what about the .txt file? 15:41:03 * SamB toys with using wider columns in various two-columm help screens 15:41:13 (in wider terminals) 15:41:43 * SamB thinks he dislikes it 15:55:27 -!- Wop has quit [Ping timeout: 248 seconds] 16:00:14 oh, what do you know ... it works 16:02:05 -!- Pingas has quit [Quit: Leaving] 16:08:03 -!- Wop has joined ##crawl-dev 16:21:48 -!- HangedMan has joined ##crawl-dev 16:24:51 cherubs meleeling seems to crash pretty reliably 16:30:52 -!- Ragdoll has quit [Quit: Up, up and away!] 17:32:55 03SamB * rfa595cb8c6a8 10/crawl-ref/source/mon-act.cc: Add a couple of comments to mon-act.cc. 17:32:55 03SamB * r41b618594478 10/crawl-ref/source/dat/dlua/util.lua: lua: Fix table_to_string() with boolean values. (For crash dumps, etc.) 17:33:05 03SamB * r4cd1acae612f 10/crawl-ref/source/dbg-asrt.cc: Include a screenshot when crash-dumping, near the end in case that crashes. 17:42:55 * SamB thinks that normal descriptions should be displayed using the menu system, like help text is 17:44:38 * SamB also thinks that these two things belong in a header of their own: 17:44:40 template void process_description(T &proc, const describe_info &inf); 17:44:40 template void process_quote(T &proc, const describe_info &inf); 17:50:18 03elliptic * r6578c790d305 10/crawl-ref/source/l_view.cc: Lua function view.is_safe_square(). 17:50:29 03elliptic * r7941f2ef976c 10/crawl-ref/source/l_moninf.cc: Another player lua function, for checking whether a monster is sleeping or paralysed. 17:50:29 03elliptic * r0b85a9d3aca1 10/crawl-ref/ (docs/options_guide.txt source/dat/lua/autofight.lua): Autofight improvements. 17:58:34 -!- MarvinPA_ has joined ##crawl-dev 18:00:46 elliptic: personally i prefer knowing what autofight is going to attack 18:00:57 so im not so sure about the prioritization 18:01:11 well, the prioritization still uses information available to you 18:01:17 probably it is a good think in the situations you want to use autofight in 18:01:21 *thing 18:01:47 im just not sure, i sometimes just use autofight when i know what its going to attack and am too lazy to use a direction key/mouse 18:01:55 I personally found it really bad that autofight would switch which monster it was attacking when another monster entered melee range from another side 18:01:57 -!- MarvinPA has quit [Ping timeout: 252 seconds] 18:02:01 the only thing I don't like about your autofight is stopping on low HP, but other people seem to love that 18:02:05 actually i used to think i preferred knowing what monster it was going to attack 18:02:10 st_: it is configurable, you can remove that 18:02:12 yeah, that was quite bad 18:02:17 yeah I did 18:02:22 but having tried it, new autofight does seem better at picking stuff 18:02:26 -!- MarvinPA_ is now known as MarvinPA 18:03:19 i havent actually tried your new autofight, i just think the rules that determine what is attacked should be simple enough to figure out nearly instantly 18:03:26 oooh... elliptic: <3 <3 <3 18:03:32 like in situations where you're just spamming it while surrounded and don't care much, it does a better job than just attacking stuff clockwise from top-left 18:03:47 so even when you don't actually need to prioritise much, it does that for you anyway 18:03:47 it wasn't clockwise anyway 18:03:51 it was something weird 18:03:52 could add stuff like decreasing the priority of poisoned monsters, etc... but there's no end of possible tweaks 18:03:55 if new autofight rules are simple and different, thats okay, if they are a bit complicated that would suck 18:04:11 alefury: I tried to keep them simple 18:04:25 should be good then :) 18:04:34 and yeah, having it continue to attack the same monster you were attacking before alone seems like a huge improvement 18:04:47 MarvinPA: I don't see why 18:04:50 -!- HangedMan has quit [Ping timeout: 265 seconds] 18:05:33 well if you're hitting an orc and it's nearly dead and then a new orc comes and stands next to you and autofight hits the new orc 18:05:39 then that is bad, and that's what old autofight did :P 18:05:40 alefury: anyway if there is great outrage when people try it, the prioritization can certainly be removed/simplified 18:05:53 (and what elliptic's doesn't, afaik) 18:06:21 MarvinPA: an injured orc should have a higher priority than a hale one, obviously. But a fresh orc priest needs to be picked over an injured grunt. 18:06:25 if I were making a bot that just spammed autofight, I would use much more complicated prioritization... but for default autofight, I don't think that's so good 18:06:39 right, i dunno if it re-prioritises for stuff like that 18:06:48 but that's probably not a situation i would continue autofighting in 18:06:50 it probably is an improvement, i just wanted to bring this up before someone adds lots of upredictable ai 18:07:25 kilobyte: that's a tricky decision really... currently it always prioritizes more injured monsters over less injured, and only uses danger level to decide between monsters at the same injury level 18:07:44 this seems to work pretty well IME, though I would make it more complicated if I were making a bot, again 18:08:51 ie, it will keep pounding that mostly harmless regular orc while the wizard/priest keeps taking half your hp per spell 18:09:13 kilobyte: are you suggesting special-casing orcs 18:09:33 since I should point out that orc wizard/priest have the same danger level as regular orc by the time you get to Orc 18:10:18 priests are more dangerous imo 18:10:29 in Orc they can't take half hp per spell 18:10:32 as I said, it works pretty well in practice... certainly there are improvements that could be made, but I wasn't sure about how complicated it should be made 18:10:35 but for autofight i think that shouldnt matter much 18:10:56 kilobyte: if you want to always prioritize danger over injury level, I can come up with lots of situations where that is a bad idea 18:11:08 -!- Textmode has quit [Ping timeout: 252 seconds] 18:11:09 they ignore EV and AC so are more dangerous than their xp suggests... but sticking with tiers seems to be a good enough approximation 18:11:19 btw, how finely is injury level accessible via lua? 18:11:26 same fineness as in-game 18:11:31 i.e. four levels 18:11:32 ah, good 18:11:34 elliptic: I'd use weighting rather than a set of rules 18:12:07 i wouldnt use weighting, because its unpredictable 18:12:10 kilobyte: you mean finish off that almost dead danger level 1 monster before starting on that danger level 2 monster, but don't do that if it is only heavily wounded? 18:12:41 elliptic: mostly, yeah 18:12:43 I considered that, but it is much more complicated and less predictable 18:12:49 alefury: and why would that be a downside? 18:13:07 i think even if autofight could more often make the "right decision", always making a predictable decision would be preferabale 18:13:51 i currently often use autofight to attack the monster on the left 18:13:52 if you're pounding the autofight key, you want it to make better decisions, not ones that are predictable -- since if you're micromanaging the fight you won't use autofight 18:14:06 when i want to attack the monster on the right, i dont use autofight 18:14:14 alefury: there is a key for that, you now :p 18:14:19 kilobyte: well, you want to know what sort of decisions it is making, so that you know what situations autofight is less reliable for 18:14:23 just pressing the same key a bunch of times is more convenient 18:14:42 like, autofight is great for being berserk currently, because you nearly always want to finish off a monster for the check at extra berserk duration 18:14:51 -!- evilmike has joined ##crawl-dev 18:15:21 and it is good in general for when you want to kill each monster before moving on to the next... situations where you aren't particularly more scared of one monster than the others 18:15:34 -!- Pacra has quit [Quit: This computer has gone to sleep] 18:16:06 I like the new autofight. Trying it locally 18:16:06 evilmike: You have 1 message. Use !messages to read it. 18:16:16 and if you are particularly scared of one monster (like a unique or an early orc priest), you probably don't want to hold down autofight against it with multiple monsters around anyway :) 18:16:20 I think my favourite part (even though its small) is being allowed to autofight snails 18:16:26 without having tried it, elliptic's version sounds better than old autofight, and it sounds like it is predictable, which is pretty much perfect imo 18:16:28 !messages 18:16:28 (1/1) elliptic said (8h 4m ago): The sacrificial humans in evilmike_hellhole interrupt resting and autoexplore for some reason, any idea why? 18:16:39 having tons of popcorn around shouldn't mean you need to stop autofighting 18:16:48 that's odd. do freed slaves interrupt autoexplore? 18:17:07 kilobyte: having tons of popcorn around and a nasty unique or an early orc priest should, though 18:17:18 (and it isn't popcorn then) 18:17:22 those humans are just flavour, but they have the neutral attitude which means they can still attack you... I guess thats why they interrupt it 18:17:23 evilmike: yeah, it's an insane improvement... my gripes are of the "it can be even better" kind 18:17:28 one situation where old autofight falls short is ynoxinuls 18:17:42 and it sounds like new autofight will not be nearly as bad in that situation 18:18:03 if the ynoxinul is next to you, it will be prioritized over uninjured ufetubi 18:18:08 kilobyte: anyway, I'd definitely be interested in any suggestions you have for modifications to improve it without making it much more complicated 18:18:57 alefury: yeah... I was also considering using summoned status in the prioritization (in fact my version on the servers uses that), but I removed it because it didn't seem clearly good to me 18:19:15 danger level is enough for vampire and ynoxinul spam 18:19:23 -!- jeanjacques has quit [Ping timeout: 252 seconds] 18:22:19 summon status is good but breaks strict prioritization 18:22:32 would work with weights, I think 18:22:51 well, the way I was doing it was only going after summons when nothing else is equally close to you 18:22:56 that is, prioritizing them last 18:23:07 on the theory that they might disappear before you got to them :) 18:23:15 this is clearly a bad idea in some situations, though 18:23:30 a lot of the time I'm not convinced it matters whether a monster is summoned 18:23:44 -!- valrus has joined ##crawl-dev 18:24:03 it matters which monster is _summoning_ but that's different 18:24:44 mhh, also thats info that is not usually available to all players 18:25:13 and shouldnt be revealed to lua (unless it is also revealed to players via monster description) 18:25:13 -!- HangedMan has joined ##crawl-dev 18:26:19 is summon duration randomized? 18:26:25 yes 18:26:56 monster summons tend to last fairly long though 18:27:19 a lot longer than it take sto kill them 18:31:22 doesn't matter if you're going to kill them all, does if you're likely to need to run 18:32:09 yeah... and if you are likely to need to run, why are you autofighting? 18:32:19 true 18:32:33 -!- jeanjacques has joined ##crawl-dev 18:32:37 on the other hand, autofight is nice for reaching and missiles 18:36:24 -!- ais523 has quit [Remote host closed the connection] 18:37:48 -!- edlothiol has joined ##crawl-dev 18:44:47 contrib/install/i686-w64-mingw32/lib/libSDLmain.a(SDL_win32_main.o):SDL_win32_main.c:(.text+0x608): undefined reference to `_SDL_main' 18:46:33 -!- syllogism has quit [] 18:49:39 even more interestingly, only in LTO builds 18:52:12 and not a toolchain issue -- 0.9.1 builds fine 18:53:21 -!- valrus has quit [Remote host closed the connection] 18:56:07 -!- agolden has joined ##crawl-dev 19:02:43 so, building TOT (or the stone_soup branch) on Mac OS X with the latest Apple dev tools appears to have been broken for a while. 19:03:37 I got it building last night, attached a small patch for the Makefile to this bug: https://crawl.develz.org/mantis/view.php?id=4157 19:05:27 er... so they drop support for even compiling for any but the newest-and-shiniest release? 19:07:00 even MSVC can still target Win95 (and obviously all free tools can do that too) 19:07:04 not what I meant to do there---I think it should now set SDK_VER to 10.4, 10.5, or 10.6, the earliest SDK that's present in /Developer/SDKs. 19:07:31 oh yea, Xcode 4 dropped everything but 10.6 19:08:07 ah, so your patch checks what is installed... sounds good 19:08:35 you can still get Xcode 3.x and the earlier SDKs, but Apple's pretty much moved on. 19:09:47 can't check this though... my compile farm includes SCO sux but not apples :p 19:11:12 Yak player tile (https://crawl.develz.org/mantis/view.php?id=5002) by minmay 19:13:12 ... 19:15:57 !seen greatzebu 19:15:58 I last saw greatzebu at Wed Nov 30 23:45:29 2011 UTC (3d 1h 30m 29s ago) quitting with message Quit: greatzebu. 19:18:21 agolden: do you happen to know what happens if SDK_VER is unset? 19:18:23 kilobyte: greatzebu handles the Mac builds at the moment, right? I wonder what he/she's done to resolve the "incomplete headers" issues on 10.4 that mistydemeo mentioned. 19:18:47 kilobyte: I'll give it a try 19:19:21 no idea, sorry :( 19:23:41 kilobyte: if SDK_VER is left undefined then SDKROOT will also be bad, so you need to get rid of the options "-isysroot" and "-mmacosx-version-min" used later. 19:24:24 kilobyte: seems to be building though... 19:24:33 -!- mikee_ has quit [Quit: Want to be different? Try HydraIRC -> http://www.hydrairc.com <-] 19:26:16 kilobyte: my guess is that it's equivalent to setting SDK_VER and SDKROOT to the latest that's installed, which is great for debug builds, not for universal release builds though. 19:27:46 -!- blabber has quit [Quit: leaving] 19:30:49 A two-headed ogre comes into view. it is wielding a giant club and a giant _spiked club. 19:30:58 the capitalization part's been fixed, right? 19:32:00 seems so 19:34:34 there are still capitalization issues with ghosts/illusions though 19:36:15 I noticed a weird capitalization issue a while ago. I messaged someone I was spectating, and it showed up as "Evilmike: message" 19:37:55 -!- neunon has joined ##crawl-dev 19:38:53 oooh... seems to be someone who can be harassed about Mac SDK issues :p 19:38:58 neunon: hi :p 19:39:17 oh hello. 19:39:27 what's up? 19:39:38 hey neunon 19:39:40 hi me, also 19:39:49 I was supposed to have time to do crawl stuff todqay :( 19:39:54 'lo due :) 19:39:59 but now I have a friend coming over to play twilight struggle! so I will forsake ye all :0 19:40:34 looks like the Fruits dropped support for compiling for any but the newest version 19:40:46 Yeah, had to work around that on my Lion machine 19:40:54 -!- Cryp71c has joined ##crawl-dev 19:40:58 Evening 19:41:08 agolden here suggested a fix in https://crawl.develz.org/mantis/view.php?id=4157 19:41:10 solution for me was to install Xcode 3 in a terrible way, and then install Xcode 4.2. 19:41:44 the patch looks good to me but 1. I can't test it, and 2. what if none of these directories exist? 19:42:09 like, a 10.7 SDK 19:42:21 it seems like a hack to me, for a couple of reasons 19:42:24 or some other toolchain that uses more standard directory placement 19:43:27 first, we make it so it either automatically uses the 10.4 SDK or whatever you tell it to (SDK_VER=whatever). with this patch applied, you'd have to see the verbose build output to know what you built against. that's sort of only a problem for release engineers but it's still annoying that we can't guarantee it generates a build that runs on anything 19:43:28 neunon: kilobyte: (it's totally a hack---also worth noting that it won't make it possible to (say) build a universal binary without the hassle of installing Xcode 3) 19:43:43 yeah, exactly 19:45:00 * neunon clones Crawl 19:45:31 mm. gitorious seems a bit slow. 19:46:13 neunon: what I'd really like is for there to just be a quick and easy way to do a build for testing using recent dev tools. 19:47:02 neunon: I'm not sure there's a better way to build the universal binary than installing both Xcodes. 19:47:22 yeah, totally agreed. I've got two macs I could use for this, one which works with the current build system (has both Xcode 3 and Xcode 4.2 installed on Lion, but that's a really non-standard setup). the other has the Apple-endorsed way, which runs only Xcode 4.2 19:47:30 I'll try on both and try to find something that "just works" 19:47:40 assuming, that is, that it ever finishes the git clone 19:51:05 I can guarantee that the only way to do a universal (i386+x86_64+ppc) binary is to have both Xcode installs. You basically have to have gcc 4.2 for the 10.5 SDK, and the current Xcode only includes llvm-gcc-4.2 and clang. but we should be able to at least build *something* with the current dev tools. 19:52:03 -!- st__ has joined ##crawl-dev 19:57:06 -!- st_ has quit [Remote host closed the connection] 19:57:08 -!- Gretell has quit [Ping timeout: 245 seconds] 19:57:08 -!- Gretell has joined ##crawl-dev 19:57:15 -!- st__ is now known as st_ 19:58:07 neunon: thanks for your help. 19:58:16 no problem :) 20:04:14 The basilisk tries to attack you, but flinches away. _You are wearing: F - a shield-shaped amulet (around neck) {tried, AC+4 EV+4} 20:11:18 very temporary workaround: 'make SDK_VER=10.7 CC=llvm-gcc-4.2 CXX=llvm-g++-4.2' 20:11:28 obviously not ideal, still digging. 20:21:04 neunon: if you don't set SDK_VER or SDKROOT (and don't include the "-isysroot" and "-mmacosx-version-min" options to the compiler) then gcc will just use sensible defaults. 20:22:32 Yeah, you can probably do 'make NO_APPLE_GCC=YesPlease' and it'll build 20:22:53 or, hm. I guess NO_APPLE_GCC was removed sometime ago 20:24:01 that was really an option at some point? 20:24:10 neunon: I was thinking something like that though, one option to wrap everything else make DEFAULT_SDK=YesPlease --> go down a simplified path that doesn't set SDK_VER, SDKROOT, etc. 20:24:15 with a legit value of "YesPlease"? 20:25:11 well, it was literally just 'ifdef NO_APPLE_GCC' 20:25:14 neunon: actually DEFAULT_SDK is a terrible name for that, I mean USE_DEFAULT_SDK or something like that. 20:25:15 I just use YesPlease for humor. 20:25:49 awww, i falsely attributed that humor to the apple people 20:26:09 nah, it was just part of our makefiles. apple are a humorlous bunch. 20:26:14 *humorless. 20:26:16 jeez. 20:28:02 GNU make is such an unflexible build tool. you pretty much -need- outside scripting in order for it to work. 20:30:33 -!- alefury has quit [Remote host closed the connection] 20:30:38 and that's with adding a whole bunch of flexibility over BSD make already 20:31:03 Yeah, sadly. 20:46:46 OK, think I've got a reasonable fix. Testing a few different build variants. 20:48:59 -!- jeanjacques has quit [Ping timeout: 252 seconds] 20:49:28 pushed. 20:51:05 03neunon * r90122a2781c1 10/crawl-ref/source/Makefile: Makefile: add NO_APPLE_GCC command-line option 20:51:15 03neunon * r30551d36c08f 10/crawl-ref/source/Makefile: Revert "Added make support for OSX Lion" 20:51:15 03neunon * r3a248b42287b 10/crawl-ref/source/Makefile: Makefile: make build work on Xcode 4.2 (Lion) 20:58:25 -!- agolden has quit [Quit: agolden] 20:59:03 -!- Cryp71c has quit [Ping timeout: 240 seconds] 20:59:38 -!- agolden has joined ##crawl-dev 21:04:09 -!- HangedMan has quit [Ping timeout: 265 seconds] 21:07:34 -!- HangedMan has joined ##crawl-dev 21:10:26 -!- MarvinPA has quit [Ping timeout: 252 seconds] 21:22:16 Xcode project seems broken too. might as well fix that next. 21:24:15 demonblade the Skirmisher (L1 MuCK) (D:1) 21:31:00 -!- agolden has quit [Quit: agolden] 21:43:20 -!- edlothiol has quit [Ping timeout: 260 seconds] 21:54:28 03kilobyte * raef483c2d12c 10/crawl-ref/source/main.cc: Unbreak Windows tile -fwhole_program builds. 21:57:44 -!- elliptic has quit [Read error: Connection reset by peer] 21:58:01 -!- elliptic has joined ##crawl-dev 22:00:39 -!- Cryp71c has joined ##crawl-dev 22:27:10 Incorrect prompt for casting refrigeration with friendly ice beasts in sight. (https://crawl.develz.org/mantis/view.php?id=5003) by elliptic 22:30:53 -!- Cryp71c has quit [Quit: Leaving] 22:36:51 -!- ainsophyao has joined ##crawl-dev 22:43:16 I thought YesPlease was just the conventional "yes", borrowed from (I suppose) the Linux tree? 22:52:04 -!- SamB has quit [Ping timeout: 260 seconds] 22:54:08 -!- SamB has joined ##crawl-dev 22:55:17 -!- medgno has quit [Quit: Quitting] 22:58:02 maybe it will refuse to compile if you are insufficiently polite in the makefile 23:15:03 -!- HangedMan is now known as HangedEye 23:18:40 -!- Wensley has joined ##crawl-dev 23:26:30 03dolorous * r084a1af8a443 10/crawl-ref/source/spl-damage.cc: Don't display the warning prompt for Ozocubu's Refrigeration for cold-immunes. 23:26:40 03dolorous * rc28662135877 10/crawl-ref/source/spl-damage.cc: More O's R fixes: no message spam or sanctuary breakage for cold-immunes. 23:26:49 <3 dolorous 23:39:07 Adeon: no, I'm pretty sure it's just a somewhat silly convention 23:39:21 it works just fine if I pass just y in for that kind of thing 23:39:42 (generally "no" would also work, but that would be evil ;-P) 23:46:44 does ozo's wake up sleeping stuff yet 23:47:07 also it'd be cool if someone wanted to code that partial-screen-flash thing, just saying 23:59:02 -!- valrus has joined ##crawl-dev