00:38:48 Kilobyte, re: the wall tiles. 1) the walls in the latest dev build are brightened up quite a bit, 2) the small screenshots in the tiles news update show the first dev release of the wall tiles, prior to that update, and 3) the bright white background of the webpage behind the screenshots make them look much, much darker than they look against the black background of the crawl client. 01:08:36 -!- eith has quit [Ping timeout: 260 seconds] 01:14:31 -!- ogaz has quit [Ping timeout: 240 seconds] 01:59:16 -!- Zaba has quit [Ping timeout: 248 seconds] 02:01:47 -!- Zaba has joined ##crawl-dev 02:08:15 -!- Siber has quit [Ping timeout: 258 seconds] 02:13:03 -!- Kil2 has quit [Read error: Connection reset by peer] 02:18:25 -!- MarvinPA_ has joined ##crawl-dev 02:21:45 -!- MarvinPA has quit [Ping timeout: 240 seconds] 02:21:49 -!- MarvinPA_ is now known as MarvinPA 02:32:28 Demonspawn sharp spines lvl 1 are useless (https://crawl.develz.org/mantis/view.php?id=1985) by serg271 02:36:39 Moin! 02:36:39 Napkin: You have 5 messages. Use !messages to read them. 03:45:57 Napkin <3 03:46:02 !coffee Napkin 03:46:02 * Henzell hands Napkin a mug of irish coffee, brewed by the Serpent of Hell. 03:46:07 hey, moin greensnark :) 03:46:11 how's it going? 03:46:13 !coffee greensnark 03:46:14 * Henzell hands greensnark a mug of irish coffee, brewed by Crazy Yiuf. 03:46:14 Good, good 03:46:33 I'm way ahead of schedule and essentially have three days off :) 03:46:41 Also I finally have a gratis iPad 03:46:48 Talk about overrated hardware :P 03:47:22 woah! 03:47:29 it's evil ;> 03:48:12 I should jailbreak it :P 03:49:38 gratis? How did you get it? 03:50:10 and, things you have to jailbreak are good for nothing 03:50:41 overrated or overheated? 03:50:44 I have a htc desire. 03:50:53 just got my N900, and I'm installing Crawl's build-depends as we speak 03:51:09 I don't think I can build crawl for Android. 03:51:12 That would be cool though :) 03:51:29 KiloByte: It's a customer's device 03:51:37 I do freelance iPhone dev 03:51:55 ! 03:51:59 * due makes the sign of the cross at ge. 03:51:59 so, when can we expect iDCSS? 03:52:01 ... greensnark. 03:52:09 bhaak: someone is alreayd working on an ipad port 03:52:22 bhaak: Too much work :) 03:52:24 I wouldn't be surprised if greensnark is faster :D 03:52:37 I like Crawl just fine on laptop and desktop 03:52:42 This thing is a pain to use 03:52:43 Well, their's is tiles port 03:53:30 You want to hold it in one hand you better have a grip of IRON, because it's pretty heavy and the gripping surfaces have been coated with teflon 03:53:51 So then you hold it in both hands and have to stretch your fingers to use the screen :P 03:53:54 the default terminal on Maemo is horizontal on a wide (800x480) display, and it has a fat menu on the top and a fat toolbar on the bottom 03:54:25 do you have the wifi or 3g version? 03:54:35 the 3g version feels pretty heavy 03:54:35 bhaak: The customer is a cheapskate, guess which version :P 03:54:39 lol 03:54:52 I still need to figure out how to make it sane, to get a console without 353479572398x2 window 03:55:13 KiloByte: What's the default term size ? 03:55:34 KiloByte: Also, there was a printf format issue with the turncount display, if you want to fix that quick :) 03:56:05 ixtli is working on an iPad port 03:56:27 bhaak: got both UMTS (and lesser crap) and Wifi 03:56:42 Also generalizing a lot of he code to make it easier to port to other platforms too, if I understand correctly 03:57:43 didn't put a SIM card into it yet so I didn't test phone capabilities (and internet access costs a fortune here), but wifi works just fine 03:57:46 Keskitalo: Yes, he did a lot of tiles refactoring 03:59:09 Tiles + touchscreen interface is an interesting thing.. the input/interface side of Crawl could overall benefit from a little generalization too :) 03:59:10 greensnark: I fixed it at 3am yesterday, but went to sleep. I replaced all %ld/lu with either %d or PRId64, and it seems to work. 03:59:19 greensnark: except for some fsim stuff 03:59:37 fsim was born for a life of pain, we can let it fulfil that destiny for a little longer 03:59:43 the ipad is surely a nice device but currently I just don't have a usecase for it. I wouldn't want to take it with me, watching videos I do at home or on the road on a psp. so what's left? 03:59:46 greensnark: do you know if it indeed needs longs there? How many testing turns can one expect at the most? 03:59:50 how about we just stop using printf 04:00:10 KiloByte: It used to be 40k, but Crawl combat has become about 10x slower since 0.1 04:00:34 Keskitalo: Hey there. 04:00:37 ah, I thought you may set it to hundreds of millions of turns 04:00:40 stdargs = suck 04:00:45 KiloByte: I'm excited about Maemo support too, which feels a little silly as I dont' have a Maemo device 04:00:49 It's actually on hold until there is a 4.0 sdk 04:00:50 The Pentium M I tested on in the 0.1 days was able to do 40k fsim turns about 5x faster than my current Core 2 quad 04:00:55 Cause its gonna change the architecture. 04:00:59 I do have DS though :) (but no homebrew kit) 04:01:07 this is one of the few places where I support using C++ features 04:01:08 But I contributed most of the relevant changes back to master already, anyway :) 04:01:08 sorear: why? C++'s overloading sucks, but stdargs works just fine. 04:01:29 KiloByte: quick, what's the portable format modifier for ssize_t 04:01:32 ixtli: Ah, good to know! Yeah, I saw that there was a lot of work done already. 04:01:46 I haven't actually looked at the code though :) 04:01:46 if you get it wrong, you pay with a runtime crash on some obscure platform 04:01:54 sorear: there's none, just as for most types. 04:02:12 sorear: I want those obscure platforms named! :) 04:02:17 bhaak: The iPad screen also looks disastrously grainly compard to iphone 4 04:02:33 bhaak: win32 with msvc libc 04:02:50 greensnark: lol. yeah, I heard that, too. haven't seen a live iphone4, though 04:03:04 i have a trove asking me for a scroll of vorpalize weapon 04:03:14 and it won't let me in because all i have is a scoll of vorpalise weapon :( 04:03:37 MarvinPA: argh, bug 04:03:41 MarvinPA: unfixable too, sorry 04:03:48 damn :( 04:03:51 bhaak: but usually it just junks the upper half of %ld unless the compiler has it always zeroed. gcc-4.0 and gcc-4.4 -O0 have it always 0, gcc-4.4 with -O2 does not. 04:04:33 Blame dpeg 04:04:59 -!- JamezQ has quit [Quit: Leaving.] 04:05:00 KiloByte: what happens if you cast it first? (long)ssize_t 04:05:11 that's why I'd want Napkin to build the official Windows binaries, even though with this commit I'll remove all traces of %ld 04:05:38 greensnark: Can you cherry-pick 5696672dd57be7a9c0cf into 0.7? Thanks. 04:05:47 (Still pushing0 04:06:11 03due * rca75baaf1934 10/crawl-ref/source/dat/des/portals/trove.des: Fix bad ize/ise on vorpalise with Troves. 04:06:28 bhaak: it's still (long). And apparently it somehow doesn't work on win32 in some cases. 04:06:37 ah, no need for me to mantis it at least then :P 04:06:49 KiloByte: should probably be (unsigned long). i'd have to check the standard on that. c sucks 04:07:11 bhaak: signed for %d, unsigned for %u 04:07:53 bhaak: but the bug lies in the standard declaring that "%ld" means long, and msvc's libc having it mean "long long". 04:08:33 bhaak: but fortunately C99's types do work as they are supposed to, and all of our C++ compilers do support them 04:08:43 hmm, and how do you format 'long long' variables? %lld? 04:09:01 oh, wait 04:09:07 bhaak: so printf(PRIu64, (uint64_t)foo) works even there 04:09:14 %ld means 'long long' ? 04:09:18 in msvc 04:09:38 bhaak: it appears to do so, at least 04:10:10 why o why? 04:10:13 bhaak: but if we remove all uses of "%ld", it won't matter the slightest what it means 04:10:52 bhaak: remind me, what company does msvc libc come from? What is their stance on portability and standards support? 04:10:55 but you get compilation warnings 04:11:20 I don't answer rhetorical questions : 04:11:41 I don't claim I fully understand this issue... 04:12:02 and it eludes me why it worked with very old gcc. 04:13:54 when it was first mentioned here, I tested it in cygwin and mingw32. cygwin worked correctly, but mingw32 not. the old gcc probably linked to another lib 04:17:42 due: Real soonish now 04:18:19 hrm, what's up with @#$%^&* sourceforge? pulls keep failing all the time 04:20:05 hehe, I have never problems pulling :) 04:23:22 -!- purge has joined ##crawl-dev 04:23:59 I don't have problems pulling 04:24:05 instead, I have problems pushing 04:24:24 so I spend 4 hours working on a patch, then it gets delayed 3 days, and duplicate work! 04:24:51 greensnark: That's fun, just a rather annoying bug 04:28:02 'x>' shouldn't consider current square when multiple > are on the screen (https://crawl.develz.org/mantis/view.php?id=1986) by Nexos 04:33:47 -!- Mu_ has joined ##crawl-dev 04:34:56 DCSS definitely needs an Android port 04:35:28 Maemo/Meego will die sooner or later.. 04:35:38 -!- syllogism has joined ##crawl-dev 04:35:50 Napkin: when I was young, we thought the same about DOS :) 04:35:57 Yeah, I'd actually be interested in an Android port 04:36:22 too few provider do Maemo/Meego - while Android is growing like crazy 04:36:45 bhaak: It would be nice if livelog could record game start time 04:37:15 To make it easier to match back to the corresponding game, I mean 04:37:49 Also recording the place would be nice 04:38:18 How about recording things like gender and alignment as well; max information possible per livelog entry? :) 04:38:19 KiloByte, i switched my motorola milestone with an n900 - solely because of the solid OS below 04:39:09 greensnark: wrong channel 04:39:17 bhaak: You worried about that? :P 04:39:30 yes :) 04:39:37 !coffee bhaak 04:39:37 * Henzell hands bhaak a cup of black coffee, brewed by Ijyb. 04:39:38 Ok, let's take it to #Unnethack 04:39:50 03by * r4bde4b17c7e2 10/crawl-ref/source/effects.cc: Remove random staff ID over time. 04:46:53 with n900's keyboard, playing console crawl sucks :( 04:47:52 I don't even 8have* a keyboard. 04:50:24 it has only up, left, down, right arrows, so I'd have to use vi keys 04:50:39 but laptops suck exactly the same way too 04:50:44 and that's why I don't use laptops 04:54:47 hehe 04:54:52 true 04:55:18 but, the longer you try the better it works 04:55:27 The numberpad is a ploy to make you waste a whole hand on the wilderness to the east of the Real Keyboard 04:55:42 I have a numberpad on my desktop and I never use it 04:55:53 It's meant for Satan worshippers, obviously 04:56:05 Possibly for people allergic to cats too 04:56:17 like blinkers in cars ;> 04:56:18 Cat lovers should avoid the numpad at all costs 04:57:39 -!- MarvinPA_ has joined ##crawl-dev 04:58:15 the true abomination is the old "grey" set of keys that duplicate most of proper arrow keys 04:58:29 s/old/new/ 04:58:53 good old keyboard didn't have that crap 04:59:21 it appeared in IBM AT only because some bozos insisted on having numlock _on_ 04:59:36 numlock is another abomination that should be ripped from all keyboards 04:59:46 Just saw off that entire section of keys, I say :) 05:00:12 And make sure you wash your hands thoroughly if you accidentally touched the numpad :P 05:00:15 does anyone play crawl with a joystick? 05:00:17 the problem is, it's between two usable sections: the main part and the numpad 05:00:30 Haha, by 05:00:34 No, it's very useful to have around 05:00:41 i use a flightstick 05:00:47 The numpad holds my keyboard up 05:01:00 If I sawed it off, it would be too narrow to hold comfortably 05:01:28 by: I still have two 1980 joysticks, although one is in parts that are probably incomplete after I tried to repair it ~10 years ago, and the other one was never ever used 05:01:49 -!- MarvinPA has quit [Ping timeout: 276 seconds] 05:01:49 -!- MarvinPA_ is now known as MarvinPA 05:01:49 I never touched any of newer stuff, though 05:02:14 which may indeed be better, who knows... perhaps the technology advanced a bit? 05:02:57 I wish I still had a tac-2 05:03:08 I wonder what genius came up with the name "joystick" 05:03:13 It's right up there with "Wii" 05:03:22 KiloByte: there's newer stuff that is still good: http://www.speedlink.com/?p=2&cat=314&pid=21424&paus=1 05:03:38 "The name "joystick" is thought to originate with early 20th century French pilot Robert Esnault-Pelterie" 05:03:45 :P 05:03:52 what a surprise 05:03:58 -!- casmith789 has joined ##crawl-dev 05:04:03 a french coined the name *joy*stick 05:05:10 bhaak: unlike keyboards, I don't have a preference for old joysticks, I simply never had an urge to use one for ~25 years 05:06:00 the last time I used one was for Freespace 2 05:14:43 -!- dpeg has joined ##crawl-dev 05:14:48 hey ploog 05:15:06 Good day Mr Brown 05:17:28 FR: "Something about spiny scales isn't perfect. I suggest to remove them." 05:17:38 I suggest to remove players. 05:19:01 Close it and ignore, basically. 05:19:24 I replied once, so that Cryp71c can see it. 05:19:43 Close and ignore == fire and forget? :) 05:20:15 make it like ubuntu does: mark as "opinion" and close :) 05:21:01 bhaak: <3 05:21:12 "Ill-conceived, misguided opinion." 05:22:00 Can we have a "Smoking Crack" resolution in Mantis 05:22:22 Hey lobsterman :) 05:22:41 scrollback should be cleaned, when you start a new tiles char 05:22:58 Napkin: Old game messages linger? 05:23:04 yeah 05:23:13 There's always one more game-restart bug :) 05:23:14 I see the "you sure want to quit?" and everything else 05:23:27 The game doesn't like people who quit :P 05:23:32 hehe 05:23:33 not in ctrl+p though, only on screen 05:24:22 dpeg: I hear you've been fireballing my javelineers? 05:24:28 rc1 tiles do look a bit dark - but apart that awesome! 05:24:38 love the new water and walls! 05:24:43 "Do you really want to quit? (y/n)" User: y. "Y or n, please." User: Y. "Are you sure? (yes/n)" User: yes. "I don't think you should quit. Quitters never win. Surrender denied." 05:24:56 greensnark: yes, but they retaliate with crashes! 05:25:13 That javelineer had nuclear javelins. 05:25:22 does autoinscribe = scrolls? of identify:@r1 05:25:23 autoinscribe = scrolls? of remove curse:@r2 05:25:23 autoinscribe = potions? of healing:@q1 05:25:23 autoinscribe = potions? of heal wounds:@q2 05:25:29 omg! sorry for spam 05:25:35 aggh! an angband player! 05:25:35 damn windows copy & paste 05:25:49 does tile_better_transparency still do anything? 05:26:11 It activates the placebo effect to make you think the tiles look better 05:26:37 lol 05:26:58 no more in use then? mystery option :D 05:27:27 You should ask Enne 05:27:40 03pointless_ 07stone_soup-0.7 * rd5f161ac7dda 10/crawl-ref/source/mon-abil.cc: [1951] Fix enchantment duration averaging during slime merges 05:27:42 03pointless_ 07stone_soup-0.7 * r8697e83b9222 10/crawl-ref/source/ (abl-show.cc godabil.cc): [1971] Fix canceling sunlight at the prompt still costing 05:27:46 03pointless_ 07stone_soup-0.7 * r5960ff90692f 10/crawl-ref/source/ (beam.cc mon-stuff.cc): Actually fix death messages for giant spore/ball lightning kills 05:27:47 03due 07stone_soup-0.7 * rce848880e30a 10/crawl-ref/source/dat/des/portals/trove.des: Fix bad ize/ise on vorpalise with Troves. 05:28:27 -!- ortoslon has joined ##crawl-dev 05:30:01 -!- Celentes has joined ##crawl-dev 05:32:11 * KiloByte types "time make debug" on the phone. Now we'll see the truth. 05:41:45 03kilobyte * r956b740cff3b 10/crawl-ref/source/ (29 files in 2 dirs): Get rid of more longs and all "%ld"s, new gcc + Windows + them = bad mojo. 05:54:34 So this weird turn count thing is only on Windows? 05:54:37 I can't get a repro here 05:56:19 windows _and_ semi-modern gcc 05:56:21 Yeah, bad format for num_turns 05:57:09 Let me pick that fix 05:57:11 in non-optimized builds the top part tends to be zeroed, so it doesn't show up there as well 05:57:11 Yes, but should be fixed anywayz 05:57:13 I'd prefer Napkin to build the official binary using the ancient version of gcc he uses for Windows builds, since the msvc libraries it links again seem to don't have it problems 06:01:07 actually, not that ancient, KiloByte: 06:01:08 root@ipx14623:~> i686-pc-mingw32-gcc --version 06:01:08 i686-pc-mingw32-gcc (GCC) 4.4.0 06:01:18 -!- Pseudonut has quit [Remote host closed the connection] 06:01:32 been using that for quite a long time already 06:01:53 since I switched from etch to lenny 06:05:56 i686-pc? 06:06:53 03kilobyte 07stone_soup-0.7 * rd1a87e761c23 10/crawl-ref/source/ (29 files in 2 dirs): Get rid of more longs and all "%ld"s, new gcc + Windows + them = bad mojo. 06:06:53 Napkin: Can you add a build for the new Windows installer to the update scripts on CDO? :) 06:08:19 Napkin: there are two mingw32 packages: mingw32 (4.2.1.dfsg-2) and gcc-mingw32 (4.4.4-0.1), and both exhibit that problem for me 06:08:29 Napkin: Also would you mind greatly if I tossed some of those "press RETURN to continues"? :) 06:09:02 the prefix is i586-mingw32msvc in both cases 06:09:48 KiloByte: I know, that why I had to dump the debian packages and install manually, guided by these instructions: http://www.nongnu.org/mingw-cross-env/ 06:10:22 greensnark: why don't you just add a useless parameter to not be asked to press any key? ;) 06:10:34 Napkin: Ok! 06:10:43 I was actually considering an expect script to press Return :P 06:10:50 hehe 06:10:51 Been a long time since I messed with expect :P 06:11:06 if test -n "$1" 06:11:06 then 06:11:06 WAIT_KEY=0 06:11:08 fi 06:11:12 :D 06:11:22 Eino complained quite a while back 06:12:17 Süßkraaaam! \o/ 06:12:25 :P 06:15:12 Napkin: that explains the issue, your toolchain is different from mine 06:16:19 "rotting rotting hulk corpse" -- how would you name it? "rapidly rotting ..."? 06:16:41 rotting^2 hulk corpse 06:19:14 Why is dpeg filing bugs with a Jun 30 build? 06:19:31 Erm 06:19:35 old people like old things? 06:19:39 I was cursing and trying to replicate the zombie beam bug and then I look at the date and realise it's all dpeg's fault :P 06:20:04 Erm^2 06:20:07 damn, no süßkram! 06:20:24 btw, have you decided on a codename for the next release yet, greensnark? :> 06:20:38 "This space intentionally left blank" 06:20:41 Too wordy? :) 06:20:59 Some long term from the ayurvedas? 06:21:11 How about "m00se bites are very painful, you know" 06:21:25 Sorry, I feel in a very flippant mood today :) 06:21:26 one of those cool long parts vimanas are made of? 06:21:48 "Amazing Avocado" 06:21:55 "Next Release". wait 06:21:55 We need to parody Ubuntu release names 06:22:45 isn't "Stone Soup" enough? 06:22:54 Obviously not 06:23:25 The treads are worn right off "Stone Soup". You drive it at high speed, you're begging for a blowout 06:26:03 "Amazing Avocado" is awesome! ;D 06:26:15 :) 06:26:33 hrm, "rapidly rotting hulk corpse" is hard to parse 06:26:45 "decaying rotting hulk corpse"? 06:26:54 "decomposing rotting hulk corpse" 06:27:15 I still like rotting^2 :P 06:27:23 "amazing rotting hulk corpse" 06:27:28 :P 06:27:48 Hey, we have exploding darts that explode, a little tautology never hurt anyone 06:28:08 sooo.. what do I need to add to get these new high-tech windows installer built? 06:28:38 sounds like a case of `make package-windows` 06:28:43 Napkin: you just used a self-contradiction 06:28:47 * KiloByte nods at greensnark. 06:28:55 just tack CROSSHOST= at the end 06:29:18 with the name of your compiler prefix 06:29:25 * Napkin looks suspiciously * 06:31:07 what will the filename be? 06:32:13 tiles or non-tiles? or does it expect both binaries? 06:32:27 Outfile "crawl-${VERSION}.setup.exe" 06:32:51 pardon? 06:32:53 it does the build twice, you don't need to specify anything 06:33:16 Napkin: if you want to change the filename, edit util/crawl.nsi and change that line. 06:33:40 i'm trying to prevent changing anything 06:33:40 it currently produces files like crawl-0.7.0rc1.setup.exe" 06:35:16 my command like is just: make package-windows CROSSHOST=i586-mingw32msvc 06:37:57 greater rotting hulk corpse 06:39:24 also, "rotting chunk of rotting hulk flesh" 06:40:49 could make rotting hulk corpses start in rotting state, and just drop the extra "rotting"? 06:41:06 I like by's idea 06:41:14 and rot away that much faster, right? 06:41:24 yes, why not 06:42:49 is there any important difference between rotting and non-rotting "red" corpses? 06:42:59 aside the time they have left 06:43:05 for ghouls, I guess 06:50:39 -!- Ekaterin has joined ##crawl-dev 06:58:57 -!- dpeg_ has joined ##crawl-dev 06:59:40 is -DSAVE_DIR_PATH=\"~/crawl\" ignored in windows, KiloByte? 06:59:59 Napkin: It saves in the user's home dir on Windows, I think 07:00:22 it's running right now 07:00:32 X:\Users\cowboy in the recent Windowses, some insanely complicated path on older Windows 07:00:38 seems to me, I can no longer sort, rename and add stuff - like the changelog. 07:01:02 Oh, hm 07:01:11 let's wait and see 07:01:14 Can you drop it into the directory structure at the right place before the build? 07:01:19 no 07:01:33 target does a rm -rf on the build directory first 07:01:45 No, in the Crawl source tree 07:02:04 I think all docs should be copied 07:02:40 well 07:03:36 admitted, it's not needed anymore - but where would I put crawl-sprint.bat? 07:03:59 Hmm 07:04:15 licence.txt should be renamed to LICENCE.txt, according to CREDITS.txt and README.txt 07:04:40 changelog is no problem 07:05:01 let's wait and see, what exactly is copied 07:05:04 next try... 07:05:08 Süßkram! ;) 07:05:18 (Very) ugly things no longer mutate due to proximity (https://crawl.develz.org/mantis/view.php?id=1987) by dolorous 07:05:19 but, one annoying thing 07:05:23 my -j2 was ignored 07:05:43 my 07:05:44 nice make -j2 -C source CROSSHOST=i686-pc-mingw32 CHOWN=true CHMOD=true USE_MERGE_BASE="${PRIMARY_BRANCH_REMOTE}" TILES=y WIZARD=y DESTDIR=${INSTALLDIR} package-windows 07:05:47 resulted in 07:05:52 make build-windows 07:05:52 rm -rf crawl-win 07:05:53 make TILES=y DESTDIR=crawl-win SAVEDIR='~/crawl' install 07:05:54 --- 07:07:49 That's weird $(MAKE) is supposed to pass MAKEFLAGS automatically 07:13:11 -!- dpeg has quit [Quit: Lost terminal] 07:15:01 trying to put TILES=y can have weird results. Same for DESTDIR. 07:15:38 actually, both TILES and DESTDIR get overridden 07:17:22 you can't redirect nsis' output file somewhere else; perhaps a hack with -Doutdir=XXX could work but it would produce errors when not specified 07:18:20 greensnark: you didn't pick my fix for ?/m gold yet, right? Good. 07:18:39 it causes crashes on reload due to the temp monster's item being unlinked 07:18:40 KiloByte: I didn't even notice that one :) 07:19:12 What was the commit? :P 07:19:16 ah, I didn't push it yet, good. You didn't heard that line, it was all your imagination. There is no such commit. 07:19:34 :P 07:19:45 I guess SF rejecting your push finally came in handy :P 07:20:37 KiloByte, the rm -rf on crawl-win at the beginning of the build is not so nice 07:20:51 that's what a make clean/distclean is for 07:21:47 except that I need to nuke that dir before the first build but not before the second 07:21:58 but yeah, I may put "make clean" there 07:22:21 did someone fix that turn count thing 07:23:05 syllogism: Yes 07:23:52 oh, don't see the commit for some reason 07:24:12 That's becaue KiloByte disguised his commit message 07:24:31 He's like the Voodoo lady from Monkey Island 07:24:43 I named it for the cause, not one of symptoms. 07:24:49 oh that one, hadn't pulled it yet 07:25:05 and couldn't reproduce it in 0.8 07:25:31 03kilobyte * r3a57810f6f89 10/crawl-ref/source/ (effects.cc itemname.cc itemprop.cc mon-stuff.cc mon-stuff.h): Make rotting hulk corpses pre-rotten. Drop rotting^2. [1984] 07:27:40 @?? hellephant 07:27:40 hellephant (04Y) | Speed: 10 | HD: 20 | Health: 140-200 | AC/EV: 13/10 | Damage: 45, 20, 15 | Flags: 05demonic, !sil | Res: 06magic(133) | XP: 3634 | Sp: fire breath (3d40), blink. 07:27:50 are they supposed to leave corpses 07:29:39 -!- dpeg has joined ##crawl-dev 07:29:55 other demonic animals do leave them, so I guess that yes 07:30:22 9 chunks of demigod flesh mmm 07:31:30 syllogism: you created multiple demigods, right? 07:31:53 was acquiring food as a ghoul 07:32:05 ... 07:32:07 heh 07:32:33 dummy monsters probably should be exempted :p 07:34:21 the N900 build took 69 freaking minutes :( The CPU is fast, but with all the memory filled by the crap that's running by default, it was a massive swapfest :( 07:36:13 Napkin: should I allow redirecting the nsis outfile somewhere? If we assume the .nsi is always called from the makefile, -Doutdir= being mandatory is not that bad. 07:36:43 well, I don't know 07:37:21 this is the file I got: crawl-0.7.0-rc1-16-geaae244.setup.exe 07:38:10 that name is fine - to be conform to the other stuff, i guess it would be good to include -win32 and .setup. is not necessary 07:38:32 Napkin: Players might confuse it with the game executable name otherwise? 07:38:47 Not that we use version in the game executable name 07:38:49 README.txt, LICENCE.txt, README.PDF and CREDITS.txt is missing, KiloByte 07:39:00 Bad KiloByte :) 07:39:12 -!- dpeg_ has quit [Quit: Lost terminal] 07:40:25 Napkin: I copy recursively the whole dir, so they must have been somehow skipped by "make install" 07:40:38 they always are, KiloByte 07:40:53 skipped, I mean 07:46:10 was the WIZARD=y parameter actually passed to child makefiles? 07:48:05 !tell Enne Minor issue -- When I start rc1 tiles client in Windows 7, a tiny amount of the bottom of the crawl window is hidden behind the task bar(just enough that i need to move it up a little) 07:48:06 purge: OK, I'll let Enne know. 07:48:40 no, KiloByte, it wasn't 07:49:05 -!- dpeg_ has joined ##crawl-dev 07:49:27 My connection is so damn unstable today, don't expect me to see me later :) 07:49:48 KiloByte: anyways, this missing of the textfiles from the main directory is the reason i'd like you to remove the rm -rf crawl-win, please? 07:51:04 -!- MarvinPA has quit [Read error: Connection reset by peer] 07:51:09 Napkin: uhm, I ran it with WIZARD=y V=y and it does get there 07:51:34 -!- MarvinPA has joined ##crawl-dev 07:51:55 KiloByte: not for me, KiloByte :-O 07:52:11 i got 07:52:25 make TILES=y DESTDIR=crawl-win SAVEDIR='~/crawl' install 07:52:26 and 07:52:27 what clean does _is_ rm -f for all files, and -rf for all dirs 07:52:32 make TILES= DESTDIR=crawl-win SAVEDIR='~/crawl' install 07:52:51 yes, clean 07:53:01 but i didn't call target clean, i called windows-package 07:53:20 --- 07:53:22 make build-windows 07:53:22 make[1]: Entering directory `/home/services/crawl/source/mingw32/crawl/crawl-ref/source' 07:53:25 rm -rf crawl-win 07:53:28 make TILES=y DESTDIR=crawl-win SAVEDIR='~/crawl' install 07:53:30 --- 07:53:53 build-windows called by package-windows 07:54:48 my exact command line: make package-windows CROSSHOST=i586-mingw32msvc WIZARD=y V=y 07:54:51 apart from it blocking the ability to add content before the build, it's generally a bad habit to call clean in a creation target - isn't it? 07:55:30 the alternative would be to copy everything to a temp dir 07:55:53 i don't get - why not just create crawl-win and put the stuff there? 07:56:05 instead of "deleting folder, create folder, put stuff there"? 07:56:25 but even then, it has to be cleaned if it has to be rebuilt 07:56:42 if you rebuild, you do a "make clean" first? 07:57:24 or maybe a separate clean target, if you want to keep the object files of everything else? 07:57:37 the whole rest of the makefile, both neunon's weirdness and standard automake ones, rebuild things when a dependency is altered 07:58:09 yes, and? 07:58:11 -!- dpeg has quit [Quit: Lost terminal] 07:59:26 ok, trying without "TILES" and "DESTDIR" this time 07:59:54 again no WIZARD=y, KiloByte 08:00:21 nice make -j2 -C source CROSSHOST=i686-pc-mingw32 CHOWN=true CHMOD=true USE_MERGE_BASE="${PRIMARY_BRANCH_REMOTE}" WIZARD=y package-windows <- anything still wrong? 08:00:55 where are you looking at, at the compile flags or at what is echoed? 08:01:18 what is echoed 08:01:31 --- 08:01:32 you need to use V=y if you want to see the actual output 08:01:39 nice make -j2 -C source CROSSHOST=i686-pc-mingw32 CHOWN=true CHMOD=true USE_MERGE_BASE="${PRIMARY_BRANCH_REMOTE}" WIZARD=y package-windows 08:01:45 make: Entering directory `/home/services/crawl/source/mingw32/crawl/crawl-ref/source' 08:01:47 or run "ps axw" 08:01:48 make build-windows 08:01:51 make[1]: Entering directory `/home/services/crawl/source/mingw32/crawl/crawl-ref/source' 08:01:54 rm -rf crawl-win 08:01:56 make TILES=y DESTDIR=crawl-win SAVEDIR='~/crawl' install 08:01:59 --- 08:02:02 you are kidding me, right? 08:02:53 it works for me (as in: builds a wizard-enabled build), shows the flag if I add V=y, shows up in "ps axw" 08:03:28 and indeed, doesn't get echoed, like all other variables inherited from parent make 08:04:11 ok *sigh* 08:04:21 breathe flames should not "bolt" through multiple enemies its a little ridiculous 08:04:51 maybe it "bolts" around them 08:06:10 Well, somehow it manages to kill them as well anyways 08:07:04 Flame breathe:2 on demonspawn just line up gnoll/orc packs early game and 1 shot them all potentially -.- 08:07:45 -!- dpeg_ has quit [Quit: Lost terminal] 08:28:52 -!- Celentes has quit [Ping timeout: 276 seconds] 08:29:58 -!- Textmode has joined ##crawl-dev 08:34:04 anyways, this needs to be done first before the installer should be shipped: 08:34:21 - rename licence.txt in main directory to LICENCE.TXT 08:34:49 - rename licence.txt in main directory to LICENCE._txt_ 08:35:03 - add CREDITS.txt, README.txt, README.pdf and LICENCE.txt to the installer 08:35:09 --- 08:35:47 Napkin: Thanks for testing the installer <3 08:35:50 03greensnark * rb7cedc63f4d4 10/crawl-ref/source/ (l_crawl.cc macro.cc macro.h main.cc): Fix Aliza's macro crash (!lm Aliza type=crash 1) 08:36:30 oh, and it would be nice, if the -jN could be passed on internally, while creating the installer 08:36:34 "make install" handles only the licenses for contribs (docs/license/) 08:37:24 yes, we always copied them manually for releases and other zips, KiloByte 08:37:25 it is, that's what +make does -- and merely passing -j would make the number of jobs explode 08:38:06 in fact, if you use +make, it forks a helper daemon named "jobserver" which manages submakes 08:38:24 (ok, it is forked on any -j, not just +make) 08:39:03 and sadly, what is echoed doesn't show the + 08:39:05 you mean i just missed that my -j2 was actually being active? i'll check again 08:39:23 you can definitely tell my speed 08:39:52 need to disable caching for that 08:40:11 both builds take 5m together for me, without -j4 that would be half an hour 08:40:22 and I use no ccache 08:41:48 hmm, an interesting warning on arm: hints.cc:208: blah blah case label smaller than the minimal possible value for the type 08:42:14 and we compile with -fsigned-char, so it's not the cause 08:44:59 arg - you added /source/crawl-win to .gitignore?! 08:45:38 yes, we definitely don't want a staging dir in the repository 08:45:52 and neither make clean nor distclean removes the folder 08:46:18 the last issue is a bug, already fixed but not committed 08:47:43 erm.. what am I doing wrong, that a make distclean doesn't remove source/*.{d,o} ? 08:48:42 ok, TILES=y was missing 08:49:43 please have it also remove the *.setup.exe, KiloByte 08:51:38 git clean -dfx \o/ 08:51:42 nevermind, KiloByte 08:53:14 oh, one more important thing, KiloByte - I need it to not ignore USE_MERGE_BASE 08:53:39 that's very important, otherwise it'll count my local commits in versioning too 08:56:58 -!- MarvinPA has quit [Remote host closed the connection] 08:57:50 -!- Spads has joined ##crawl-dev 08:58:28 -!- Spads has left ##crawl-dev 08:59:19 pox on those who say "licence" instead of "license"! :p 09:04:10 -!- Cryp71c has joined ##crawl-dev 09:06:11 Napkin: USE_MERGE_BASE appears to work, at least in build.h. 09:11:50 oh, but the VERSION is different 09:12:10 i had -16 instead of the -14 09:14:41 same bug happens for "package-source" 09:17:43 perhaps I should rename "crawl-win" to something like "build-win" to be less generic? "package-source" uses "build". 09:18:30 build-win sounds good, or maybe even just build 09:19:30 btw, do we still need/want to deliver crawl-sprint.bat ? 09:20:00 I doubt that, the menu allows you to choose sprint now 09:21:23 I' 09:22:00 me too 09:22:28 I'll keep build-win, since unlike the source, it makes sense to leave that dir after build. There's little that can fail during zipping and tarring, but it's likely we'll play with nsis settings. 09:29:14 works now, I built it with USE_MERGE_BASE=0.6.0, and it created crawl-0.6.0-a2.setup.exe 09:29:46 (-a2 being the forking point between master and 0.6) 09:34:52 commit, i'll try :) 09:35:29 WTF... an optimized build on my N900 was _faster_ than the debug one 09:36:09 Does it also run faster :P 09:36:15 in a sec, I'm rebuilding win again, this time deleting $INSTDIR\README.txt not README.txt 09:36:39 roger 09:36:56 i've adjusted the scripts to use git clean -dfx in the meantime 09:37:20 and removed crawl-sprint.bat from all 5 branches 09:37:38 greensnark: since the compilation didn't max the CPU but swapped like hell, optimized=smaller code could indeed take less time 09:37:49 Ah 09:37:59 especially to link, since linking C++ code takes oodles of memory 09:38:09 Also lack of debug info should make objects a lot smaller 09:38:15 is there no arm crosscompiler in debian? 09:39:17 "Failed to set video mode: couldn't find matching GLX visual" 09:39:22 crap 09:40:01 did you try fullscreen? 09:40:06 Napkin: there is, but cross builds are for wimpy iPads and that kind of stuff 09:40:20 well.. at least distcc? 09:40:24 I typed ./crawl without doing anything 09:40:47 and mount some diskspace via nfs ;) 09:40:58 but 256mb of ram is difficult to counter ;) 09:41:01 Napkin: of course, but that's not the point right now :p 09:41:46 and if you get Wifi LEAP working, with the default daemons, please beep me :D 09:42:16 using wpa_supplicant works, but isn't such a great idea... 09:49:20 my wifi AP doesn't do LEAP 09:49:57 but at least according to the Wikipedia page, LEAP is as bad as WEP1 09:50:56 one last thing: I renamed crawl.exe to crawl-console.exe, but I'm not so sure about that 09:51:58 on Unix, there's "crawl" being console and "crawl-tiles" being, well, tiles, since it makes sense there (the console version is almost always run from a terminal) 09:52:20 on Windows, at least one player was confused by that 09:56:14 03kilobyte * r554cce172dcd 10/crawl-ref/source/ (makefile util/crawl.nsi): Windows installer changes: 10:02:07 does 0.7 have changes to the random monster energy thing? 10:02:23 sounds good, KiloByte 10:02:44 the startmenu entries have the same name, right? 10:04:16 stabwound: on double attacks 10:04:18 no* 10:04:56 ah, okay 10:05:00 I guess that's better than nothing 10:06:11 03kilobyte 07stone_soup-0.7 * rfd4cbc9dae9f 10/crawl-ref/source/ (makefile util/crawl.nsi): Windows installer changes: 10:24:18 !tell purge File a mantis bug with the resolution you're at and the unnaturally large taskbar size that you're using. ;) 10:24:19 Enne: You have 1 message. Use !messages to read it. 10:24:19 Enne: OK, I'll let purge know. 10:26:19 03kilobyte 07stone_soup-0.7 * r82e4d82a261c 10/crawl-ref/source/command.cc: Fix crash on ?/m gold, where "gold" is a partial match for a mimic type. 10:26:21 03kilobyte 07stone_soup-0.7 * r4fe3e11c7cd4 10/crawl-ref/source/ (effects.cc itemname.cc itemprop.cc mon-stuff.cc mon-stuff.h): Make rotting hulk corpses pre-rotten. Drop rotting^2. [1984] 10:26:22 03greensnark 07stone_soup-0.7 * r92f521c42930 10/crawl-ref/source/ (l_crawl.cc macro.cc macro.h main.cc): Fix Aliza's macro crash (!lm Aliza type=crash 1) 10:26:58 -!- flowsnake has joined ##crawl-dev 10:33:35 -!- eith has joined ##crawl-dev 10:39:53 03kilobyte * r2853989995da 10/crawl-ref/source/command.cc: Fix crash on ?/m gold, where "gold" is a partial match for a mimic type. 10:52:34 -!- Ekaterin has quit [Quit: Leaving] 10:55:07 Enne: I am pretty sure I am using the default taskbar size. Windows 7 has large icons on the taskbar by default which I think makes it larger than xp 10:55:07 purge: You have 1 message. Use !messages to read it. 10:55:29 I'll post a tracker item 10:55:59 Right now Crawl just checks the resolution, subtracts some magic number, and then centers the screen. 10:56:59 Maybe I should find some less hacky way to do that. 10:57:03 Yeah, I think its centered, its just a windows 7 issue 10:57:16 I can set the taskbar icons to small which fixes it, but thats not the default 10:58:12 Also, i am in a strange resolution... 1680x1050 which might be causing it 10:58:20 lemme try a normal res... 10:59:58 I see in 1024x768 it fullscreens nicely :) 11:00:05 1680x1050 is a very normal res :P 11:00:44 It's just the tall taskbar. 11:00:46 syllogism: Ah. I was told it was not an ordinary resolution 11:07:47 We have a whole bunch of tty crashes that are really easy to trigger by resizing terminal 11:08:03 So all you have to do is collect scrolls of acquirement, save and acquire/crash until you have all good stuff 11:09:10 We need to start recovering from some of these error conditions instead of just ASSERT-crashing 11:10:02 alternatively, have a separate assert routine 11:10:05 that does save+quit 11:10:13 for things that don't indicate memory corruption 11:10:19 like terminal issues 11:12:50 On demand save+quit is also quite dangerous 11:13:05 You can use force_more to trigger a more on messages that indicate you're about to be hosed 11:13:15 And then do whatever forces the save+quit 11:13:35 Although that may be fixed with by's new messaging, I should check with him 11:13:47 hmmm 11:14:05 do things actually not happen until their message scrolls by? 11:14:16 that seems kinda broken 11:14:27 I think it's different now 11:14:40 But in older crawls mpr() could block until the player dismisses a force-more 11:14:51 So all you need to do is set force-more on things like banishment messaging 11:15:07 I think messages are now accumulated and flushed later, so this may no longer be an issue 11:15:20 yeah, that would make sense to me 11:22:19 I think force_more causes messages to be flushed at that point 11:22:51 This discussion is relevant to me, as well. 11:23:04 my point is just that the messages should appear after the event already occurred within the game 11:23:11 Because iOS devices barely get any time when they're told they're going to be closed. 11:23:15 Definitely no user input. 11:23:34 so putting a force_more on banishment and then forcing a save should leave you in the abyss 11:23:37 etc 11:24:32 I'm pretty sure it's not like that now 11:26:20 if there's a "more" in the middle of a turn, the map view and status area need to be up-to-date with the last displayed message, so it's not as simple as just queueing up a turn's messages 11:26:45 by: Ah, ok 11:28:25 alternately 11:28:38 should separate game state from display state, like for server tiles 11:28:45 maybe we could just disallow pausing the game like the mid-turn 11:28:59 keep force_more, but have it give a prompt once the current turn is completely over 11:29:09 whenever the player would normally have input again 11:43:59 greensnark: I'd really prefer not to remove those cgotoxy asserts. 11:44:10 Enne: Oh 11:44:12 Tiles problems? 11:44:20 I put them there to catch console devs who assumed that everybody had a character display and could just write off the end of whatever region they were in. 11:44:37 03greensnark * rb8261bfa3512 10/crawl-ref/source/ (libutil.cc menu.cc output.cc view.cc viewmap.cc): Fix tty Crawl crashing or CPU-pegging when the user shrinks their terminal. 11:44:53 Sorry, I pushed that a while ago. What are the alternative fixes? 11:45:41 I'd like to sort these out before the tournament, because these crashes are very abusable 11:45:53 Granted, we'll get botspam in here if someone does try to abuse it 11:46:11 In the short term, you could push that to 0.7 and leave them in master to sort out a better approach later. 11:46:21 Ok, will do 11:46:49 Do we have a minimum terminal size? 11:46:57 We do, 61x24 11:47:06 But the game doesn't try to stop you resizing down after you start it 11:47:23 And it's not entirely clear how to prevent that easily 11:47:27 the game should at the very least, catch SIGWINCH 11:47:33 It does 11:47:49 and just refuse to continue until you resize to the original size 11:48:08 Hm, block in the signal handler? 11:48:21 We could print "Terminal too small" if there's room for it. 11:48:23 The main code could be in a force-more-like scene at that point 11:50:12 03greensnark 07stone_soup-0.7 * r0ced1043e1e1 10/crawl-ref/source/ (libutil.cc menu.cc output.cc view.cc viewmap.cc): Fix tty Crawl crashing or CPU-pegging when the user shrinks their terminal. 11:52:45 are there other abusable situations other than force_more? 11:54:11 any lua hook could probably ask for input 11:54:34 Yes 11:55:45 hmmm, or at the very least stick a "for i = 1, 100000" or something in there 11:56:03 Crawl will throttle Lua scripts that are trying to use too much CPU 11:56:04 03greensnark * rbba8f8c213a7 10/crawl-ref/source/ (libutil.cc output.cc view.cc): Reinstate view bounds asserts; these crashes should be fixed by more robust handling of SIGWINCH in 0.8. 11:56:16 greensnark: right, but the cpu time isn't the issue, it's the real time 11:56:27 But that doesn't solve the case where the user wants to avoid the game handling HUP gracefully 11:57:01 doy: Real time? 11:57:35 greensnark: i mean, even if we disallowed prompting during lua hooks, people could still stick something in to delay them long enough to hup at the right time 11:57:35 -!- DocAquatic has joined ##crawl-dev 11:57:55 Oh ok, we're talking about the same thing then 11:57:58 yeah 11:58:50 Good afternoon, Crawl developers :) 12:00:06 i still don't think that a safe+quit handler would make things any worse, and it would get rid of a couple of abuses 12:00:56 but yeah, to really handle this properly, we really do need to completely decouple the view from the game state 12:02:09 Save+quit would fix some abuses but enable some others :/ 12:02:15 So it just moves the problem around a bit 12:02:18 what others would it enable? 12:02:25 The force-more scene :) 12:02:37 how is that not possible with the current state? 12:02:44 you just go back a little further 12:02:51 Oh, right 12:02:59 Yeah, hm, that makes sense 12:04:58 Let's do that now and see how it works out 12:05:19 just enable it for the cgotoxy asserts for now, i guess 12:05:49 figuring out what's a safe place to use that at and what isn't is probably a bit tricky 12:09:25 as for the winch stuff, could we just have the signal handler clear any waiting prompts, replace the screen with a "resize your terminal back to this big" message (optionally with a key combination to force it back to the right size), and disable all input until that's done (maybe except for saving)? 12:09:30 would that make sense? 12:11:09 Yes, but it'll need some infrastructure 12:11:33 The signal handler will need to set a flag so that prompts unwind correctly, like the HUP-seen flag 12:11:42 yeah 12:11:45 We should definitely do it though 12:11:56 and i guess the level up prompt will need some special handling too 12:11:58 Otherwise we'll just keep seeing similar crashes sneaking in in future :) 12:12:01 yeah 12:12:05 well 12:12:06 Level-up has special handling already 12:12:19 ideally, it would just work, regardless of how you resized the terminal 12:12:23 (above some minimum size) 12:12:27 but... 12:14:46 Oh, we handle sane cases already 12:15:06 obviously we don't, if we're getting crashes 12:15:07 (: 12:15:18 Sane cases as in term sizes that are big enough 12:15:30 oh, the crashes are just for terms that are too small? 12:15:34 Yes 12:15:38 ah, okay, yeah 12:15:50 that should definitely just not be allowed 12:15:53 Also some edge cases that were mishandled like really long level-map titles 12:15:58 you'll get asserts from gotoxy if smaller than 80x24 12:16:07 i think 80x24 is a reasonable minimum 12:16:17 Some weird people try to play on phones :P 12:16:36 heh 12:16:37 well 12:16:47 The game rejects <61x24 on startup. If that's not enough we should change that check to whatever is enough 12:17:05 !lm * type=crash sprint 12:17:05 8. [2010-07-15] MarvinPA the Brawler (L11 TrBe) ASSERT(x >= 1 && x <= sz.x) in 'libutil.cc' at line 746 failed on turn 2620. (Sprint) 12:17:16 MarvinPA <3 12:17:25 He's a bug magnet :) 12:17:36 We need more MarvinPAs. Can we clone him or something 12:29:30 03enne.walker * rfdbd60ccb78a 10/crawl-ref/source/startup.cc: Revert "Reserve extra space for tiles save game menu. Fixes mantis #1840" 12:29:33 03enne.walker * r61540fe7fc43 10/crawl-ref/source/startup.cc: Always use max lines in tiles startup menu. 12:29:34 03enne.walker * rdd0fe08b9ad4 10/crawl-ref/source/startup.cc: Remove magic numbers from startup menu. 12:29:42 03enne.walker * r3b999c2eeeaf 10/crawl-ref/settings/tiles_options.txt: Remove better transparency option from settings. 12:32:13 -!- eith has quit [Ping timeout: 240 seconds] 12:40:39 03greensnark * rd2529118e670 10/crawl-ref/source/ (9 files): Add ASSERT_SAVE macro for asserts where it should be safe to emergency-save the game when the assert fails (doy). 12:42:12 Enne: Do you see messages from previous games in tiles if you quit and start a new game without restarting Crawl? 12:42:34 "to save myself a recompile" :) 12:42:36 I did 12:42:42 Ah, damn 12:42:42 start game, quit, start new game 12:42:45 * greensnark recompiles. 12:42:48 old messages are in the message area 12:42:54 but not in the ^P screen 12:43:00 Even after the clrscr() I added? 12:43:09 that was yesterday btw 12:43:11 haven't tride today 12:47:32 I can't repro that, but there's also all the debug messages. 12:47:45 is the quad damage in 0.7? 12:47:55 (Although, apparently I'm failing to turn wiz mode off with wiz_mode = no?) 12:47:58 it would be more likely to show up without clear_messages=true, maybe? 12:48:28 I don't see it either 12:48:36 I guess Enne fixed it with the clrscr 12:48:50 I tried with and without clear_messages, looks ok 12:50:33 03enne.walker 07stone_soup-0.7 * rd7ed3d8424b1 10/crawl-ref/source/startup.cc: Remove magic numbers from startup menu. 12:50:40 03enne.walker 07stone_soup-0.7 * rd2e5582ddb70 10/crawl-ref/source/startup.cc: Always use max lines in tiles startup menu. 12:50:41 03enne.walker 07stone_soup-0.7 * r913a2afca6f0 10/crawl-ref/settings/tiles_options.txt: Remove better transparency option from settings. 12:52:17 -!- bhaak is now known as notbhaak 12:52:31 -!- notbhaak is now known as bhaak 12:53:45 So I can create a character called '', without the quotes 12:53:59 The game seems quite happy with it too 12:57:03 -!- ogaz has joined ##crawl-dev 12:57:08 Casting summon elemental on metal walls should create an iron elemental (https://crawl.develz.org/mantis/view.php?id=1988) by coolio 12:57:16 Hah 12:57:31 03greensnark 07stone_soup-0.7 * rf9466b9c4ced 10/crawl-ref/source/ (dgn-overview.cc dgn-overview.h main.cc): [1980] Clear overview in _reset_game. 12:57:35 * greensnark blames due for encouraging the elemental fetishists. 12:57:40 03felirx 07stone_soup-0.7 * r21093ec76e9d 10/crawl-ref/source/newgame.cc: In species/bg menu, '+' will now either pick viable or duplicate '*' behaviour if no previous choice has been done. 12:57:55 Someone should point out that metal != iron in many cases 12:58:40 A rusty copper elemental is created! It collapses under its own weight. 12:58:41 Enne: Any more immediate changes to tiles or shall I run the builds? 12:59:34 i originally described the iron as demonic but the word was deleted when they got turned into real monsters :3 12:59:44 demonic iron? 12:59:53 they're specifically from dis you see 13:00:01 ichor running through its conduits? :) 13:00:04 it's like regular iron, only evil 13:00:08 Hehe 13:00:09 :D 13:00:32 greensnark do you know if the quad damage is in 0.7 i was gonna stick one in this map for fun 13:00:33 a hostile iron elemental then 13:00:39 Mu_: It is 13:00:43 cool 13:00:56 But not usable by monsters :'( 13:00:59 :( :( :( 13:01:13 Mu_: just don't put it anywhere where it can actually be useful :P 13:01:14 Cerebov with quad damage would be fun 13:01:35 i'm gonna put it in that sword in the stone vault 13:01:39 Heh 13:01:50 I don't think summon elemental should be able to dig through metal walls 13:01:53 Does it have permarock glass? :P 13:02:07 ogaz: No danger of that happening :) 13:02:43 Ok, I shall update the builds 13:10:28 it would be pretty easy to get it out of that vault since teletraps don't do anything 13:11:42 oh :( 13:12:26 though anyone with LRD probably also has apportation anyway 13:12:41 ye 13:13:08 just make it airdrop 10 cerebovs when you pick it up 13:13:14 Use a trigger to gate in six pan lords ... damn 13:13:24 syllogism sometimes thinks spookily like me :P 13:14:11 not sure how i'd do that 13:14:22 Mu_: item pickup trigger? 13:14:43 ye 13:15:07 Or you can trigger it when they just break the glass :P 13:15:18 So that they have added incentive to pick up the quad real fast :P 13:15:24 I hate you people :( 13:16:14 Hey, we're good salespeople. If they wanted the quad that badly before the injection of pan lords, they'll want it even more after 13:19:47 Napkin: git clean -dfx should happen after the zip is moved to the trunk builds dir 13:20:00 I've made the change to update.sh 13:20:22 arg, true 13:20:31 Should I commit it? 13:20:39 i need to adjust that on every branch :-O 13:20:45 yes, --amend it, please 13:20:48 Ai caramba :) 13:21:50 Ok, amended 13:21:59 thanks :) 13:29:20 Oh, you have non-tiles on a different branch :) 13:29:33 -dfx strikes again :P 13:30:14 I had no other choice ;) 13:30:58 Ok, committed to both places 13:31:20 What command do I use to build the magic windows installer 13:31:33 not finished yet 13:31:40 it's in branch installer-0.7 13:31:43 Ok 13:32:13 There's a hilarious review of the second Twilight movie: http://skepchick.org/blog/2010/07/reclipse-yes-the-shit-still-sucks/ 13:35:28 Ok, rc2 incoming tomorrow, Jul 17 13:35:42 Any critical fixes would do well to make it in before that :) 13:35:45 ok, i'll try to have the installer finished tonight 13:35:55 Thanks! 13:38:37 -!- Pseudonut has joined ##crawl-dev 13:40:07 The worst movie I can recall seeing in recent times is Van Helsing(with hugh jackman) 13:41:30 Heh, Van Helsing was pretty dreadful, yeah 13:42:13 i think i'd probably have to say spiderman 3 13:42:25 (although i didn't see van helsing) 13:42:32 (or twilight) 13:44:13 Judging by the reviews, I should stay well clear of Twilight 13:44:30 I was once roped into watching a Vin Diesel movie 13:44:51 I managed to escape shortly before they drove ten cars into narrow tunnels at 200mph though 13:46:45 That new Avatar: The Last Airbender has moved up to 8% on rotten tomatoes :D 13:46:54 (was 7%) 13:48:13 greensnark: Vin Diesel has one excellent movie in my book(entertainment wise) -- Cronicles of Riddick 13:48:56 riddick was pretty meh, i thought 13:49:02 not terrible, but 13:49:07 pitch black was pretty good though 13:50:59 Then I saw Babylon A.D. and I gave up on him 13:51:09 after wasting the money :| 13:57:15 03greensnark 07stone_soup-0.7 * r0c6c0ef7bc3c 10/crawl-ref/README.txt: Remove mention of Ctrl-T as the key to start the tutorial. 13:57:21 03greensnark 07stone_soup-0.7 * r72a878dddae3 10/crawl-ref/ (README.pdf source/util/docs/README.tex): Remove mention of Ctrl-T for tutorial from README.pdf 14:18:24 -!- eith has joined ##crawl-dev 14:33:04 -!- ortoslon has quit [Ping timeout: 258 seconds] 15:00:04 -!- joosa has quit [Ping timeout: 248 seconds] 15:00:42 -!- ortoslon has joined ##crawl-dev 15:33:27 -!- joosa has joined ##crawl-dev 15:49:10 -!- Textmode has quit [Ping timeout: 265 seconds] 15:50:08 greensnark: No more changes that I know about. :) 15:57:08 -!- nrook has joined ##crawl-dev 16:14:15 Ok, I lied. One more commit. ;) 16:14:24 03enne.walker * r3699c76821c6 10/crawl-ref/CREDITS.txt: Add Ilkka Koskela (coolio) to the credits. 16:14:36 -!- Textmode has joined ##crawl-dev 16:14:57 oh, I almost hoped he was the rapper 16:15:35 03enne.walker 07stone_soup-0.7 * r8c4d6e6caffe 10/crawl-ref/CREDITS.txt: Add Ilkka Koskela (coolio) to the credits. 16:23:58 -!- Eronarn has quit [Ping timeout: 260 seconds] 16:29:13 -!- purge has quit [Ping timeout: 240 seconds] 16:32:49 -!- Danei has quit [Read error: Connection reset by peer] 16:38:57 -!- Eronarn has joined ##crawl-dev 16:43:28 KiloByte: /var/www/crawl.develz.org/htdocs/trunk/crawl-0.7.0-rc1-27-g8c4d6e6.setup.exe 16:43:40 erm... 16:43:43 KiloByte: http://crawl.develz.org/trunk/crawl-0.7.0-rc1-27-g8c4d6e6.setup.exe 16:44:47 greensnark: update-mingw-installer-0.7.sh 16:44:58 -!- Eronarn has quit [Ping timeout: 260 seconds] 16:45:57 due: what do you think how to best add the new installer to our trunk page? 16:46:42 -!- Eronarn has joined ##crawl-dev 16:50:09 -!- bmh has joined ##crawl-dev 16:56:49 KiloByte: any way to enable more compression in the installer? compression of setup.exe is almost = 0 16:57:29 KiloByte: both zips together -> 13mb... setup.exe with same content -> 26mb 17:03:56 Napkin: uhm, it's <8MB for me 17:04:58 erm - true 17:05:02 the .nsi includes: SetCompressor /SOLID lzma 17:05:10 now i wonder, what my "beta tester" was talking about :D 17:07:01 lol, he blames it on windows ;> 17:10:17 user error :P 17:15:29 -!- bmh has left ##crawl-dev 17:15:42 -!- Cryp71c has quit [Quit: Leaving] 17:16:22 hmm, somehow my build is nearly a megabyte smaller 17:18:26 Fedhas should protects your food from harpies (https://crawl.develz.org/mantis/view.php?id=1989) by galehar 17:23:04 hmm - how do I compile without using system-wide sdl? 17:23:33 Napkin: Mangle the makefiles to actually use the include files from contrib? 17:23:38 Napkin: They've been broken for ages. 17:23:55 well, i think there is a variable to set, but I forgot 17:26:03 KiloByte: what do you think about debian packages (stable) of trunk/prerelease? 17:27:27 Napkin: definitely a good idea 17:27:47 i think 0.6 is in sid already - so we should have a ready debian/ folder already 17:28:39 Zao: specify BUILD_FOO=y 17:31:30 Napkin: I see you install the whole 'git log', it has 4.5MB uncompressed 17:31:42 -!- ortoslon has quit [Ping timeout: 258 seconds] 17:31:49 true 17:31:55 KiloByte: Ooh, handy. 17:32:30 KiloByte: oh, stop looking for why it's different in size - my fault 17:32:39 Zao: I fail to see why you would use contrib libraries on any Unix system 17:32:41 i was just shocked to think it's not compressed at all 17:32:45 KiloByte: Sounds a bit odd though, as it builds contrib already, but things like tool/*.cc doesn't have the include paths. 17:32:56 KiloByte: Non-admin. No internet connection. 17:33:06 Zao: and in all distributions I know, the Security Police will shout at you 17:33:13 ah 17:34:35 hmm.. the (Read options from unknown.) "bug" is quite inconvenient 17:34:55 yeah, and it's been there since forever 17:35:30 omg 17:35:41 make -j9 -C crawl-ref/source WIZARD=y TILES=y CHOWN=true CHMOD=true DESTDIR=$HOME/Games/Crawl install 17:35:56 ended in morgue & saves being saved in $HOME 17:36:09 next try... 17:37:54 *giggle* I have also no idea where my macros just got saved :D 17:38:24 not in my $HOME, not in $HOME/Games/Crawl/settings 17:39:06 ahh.. "Couldn't open settings/macro.txt for writing!" 17:40:49 ok, let's read makefile again for instructions - DESTDIR seems not to be designed like this 17:42:03 let's try prefix 17:42:58 let me guess - I have to set SAVEDIR and DATADIR manually, since I am not in windows? 17:44:51 yes, neither DESTDIR nor prefix only works. morgue & saves in $HOME and macros can't be saved. 17:45:07 May I mention, that this is messy? 17:46:45 especially, because I have to specify DATADIR=data/ SAVEDIR=saves/ now, which seems to be exactly what I am trying to prevent.. 17:49:20 :/ 17:49:49 hehe, still "Couldn't open settings/macro.txt for writing!" 17:50:50 just for the record - this has nothing to do with unix write permissions 17:52:34 well, ok, at least the settings are read from $HOME/Games/Crawl/data/settings 17:53:03 --- 17:53:06 cd $HOME 17:53:10 mkdir settings 17:53:13 crawl 17:53:32 > file settings/macro.txt 17:53:32 settings/macro.txt: ASCII text 17:53:36 --- 17:58:11 jeeeezz.. 17:58:33 > grep macro_dir Games/Crawl/data/settings/init.txt 17:58:33 macro_dir = /home/chojin/Games/Crawl/data/settings/ 17:58:35 --- 17:58:42 that is being ignored as well... 17:59:47 ln -s $HOME/Games/Crawl/data/settings/init.txt $HOME/.crawlrc <- didn't help either... 17:59:50 help? 18:02:42 ok.. giving up... 18:02:52 crawl -macro /home/chojin/Games/Crawl/data/settings/ <- didn't work either 18:04:57 -!- Cryp71c has joined ##crawl-dev 18:05:45 seems like "cd $HOME/Games/Crawl/data && crawl" is the only way I can play with a local install. 18:07:07 sorry, cd $HOME/Games/Crawl/data && ../bin/crawl 18:17:38 not acceptable 18:21:52 Do we only use the cwd and not the executable directory when looking for data? 18:22:08 Oh, nevermind, you put the data elsewhere. 18:25:38 well, i had to, Enne 18:26:15 otherwise morgue and saves landed in $HOME instead of the folder i specified as prefix 18:26:43 and actually, everything is loaded from the data dir, apart from the macro.txt 18:26:55 which is always in $PWD/settings/macro.txt 18:32:13 thinking about it, morgue & saves probably landed in $PWD too - $HOME was just a coincidence 18:33:35 ich hab bock auf 'nen kaffee! 18:33:54 huch, wrong channel :-$ 18:43:30 -!- nrook has quit [Quit: Lost terminal] 18:48:45 -!- Danei has joined ##crawl-dev 18:53:55 Artefact amulet has "guardian spirit" property twice (https://crawl.develz.org/mantis/view.php?id=1990) by Mr K 18:55:57 -!- flowsnake has quit [*.net *.split] 18:55:57 -!- syllogism has quit [*.net *.split] 18:55:58 -!- Sequell has quit [*.net *.split] 18:55:59 -!- lorimer has quit [*.net *.split] 18:58:43 -!- ogaz has quit [Read error: Connection reset by peer] 18:59:31 -!- ogaz has joined ##crawl-dev 18:59:48 -!- flowsnake has joined ##crawl-dev 18:59:48 -!- syllogism has joined ##crawl-dev 18:59:48 -!- Sequell has joined ##crawl-dev 18:59:48 -!- lorimer has joined ##crawl-dev 19:06:08 it was worth the trouble though - the new tiles look awesome! :D 19:07:47 !tell due Mennas has no description in game?? 19:07:47 Mu_: OK, I'll let due know. 19:07:48 evening 19:07:48 due: You have 1 message. Use !messages to read it. 19:07:51 o 19:07:52 :p 19:07:55 Mu_: Huh, really? 19:08:00 moin due :) 19:08:01 ye 19:08:09 Let me make a new build and check it out 19:08:16 maybe it was too long?? 19:08:25 No 19:09:23 i remember seeing a description, may have imagined it though 19:09:43 ye when he was still just a custom 19:10:17 @?? draconian knight 19:10:18 green draconian knight (10d) | Speed: 10 | HD: 16 | Health: 111-145 | AC/EV: 12/12 | Damage: 15 | Flags: evil, spellcaster, cold-blooded | Res: 06magic(85), 03poison | Chunks: 07contaminated | XP: 2485 | Sp: b.cold (3d25), b.draining (3d20), animate dead, teleport self / b.fire (3d25), b.cold (3d25), cantrip, b.lightning (3d21), b.draining (3d20), teleport self / b.fire (3d25), pain (d16), invisibility, anima.. 19:15:15 it's really too bad makhleb and trog sprint piety rates won't be fixed for the tournamnet 19:16:22 hmm... is there a but when using X in tiles? it's not always centering on the cursor 19:16:38 -!- Pseudonut has quit [Quit: Computer went to sleep] 19:18:08 hmm... scrolling the map (X) is generally broken here 19:18:17 anyone else seeing that? 19:20:10 ok, i can reproduce it.. as soon as the cursor is no longer centered (because of reaching the side of the map), it no longer scrolls the screen in the 90° direction 19:22:16 unrelated to my view_lock = false setting, btw 19:24:05 -!- ogaz has quit [Ping timeout: 264 seconds] 19:24:21 Tiles only? 19:25:53 !tell dpeg I worry BiA/greater servants are going to ruin Sprint in tournament; they costs ~4-5 piety to use and killing almost any monster increases piety by at least 3 19:25:53 syllogism: OK, I'll let dpeg know. 19:28:21 i think so, Enne - haven't tried console yet though 19:28:49 Enne: generally said: as soon as the cursor is no longer being centered, because you reached a border, the view_port is no longer moving 19:28:49 It's possible. The tiles X-mode diverges quite wildly from the console one. 19:29:12 I'm off for the night, but if nobody else fixes it before tomorrow, I'll take a look. 19:29:17 -!- syllogism has quit [] 19:29:25 g'night :) 19:29:56 i'll check console in the meantime 19:30:23 Hm, increase piety cost for Sprint, and reduce piety gain for servent kills? 19:35:15 i should prolly specify all the tiles the sprint map should use for the different subvaults 19:39:08 view_port stuck when examining level map (https://crawl.develz.org/mantis/view.php?id=1991) by Napkin 19:40:45 -!- Mu_ has quit [Quit: Defecator, may everything turn out okay so that you can leave this place.] 19:47:23 -!- pointless_ has joined ##crawl-dev 19:47:34 hey pointless_ 19:47:39 hi 19:52:54 How are spriggan riders or whatever implemented? 19:54:53 -!- eith has quit [Ping timeout: 276 seconds] 19:59:45 -!- MarvinPA has joined ##crawl-dev 20:29:04 ouch.. unseen horror in d8 is sucky... 20:29:16 stupid 20:31:37 -!- Pseudonut has joined ##crawl-dev 20:31:44 -!- Totaku has quit [Read error: Connection reset by peer] 20:32:28 -!- Totaku has joined ##crawl-dev 20:58:04 -!- Twinge has quit [] 21:09:31 when a "hints mode" game completes, it does not return to the main menu, like tutorial or normal game does 21:10:00 the tutorial btw suggests to read the backlog by clicking the message area - but then doesn't give a clue how to close it again. 21:16:50 -!- jooosa has joined ##crawl-dev 21:19:26 -!- joosa has quit [Ping timeout: 260 seconds] 21:19:26 -!- jooosa is now known as joosa 21:34:22 -!- Twinge has joined ##crawl-dev 21:34:30 -!- Twinge has quit [Client Quit] 21:36:55 -!- Twinge has joined ##crawl-dev 21:38:03 -!- pointless_ has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.6/20100625231939]] 21:41:20 -!- MarvinPA_ has joined ##crawl-dev 21:44:19 -!- MarvinPA has quit [Ping timeout: 258 seconds] 21:44:28 -!- MarvinPA_ is now known as MarvinPA 22:36:22 -!- eith has joined ##crawl-dev 22:49:49 -!- flowsnake has quit [Quit: Leaving] 23:14:55 -!- Cryp71c has quit [Quit: Leaving] 23:20:58 -!- Cryp71c has joined ##crawl-dev 23:24:14 -!- GrimmSweeper has joined ##crawl-dev 23:29:11 -!- ogaz has joined ##crawl-dev 23:44:36 -!- Kil2 has joined ##crawl-dev 23:46:53 -!- Cryp71c has quit [Ping timeout: 265 seconds] 23:50:05 -!- [2]GrimmSweeper has joined ##crawl-dev 23:50:05 -!- GrimmSweeper has quit [Disconnected by services] 23:50:08 -!- [2]GrimmSweeper is now known as GrimmSweeper 23:50:38 -!- GrimmSweeper has quit [Client Quit] 23:51:06 -!- ogaz has quit [Read error: Connection reset by peer] 23:51:24 -!- GrimmSweeper has joined ##crawl-dev