--- Log opened Mon Oct 31 00:00:21 2016
20161031 00:04:15< tad__> We still need this debugger session?
20161031 00:04:34< vultraz> dont ask me
20161031 00:06:06< celticminstrel> Load Map still crashes for me.
20161031 00:06:23< tad__> AOI S01?
20161031 00:06:29< tad__> Which did you just try?
20161031 00:06:38< celticminstrel> I didn't even get that far.
20161031 00:06:44< celticminstrel> It crashes before the dialog opens.
20161031 00:07:12< tad__> Sync and rebuild? Sounds like the problem I had and they just patched.
20161031 00:07:25< celticminstrel> No, I'm saying this because that's what I just did.
20161031 00:07:52< tad__> I had to do gdb / catch throw to see it fail.
20161031 00:08:15< celticminstrel> This is a segfault, if I recall correctly.
20161031 00:09:19< tad_carlucci> I guess I can leave the Windows box idling in VS in case someone needs the debug session.
20161031 00:09:30< gfgtdf> tad__: i don't think we need that debug session anymore.
20161031 00:09:40< celticminstrel> Yes, segfault.
20161031 00:09:41< tad_carlucci> OK good.
20161031 00:09:52-!- tad__ [~lundberg@173.217.65.103] has quit [Quit: Switching to Unix to get some real work done.]
20161031 00:09:56< celticminstrel> I don't remember if I pasted a stack trace before.
20161031 00:10:46< celticminstrel> It's something to do with the sorting.
20161031 00:10:58< celticminstrel> Or at least, that's where the crash occurs.
20161031 00:12:01< vultraz> ive had no problems
20161031 00:13:09< tad_carlucci> Works on Linux (now) and Windows 10. Is it possible it's a system thing for OSX like I just hit on Linux? Can you tell what it's sorting? Any files or directories look suspect?
20161031 00:14:17-!- gfgtdf [~chatzilla@x4e32b47a.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 49.0.2/20161019084923]]
20161031 00:14:42-!- esr [~esr@wesnoth/developer/esr] has joined #wesnoth-dev
20161031 00:20:15< tad_carlucci> celticminstrel, Is it crashing at the std::sort in filesystem_boost.cpp get_file_in_dir() lines 395 or 398?
20161031 00:23:59< tad_carlucci> If so, try changing 	if (files != nullptr)    to read   	if ((files != nullptr) && (files->size() > 0))  and similar for dirs jsut below.
20161031 00:24:50< celticminstrel> It's the std::sort on line 96...
20161031 00:24:57< celticminstrel> In file_dialog.cpp...
20161031 00:25:42< celticminstrel> Called from line 499.
20161031 00:26:30< celticminstrel> Hmm, this is immediately after a call to get_files_in_dir though...
20161031 00:26:45< celticminstrel> So I'll try that.
20161031 00:27:15< celticminstrel> (Why size() > 0? Shouldn't 0-byte files be included?)
20161031 00:28:09< celticminstrel> Oh, it's a vector.
20161031 00:28:23< celticminstrel> But you know... sorting a 0-length vector should be a no-op, right?
20161031 00:28:44< tad_carlucci> or ! empty() .. just something I read on SO probably not related but why risk a runtime error if we can short-circuit it?
20161031 00:29:14< celticminstrel> Well, the implementation of sort() should short-circuit if begin == end though.
20161031 00:29:37< celticminstrel> Well anyway, I'm trying it just in case.
20161031 00:29:50< tad_carlucci> celticminstrel, should and does are two different things.
20161031 00:29:59< celticminstrel> I suppose...
20161031 00:30:17< celticminstrel> Might have something to do with my old compiler, then.
20161031 00:30:39< tad_carlucci> I was reading SO on std::sort problems and .. there you go .. older compilers and libs ... 
20161031 00:31:03< celticminstrel> clang ~3.2 if I recall correctly.
20161031 00:31:20< tad_carlucci> Not that I found one mentioning xcode or osx or clang but there was some comments about bugs long ago.
20161031 00:31:22< celticminstrel> Maybe I should just start building with scons on Mac, instead...
20161031 00:31:35< celticminstrel> Then I should be able to get it to use clang 3.9.
20161031 00:32:15< celticminstrel> Last time I tried that it didn't work, but... I think that was only because I was trying to enable the address sanitizer...
20161031 00:34:43< tad_carlucci> celticminstrel, The lines you talked about are calls to 'isort_dir_entries' which is a std::sort with a lambda. Might try rewriting the lambda as a real function there. Be sure it implements < compare exactly.
20161031 00:35:47< celticminstrel> Apparently I only have 3.8, not 3.9
20161031 00:36:05< celticminstrel> Not sure why rewriting the lambda would make a difference...
20161031 00:36:09< tad_carlucci> That I **DID** find spoken of on Google as a rare hard to find segfault because it's the compare function not a memory issue at afault
20161031 00:36:58< celticminstrel> I don't get it...
20161031 00:37:00< tad_carlucci> The blog I read spoke of the compare function not exactly implementing < and accidentially doing <= and it took him days to find it.
20161031 00:37:10< celticminstrel> Hmm.
20161031 00:37:23< celticminstrel> Well, I think it would crash on Windows if that were the case though...
20161031 00:37:47< celticminstrel> I distinctly recall getting an "inconsistent comparison operator" assertion on Windows.
20161031 00:37:52< celticminstrel> (Not in Wesnoth, something else.)
20161031 00:38:21< tad_carlucci> Its a call to translation::icompare so THAT is going to be a bit platform dependant.
20161031 00:38:39< celticminstrel> Well, locale-dependent.
20161031 00:38:57< tad_carlucci> ICU for me. Whatever Windows does there.  Dunno about your system.
20161031 00:39:01< celticminstrel> Actually the crash was in icompare, in the locale's collator or something.
20161031 00:39:32 * celticminstrel assumes either ICU or some native Apple thing.
20161031 00:39:51< celticminstrel> ie Cocoa something-or-other, or CoreFoundation or something.
20161031 00:40:17< tad_carlucci> Anyway. A faulty compare function in std::sort can lead to segfaults which look like memory/pointer issues, from what I just read.
20161031 00:40:45< celticminstrel> The compare looks okay to me though.
20161031 00:41:21 * shadowm wonders when do segmentation faults not look like memory or pointer issues...
20161031 00:42:12< tad_carlucci> Well, this guy seemed a bit miffed that std::sort would segfault and waste a few days of his time on a bad compare function.
20161031 00:43:52< tad_carlucci> I'm thinking trying a normal string compare to see if that fixes it, if so then we can look at why celticminstrel has a bad icompare and if it's just outdated or a new upstream issue
20161031 00:44:18< tad_carlucci> Unless celticminstrel uses locale filenames.
20161031 00:44:21< celticminstrel> It'd be great if mattsc or ancestral can chime in. :/
20161031 00:44:26< celticminstrel> Locale filenames?
20161031 00:45:04< tad_carlucci> Something outside the normal C/ascii charset
20161031 00:45:19< shadowm> C is a locale too.
20161031 00:45:22 * celticminstrel doesn't use any unusual locale, but probably does have filenames somewhere with Unicode characters.
20161031 00:45:24< tad_carlucci> grr
20161031 00:45:30< shadowm> A crummy one but a locale nevertheless.
20161031 00:46:08< tad_carlucci> Hey, it was perfectly fine when we used 7-bit paper tape to send files .. so be nice to it. it's OLD
20161031 00:46:13< celticminstrel> The top of the stack appears to be in "vtable for fabwueibfawieuftype_info"
20161031 00:46:23< celticminstrel> Not very useful to know, but.
20161031 00:46:49< celticminstrel> So I'm testing if changing icompare to compare works.
20161031 00:59:33< tad_carlucci> There is a report of gcc 4.8.3 segfault on sortin empty vector. no info just the report of failuire
20161031 01:07:13< tad_carlucci> [](const std::string& a, const std::string& b) { return translation::icompare(a, b) < 0; });
20161031 01:07:13< tad_carlucci> should not make any difference but probably should just read
20161031 01:07:13< tad_carlucci> [](const std::string& a, const std::string& b) { return translation::icompare(a, b); });
20161031 01:07:51< celticminstrel> It works with compare instead of icompare.
20161031 01:08:05< celticminstrel> The second version you just posted is wrong.
20161031 01:08:14< celticminstrel> icompare() follows the interface of strcmp.
20161031 01:08:42< tad_carlucci> no. TRUEi if less, else FALSE
20161031 01:08:54< tad_carlucci> So its right as is
20161031 01:09:07< celticminstrel> icompare follows the interface of strcmp.
20161031 01:09:20< celticminstrel> Positive if a>b, negative if a<b, 0 if a==b.
20161031 01:09:31< celticminstrel> So the existing code is correct.
20161031 01:10:13< tad_carlucci> So compare and icompare are side-by-side in gettext_boost.cpp .. and leaving my ken .. internationalization routines I'm not versed in
20161031 01:17:49< tad_carlucci> Reading the Boost docs I don't see anything wrong with icompare() in gettext_boost.cpp so why does it cause problems?
20161031 01:20:28< tad_carlucci> Wait a minute. Boost docs say compare() should be using level_type and needs 5 parameters but we're not. But our icompare() matches the docs.
20161031 01:21:04< tad_carlucci> celticminstrel, what version of Boost have you installed?
20161031 01:25:26< celticminstrel> I think 1.58.
20161031 01:25:59< tad_carlucci> Oh .. I see. .. compare uses std::compare but icompare uses boost.
20161031 01:26:09< celticminstrel> Yeah, I was about to mention that too.
20161031 01:27:06< tad_carlucci> So either its been fixed by 1.62 or we need to keep looking
20161031 01:29:12< celticminstrel> I can't build with scons unless I can figure out how to make it find SDL. Maybe it's not equipped to use a framework for SDL...
20161031 01:29:48< celticminstrel> I wonder if the binary in the framework is suitable to be copied over as something like libsdl2.dylib.
20161031 01:31:03 * tad_carlucci shrugs. 
20161031 01:31:14< tad_carlucci> I'd be surprised
20161031 01:31:51< tad_carlucci> No xcode dylib for boost 1.62?
20161031 01:32:26< celticminstrel> I think I'd have to build Boost myself, but not sure. At least, it's what I've done in the past.
20161031 01:32:50< celticminstrel> (Because default builds used gcc at that time.)
20161031 01:32:58< celticminstrel> (And libstdc++.)
20161031 01:33:16< celticminstrel> (Which is a problem because that libstdc++ doesn't support c++11 at all.)
20161031 01:33:17< tad_carlucci> So an hour or so jsut to test. And all we know is it's OK using std::compare but fails using boost:: ... compare()
20161031 01:33:43< celticminstrel> I think it's more than an hour to build Boost, but I can't remember. It's been awhile.
20161031 01:34:10< tad_carlucci> I got lucky and found a prebuilt for Arch and Windows 10.
20161031 01:36:04< tad_carlucci> Well, I'm out of my depth for anything more than gazing at the manual and wondering if I'm missing something. But it seems to me we know the area to examine. Might be better to rattle more knowedgable heads about it when they're around.
20161031 02:01:22< tad_carlucci> As a quick test I created a file with a UTF8 non-ascii character and checked. No problem loading that map. Not sure if that means anything though.
20161031 02:47:56-!- tad_carlucci [~lundberg@173.217.65.103] has quit [Quit: Off to resolve a merge conflict between the wife and husband branches of my real life.]
20161031 02:58:10-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161031 03:20:29-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20161031 03:20:35-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20161031 03:29:36-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has quit [Ping timeout: 260 seconds]
20161031 04:35:41-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has joined #wesnoth-dev
20161031 04:43:35-!- irker858 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20161031 04:53:11-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161031 05:42:50-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20161031 05:43:13-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161031 06:00:33-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection timed out]
20161031 06:02:55-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161031 06:06:22-!- JyrkiVesterinen [~jyrki@87-100-197-241.bb.dnainternet.fi] has joined #wesnoth-dev
20161031 06:19:55-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection timed out]
20161031 06:21:25-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161031 06:29:34-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 244 seconds]
20161031 06:30:53-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161031 07:09:42-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20161031 07:39:56-!- JyrkiVesterinen [~jyrki@87-100-197-241.bb.dnainternet.fi] has quit [Quit: .]
20161031 07:59:20-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20161031 08:02:44-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20161031 08:03:19-!- mjs-de [~mjs-de@p508C90E1.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161031 08:13:44-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection timed out]
20161031 08:14:45-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161031 08:18:50-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20161031 08:18:50-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20161031 08:19:11-!- VultCave is now known as vultraz
20161031 08:21:36-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20161031 08:30:39-!- JyrkiVesterinen [~JyrkiVest@194.157.54.14] has joined #wesnoth-dev
20161031 08:32:05-!- Jetrel [~Jetrel@2001:558:6014:1e:2422:435:dd84:bbf3] has quit [Quit: "The highest possible stage in moral culture is when we recognize that we ought to control our thoughts." - Charles Darwin]
20161031 08:38:42-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20161031 08:44:03-!- mjs-de [~mjs-de@p508C90E1.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20161031 09:29:22-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has quit [Ping timeout: 256 seconds]
20161031 09:50:02-!- vultraz [~chatzilla@124.109.10.167] has quit [Ping timeout: 244 seconds]
20161031 09:53:00-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161031 10:06:33-!- louis94 [~~louis94@91.178.241.125] has joined #wesnoth-dev
20161031 10:11:21-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has joined #wesnoth-dev
20161031 10:27:14-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20161031 10:35:46-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 268 seconds]
20161031 10:46:48-!- louis94 [~~louis94@91.178.241.125] has quit [Quit: Konversation terminated!]
20161031 10:50:07-!- mjs-de [~mjs-de@p508C876E.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161031 10:53:49-!- gfgtdf [~chatzilla@x4e36aa37.dyn.telefonica.de] has joined #wesnoth-dev
20161031 10:58:36< gfgtdf> celmin: does it also crash for you hen you replace icompare ith compare?
20161031 10:59:38< gfgtdf> 20161031 00:38:57< tad_carlucci> ICU for me. Whatever Windows does there.  Dunno about your system.
20161031 11:00:15< gfgtdf> tad_carlucci: if you have boost complied with icu support it used icu there on windows too.
20161031 11:03:18< loonycyborg> pretty sure that icu support in boost does nothing unless you're using particular function from boost.regex
20161031 11:03:51< loonycyborg> it has icu-enabled regexes, and they don't work without icu
20161031 11:04:08< loonycyborg> since I can still build wesnoth without icu we're not using them in wesnoth
20161031 11:18:44-!- mjs-de [~mjs-de@p508C876E.dip0.t-ipconnect.de] has quit [Ping timeout: 256 seconds]
20161031 11:26:37-!- mjs-de [~mjs-de@p508C961C.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161031 11:27:08< gfgtdf> loonycyborg: this is not about boost::regex this is about boost::locale for comparing strings or time formatting.
20161031 11:27:24< loonycyborg> ah
20161031 11:27:47< loonycyborg> and we're using some functions that could go through icu backend in boost.locale?
20161031 11:28:01< gfgtdf> loonycyborg: yes
20161031 11:29:10-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161031 12:58:27-!- mjs-de [~mjs-de@p508C961C.dip0.t-ipconnect.de] has quit [Ping timeout: 250 seconds]
20161031 13:06:44-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20161031 13:09:11-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161031 13:11:43-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 245 seconds]
20161031 13:11:49-!- mjs-de [~mjs-de@p508C86FC.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161031 13:15:10-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161031 13:15:49-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Read error: Connection reset by peer]
20161031 13:18:30-!- stikonas_ is now known as stikonas
20161031 13:22:40-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20161031 13:24:42-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20161031 13:38:19-!- mjs-de [~mjs-de@p508C86FC.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20161031 13:52:24-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20161031 13:52:44-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161031 14:00:52-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20161031 14:01:10-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161031 14:02:39-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Client Quit]
20161031 14:11:05-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161031 15:05:58-!- tad_carlucci [~lundberg@173.217.65.103] has joined #wesnoth-dev
20161031 15:08:47< tad_carlucci> gfgtdf, I checked and I was wrong about having ICU installed on the Linux box. I know it's not on the Windows box.
20161031 15:25:10-!- mjs-de [~mjs-de@x4db514a0.dyn.telefonica.de] has joined #wesnoth-dev
20161031 15:31:07-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20161031 15:49:59-!- JyrkiVesterinen [~JyrkiVest@194.157.54.14] has quit [Quit: .]
20161031 16:11:15-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161031 16:11:40-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161031 16:11:47-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161031 16:43:41-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20161031 17:13:16-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has quit [Ping timeout: 265 seconds]
20161031 17:22:16< tad_carlucci> celticminstrel, Can you check your system and your wesnoth build to determine if you have ICU or GNU gettext installed, and which you're building with?
20161031 17:26:53< tad_carlucci> $ yaourt -Q boost icu gettext
20161031 17:26:54< tad_carlucci> extra/boost 1.62.0-2
20161031 17:26:54< tad_carlucci> extra/icu 57.1-1
20161031 17:26:54< tad_carlucci> core/gettext 0.19.8.1-2 (base base-devel)
20161031 18:01:29-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161031 18:02:14< celticminstrel> Uh.
20161031 18:02:21< celticminstrel> Hmm.
20161031 18:02:29-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161031 18:02:41< celticminstrel> Maybe I can get that info from otool?
20161031 18:03:50< celticminstrel> It only lists libiconv...
20161031 18:05:44< celticminstrel> I have both ICU and gettext installed, for the record.
20161031 18:06:46< celticminstrel> Neither are listed in the local libs directory though.
20161031 18:10:33-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20161031 18:10:50-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161031 18:17:14-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has joined #wesnoth-dev
20161031 18:34:38-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has quit [Ping timeout: 245 seconds]
20161031 18:42:02< tad_carlucci> Look at the 'extras' package for windows, I don't see either icu or gettext so I'm presuming he build boost to use Window's native internationalization. So, next thing I'd do is find someone else using OSX and compare results with them.
20161031 18:44:25-!- iceiceice [~chris@nat-tvwna-outside-visitornet2-b-119.princeton.org] has joined #wesnoth-dev
20161031 18:44:25-!- iceiceice [~chris@nat-tvwna-outside-visitornet2-b-119.princeton.org] has quit [Changing host]
20161031 18:44:26-!- iceiceice [~chris@unaffiliated/iceiceice] has joined #wesnoth-dev
20161031 18:47:32< iceiceice> celticminstrel, i was going through my email, i got some comment on this commit from 12 days ago that i didn't see until now
20161031 18:47:33< iceiceice> https://github.com/wesnoth/wesnoth/commit/e9d72ad7ed7d6897a3807b1293ff4c37c6c5b7d9#commitcomment-19495480
20161031 18:47:44< iceiceice> i can't see it now on github, maybe it got deleted?
20161031 18:48:12< iceiceice> the reason i removed those two lines is that i moved the declaration to the header
20161031 18:48:28< iceiceice> i guess it works the other way too and sometimes in other projects you'll see like, extern definitions in .cpp files
20161031 18:48:42< iceiceice> but in wesnoth at least we usually but the declaration in the header
20161031 18:49:02< iceiceice> its not a big deal either way
20161031 18:49:24< iceiceice> have to go, i just didn't want to forget about this after i clean out my emails
20161031 18:49:38-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161031 18:49:44-!- iceiceice [~chris@unaffiliated/iceiceice] has quit [Read error: Connection reset by peer]
20161031 19:00:35< tad_carlucci> The commit iceiceice refers to moves the externs to the header, where they presently are (just checked), and moves some code to prevent stack corruption. I've also had some comments go missing for unknowable reasons (other commits).
20161031 19:28:15< gfgtdf> tad_carlucci: don't close bugs that effect a releases wesnoth version unless they are fixed is already in a released wesnoth version
20161031 19:28:40< tad_carlucci> gfgtdf, OK, sorry.
20161031 19:28:51< gfgtdf> tad_carlucci: in partuular https://gna.org/bugs/?25164 effected 1.13.5 so it shouldn't closed until 1.13.6 comes out.
20161031 19:29:42< tad_carlucci> Got it. I'll just leave 'em alone.
20161031 19:37:25-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has joined #wesnoth-dev
20161031 19:56:22-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20161031 19:59:41-!- Kwandulin [~Miranda@p5DDD2B8F.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20161031 20:51:01-!- mjs-de [~mjs-de@x4db514a0.dyn.telefonica.de] has quit [Remote host closed the connection]
20161031 20:51:35-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161031 20:59:26-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161031 20:59:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161031 20:59:49-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161031 21:07:24-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has quit [Ping timeout: 250 seconds]
20161031 21:14:36-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has joined #wesnoth-dev
20161031 21:16:00-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 256 seconds]
20161031 21:22:40-!- Appleman1234 [~Appleman1@KD106181164020.au-net.ne.jp] has quit [Ping timeout: 265 seconds]
20161031 21:25:09< celticminstrel> iceiceice, if you see this... I think I deleted the comment after I noticed the answer in my own.
20161031 21:30:23< zookeeper> so i'm gonna add a super simple feature which allows some terrain graphics rotations to be skipped, so rotations=tr,skip,br,bl,skip,tl would mean the 1st and 4th rotations would be skipped entirely. i suppose no one objects if the magic value is simply called "skip"?
20161031 21:30:56< celticminstrel> I have no idea what you mean, but "skip" seems acceptable as a magic value I guess.
20161031 21:32:04< DeFender1031> I have a vague idea what you mean, and "skip" seems acceptable to me as well.
20161031 21:32:37< vultraz> is it not possible to just not include the value at all?
20161031 21:33:02< celticminstrel> I'd wonder that as well.
20161031 21:33:28< celticminstrel> But I didn't mention it because I have no idea of the context.
20161031 21:33:45< celticminstrel> Also note, "tr,,br,bl,,tl" is not good.
20161031 21:33:52< celticminstrel> Just saying.
20161031 21:34:07< celticminstrel> Probably obvious but.
20161031 21:34:18< vultraz> no, I mean
20161031 21:34:25< vultraz> obviously, skip is in place of something
20161031 21:34:38< vultraz> so unless the list is by-index, in which case the value honestly doesn't matter
20161031 21:34:54< vultraz> then the absence of a certain value would work as much as a magic one..
20161031 21:34:56< vultraz> would it not?
20161031 21:35:01< vultraz> unless 'skip' does more
20161031 21:35:06< celticminstrel> If it's set-like, then sure.
20161031 21:35:12< celticminstrel> If it's list-like, a magic value is needed.
20161031 21:35:17< celticminstrel> I have no idea of the context.
20161031 21:35:23< celticminstrel> So I can't say which is better.
20161031 21:35:54< DeFender1031> celticminstrel, https://wiki.wesnoth.org/TerrainGraphicsWML "rotations" attribute
20161031 21:36:12-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161031 21:36:38< celticminstrel> Uh... I don't get it...
20161031 21:37:06< vultraz> so it's list-like
20161031 21:37:39< vultraz> all 6 rotations checked against the values, by index
20161031 21:37:52< DeFender1031> celticminstrel, I don't actually understand how it works either, I just know it's there and is somehow used to "simplify" the code for wall transitions
20161031 21:38:08< celticminstrel> Is it list-like? I can't even tell from that description.
20161031 21:38:42< celticminstrel> Oh wait. Suddenly I think I understand.
20161031 21:38:44< DeFender1031> though, zookeeper, what IS the point of "skip" there? Can't you just not use @R2 and @R4 to skip them?
20161031 21:39:13-!- irker220 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20161031 21:39:13< irker220> wesnoth: SJD wesnoth:master d5a84948dc53 / utils/umc_dev/README.txt: Fix typo in readme https://github.com/wesnoth/wesnoth/commit/d5a84948dc53ae09f6e2a31ad8c71650fb75facc
20161031 21:39:13< irker220> wesnoth: Charles Dang wesnoth:master 3d1839390225 / utils/umc_dev/README.txt: Merge pull request #852 from selony/patch-1 https://github.com/wesnoth/wesnoth/commit/3d18393902253fbcc9b2ecc2b213785847d98079
20161031 21:39:21< zookeeper> vultraz, i think i tried an empty value originally but couldn't get it to work
20161031 21:39:31< celticminstrel> Uhh. If there's six, why is it R0 .. R6...
20161031 21:39:35< celticminstrel> I'm confused again.
20161031 21:39:45< zookeeper> probably the parser didn't put it onto the list, or something
20161031 21:39:57< celticminstrel> zookeeper: I think there's an option for that.
20161031 21:40:13< celticminstrel> Wow, what a trivial PR.
20161031 21:42:00< zookeeper> DeFender1031, and no this is for the rare cases where you do something complicated and need to actually skip some rotations entirely
20161031 21:42:10< DeFender1031> celticminstrel, if PR is the only way for some people to commit, then even the trivial fixes will go through PR.
20161031 21:42:30< celticminstrel> zookeeper: So basically, using "skip" will cause the template to produce fewer than 6 rules?
20161031 21:42:35< DeFender1031> zookeeper, i'm not sure how skipping them differs from not using them
20161031 21:42:36< zookeeper> celticminstrel, yes
20161031 21:42:43< DeFender1031> OOOOOOOOOOOOOOOOOHHHHHHH
20161031 21:42:46< DeFender1031> that's how.
20161031 21:43:09< celticminstrel> And the reason why you can't just omit it, eg "tr,br,bl,tl" is...
20161031 21:43:15< celticminstrel> The position in the list is significant?
20161031 21:43:16< DeFender1031> TBH, I don't actually fully understand how this rotations thing works to begin with.
20161031 21:43:32< celticminstrel> eg, "tr,br" is different from "br,tr".
20161031 21:43:48< DeFender1031> celticminstrel, i think it rotates in order
20161031 21:43:51< zookeeper> celticminstrel, yeah, hexes == six directions == six rotations, so the position matters
20161031 21:44:04-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 244 seconds]
20161031 21:44:13< celticminstrel> From the description it seems like @Rn is adjusted based on the index in the list...
20161031 21:44:25< DeFender1031> menaing, 0 degrees->60->120->180->240->300
20161031 21:44:41< DeFender1031> celticminstrel, yeah, but it also rotates the map.
20161031 21:44:44< celticminstrel> So, something like, @Rn becomes @R((n+i)%6) and then is subsituted with the text at the resulting index.
20161031 21:45:04< celticminstrel> (If it's really like that then @R0 would be equivalent to @R6...)
20161031 21:45:13< zookeeper> yeah
20161031 21:45:22< zookeeper> i have some pretty wonky rotations for this walls<->castles thing
20161031 21:46:19< celticminstrel> It seems unlikely that someone would put "skip" in the image name for rotation differentiation, so it still seems fine as a magic word.
20161031 21:46:41< celticminstrel> Oh, I just remembered that I need to update the changelog...
20161031 21:47:38< DeFender1031> zookeeper, that makes sense, given how complex the interactions between them need to be
20161031 21:48:16< DeFender1031> in general, I find the terrain_graphics system clunky, complex, limited in what it lets you easily accomplish, and hard to work with.
20161031 21:48:39< DeFender1031> (that said, I don't actually have a better proposal for something that isn't those things)
20161031 21:51:17-!- louis94 [~~louis94@91.178.241.125] has joined #wesnoth-dev
20161031 21:52:29< zookeeper> yeah, either the complexity is in the engine or in WML, but the things we need to do just are inherently complicated and i don't see many ways to avoid that.
20161031 21:54:01< celticminstrel> Thinking about it, using coroutines to implement iterators might be the easiest way. (It also makes it more like Python iterators.)
20161031 22:00:08< celticminstrel> I'm going to cherry-pick the bugfixes from the lua_mapgen_stuff if no-one objects.
20161031 22:00:20< celticminstrel> vultraz, gfgtdf: ^
20161031 22:00:41 * celticminstrel missed the word "branch" there.
20161031 22:01:44< DeFender1031> zookeeper, i think the complexity of the syntax and the system itself it what bothers me... basically, The best systems are those that are simple to work with and you can use to do complex things. Some are simple to work with for simple things, but complex for complex things, which is generally fine though not ideal, but with terrain_graphics, I even find some of the simple things complex to work with.
20161031 22:06:17 * celticminstrel will link the three commits in a minute.
20161031 22:11:27< zookeeper> DeFender1031, well, sure, some of the basic functionality of the tag could be simpler i suppose (or have some convenient shortcuts)
20161031 22:12:16< celticminstrel> https://github.com/wesnoth/wesnoth/commit/e79c377b9596229d9cecaf26556394d93166374b
20161031 22:12:24< celticminstrel> https://github.com/wesnoth/wesnoth/commit/5dc888e67600481794be360becd965a9e21b0255
20161031 22:12:31< celticminstrel> https://github.com/wesnoth/wesnoth/commit/ff46718091094f8fa58f74a54035aba7f4a468dc
20161031 22:12:47< celticminstrel> Those are the three commits I'm going to merge if no-one objects.
20161031 22:12:48-!- louis94 [~~louis94@91.178.241.125] has quit [Quit: Konversation terminated!]
20161031 22:12:56< celticminstrel> Not sure if gfgtdf is around to object though... and vultraz disappeared...
20161031 22:13:14< celticminstrel> But someone else can look too if they want.
20161031 22:13:23< DeFender1031> zookeeper, not just the functionalities of each indivisual graphic, but the interactions between graphics as well. For example, getting overlapping stuff to layer properly is sometimes really annoying.
20161031 22:14:33-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20161031 22:17:49-!- travis-ci [~travis-ci@ec2-54-162-125-255.compute-1.amazonaws.com] has joined #wesnoth-dev
20161031 22:17:50< travis-ci> wesnoth/wesnoth#11816 (master - 3d18393 : Charles Dang): The build passed.
20161031 22:17:50< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/172127133
20161031 22:17:50-!- travis-ci [~travis-ci@ec2-54-162-125-255.compute-1.amazonaws.com] has left #wesnoth-dev []
20161031 22:24:46-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20161031 22:26:02-!- AI0867 [~ai@wesnoth/developer/ai0867] has quit [Remote host closed the connection]
20161031 22:26:02-!- ShikadiLord [~ignacio@wesnoth/developer/shadowm] has quit [Remote host closed the connection]
20161031 22:26:03-!- irker220 [~irker@uruz.ai0867.net] has quit [Read error: Connection reset by peer]
20161031 22:28:16< celticminstrel> Well, no-one objected, so I guess I'll push. Pretty sure I discussed some of those commits with gfgtdf before, anyway.
20161031 22:28:34< zookeeper> dammit, again
20161031 22:28:38< celticminstrel> ?
20161031 22:28:44< zookeeper> someone pushed right before me
20161031 22:28:47< celticminstrel> Oh. Sorry.
20161031 22:28:53< celticminstrel> git pull --rebase
20161031 22:29:06-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20161031 22:29:36< celticminstrel> ...why do I feel like I accidentally committed something I shouldn't've...
20161031 22:30:41< celticminstrel> Hmm, it's not there. Good.
20161031 22:30:49-!- AI0867 [~ai@wesnoth/developer/ai0867] has joined #wesnoth-dev
20161031 22:30:52< zookeeper> it's all kinds of stupid that tortoisegit's commit+push option isn't atomic.
20161031 22:31:05< celticminstrel> But it's not in the stash either. Oh well, not important.
20161031 22:31:18< celticminstrel> zookeeper: Pretty sure it can't be...
20161031 22:31:40< zookeeper> of course it can
20161031 22:31:53< celticminstrel> Huh? It was in the stash after all? Confusion!
20161031 22:32:38-!- irker070 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20161031 22:32:38< irker070> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 6f569cdcdd99 / data/lua/cave_map_generator.lua: Lua Cave Mapgen: Support comma-separated chamber locations https://github.com/wesnoth/wesnoth/commit/6f569cdcdd9901b3fb0b7651f75c243e8ae3307b
20161031 22:32:38< irker070> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 3c8f76fb5800 / data/ (4 files in 3 dirs): Lua mapgen: Fix random seed not being used to generate a scenario from a map gen https://github.com/wesnoth/wesnoth/commit/3c8f76fb5800519930ef1b8ed4bd83bbb19f4632
20161031 22:32:38< irker070> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff aa71ccaf2fef / data/lua/cave_map_generator.lua: Lua Cave Mapgen: Allow alternate clear and village terrains https://github.com/wesnoth/wesnoth/commit/aa71ccaf2fefaa2bb0e0d83ac0891ff0d13c59f0
20161031 22:32:39< irker070> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 3cb41268b881 / data/lua/cave_map_generator.lua: Lua Cave Mapgen: Allow a chamber's existence to depend on number of players https://github.com/wesnoth/wesnoth/commit/3cb41268b881958a52dbdc8dd792f4cf8ac52dad
20161031 22:32:40< irker070> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 7d3353b65013 / data/multiplayer/scenarios/Random_Scenario_Cave.cfg: Add a multiplayer cave random map https://github.com/wesnoth/wesnoth/commit/7d3353b65013081e5581d88c5ce66d7bddf94155
20161031 22:32:54< celticminstrel> What? Why'd it post the branch ones but not the master ones? :/
20161031 22:33:13< celticminstrel> Hmm, one of those commits has a wrong commit message.
20161031 22:34:44< celticminstrel> Sorry, repeat.
20161031 22:34:45< irker070> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 7e41841654a7 / data/ (4 files in 3 dirs): Lua mapgen: Use helper.rand for chamber locations https://github.com/wesnoth/wesnoth/commit/7e41841654a731f71c68972239d38ab9e5e40946
20161031 22:34:47< irker070> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff eb3d55f4340e / data/lua/cave_map_generator.lua: Lua Cave Mapgen: Allow alternate clear and village terrains https://github.com/wesnoth/wesnoth/commit/eb3d55f4340e8d5720217550f41587a355dacd39
20161031 22:34:49< irker070> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 8d8716f4c801 / data/lua/cave_map_generator.lua: Lua Cave Mapgen: Allow a chamber's existence to depend on number of players https://github.com/wesnoth/wesnoth/commit/8d8716f4c80125505cf1c0d167662cdc2e229431
20161031 22:34:51< irker070> wesnoth: Celtic Minstrel wesnoth:lua_mapgen_stuff 773a1de522b1 / data/multiplayer/scenarios/Random_Scenario_Cave.cfg: Add a multiplayer cave random map https://github.com/wesnoth/wesnoth/commit/773a1de522b1e0a56e4f74d4c7f490ec7cf6f9e6
20161031 22:38:23< celticminstrel> Whoa! The contribution graph changed colour!
20161031 22:38:41< celticminstrel> It's now all yellow and orange instead of shades of green.
20161031 22:39:01< celticminstrel> Hmm. It also says "Happy Halloween". Maybe it's a temporary holiday change thing.
20161031 22:41:41-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161031 22:43:45< Aginor> so behind the times :D
20161031 22:44:32-!- louis94 [~~louis94@91.178.241.125] has joined #wesnoth-dev
20161031 22:46:00< celticminstrel> Aginor: ...what?
20161031 22:52:25-!- Appleman1234 [~Appleman1@KD106181166169.au-net.ne.jp] has joined #wesnoth-dev
20161031 22:55:11< irker070> wesnoth: SJD wesnoth:master 9bbaf23aeea9 / utils/umc_dev/Makefile: Fix xtext javadoc path https://github.com/wesnoth/wesnoth/commit/9bbaf23aeea979e1e6ce7d3e3500835cea1a6473
20161031 22:55:13< irker070> wesnoth: Timotei Dolean wesnoth:master 0f6a779152cd / utils/umc_dev/Makefile: Merge pull request #853 from selony/patch-2 https://github.com/wesnoth/wesnoth/commit/0f6a779152cd65908c637d3bc0297a9f62a6e75f
20161031 22:55:17< irker070> wesnoth: SJD wesnoth:master 084853e262e3 / utils/umc_dev/org.wesnoth/src/org/wesnoth/preferences/AddonManagerPreferencePage.java: Add port for 1.13 in the UMC addons preferences https://github.com/wesnoth/wesnoth/commit/084853e262e31c2be497a69fff01c87aeca74001
20161031 22:55:19< irker070> wesnoth: Timotei Dolean wesnoth:master d5a1584176e7 / utils/umc_dev/org.wesnoth/src/org/wesnoth/preferences/AddonManagerPreferencePage.java: Merge pull request #854 from selony/patch-1 https://github.com/wesnoth/wesnoth/commit/d5a1584176e7e5f77e70b9693e10f17d97f2c054
20161031 22:55:51< celticminstrel> Someone I don't know is merging PRs...
20161031 22:56:10< celticminstrel> Though I guess it's obvious why he's the one to merge it.
20161031 23:04:23-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161031 23:09:54-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 256 seconds]
20161031 23:26:22-!- Appleman1234 [~Appleman1@KD106181166169.au-net.ne.jp] has quit [Ping timeout: 244 seconds]
20161031 23:26:46-!- Appleman1234 [~Appleman1@KD106181166169.au-net.ne.jp] has joined #wesnoth-dev
20161031 23:45:05-!- Appleman1234 [~Appleman1@KD106181166169.au-net.ne.jp] has quit [Ping timeout: 252 seconds]
20161031 23:49:24< Aginor> celticminstrel: halloween was yesterday
20161031 23:49:27< Aginor> :D
20161031 23:49:52< Aginor> although the office is full of leftover sweets, so I guess that's positive
20161031 23:50:52-!- travis-ci [~travis-ci@ec2-54-221-165-92.compute-1.amazonaws.com] has joined #wesnoth-dev
20161031 23:50:53< travis-ci> wesnoth/wesnoth#11821 (lua_mapgen_stuff - 7d3353b : Celtic Minstrel): The build has errored.
20161031 23:50:53< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/172141354
20161031 23:50:53-!- travis-ci [~travis-ci@ec2-54-221-165-92.compute-1.amazonaws.com] has left #wesnoth-dev []
20161031 23:54:34-!- mkdroid [~null@unaffiliated/matthiaskrgr] has joined #wesnoth-dev
20161031 23:55:32-!- mkdroid [~null@unaffiliated/matthiaskrgr] has quit [Remote host closed the connection]
--- Log closed Tue Nov 01 00:00:34 2016