--- Log opened Mon Jan 22 00:00:36 2018
20180122 00:22:45-!- irker250 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20180122 00:22:45< irker250> wesnoth: Charles Dang wesnoth:master 9614a4280a2a / src/ (31 files in 17 dirs): Cleaned up various util includes https://github.com/wesnoth/wesnoth/commit/9614a4280a2a13db260de19c1b705656b8976368
20180122 00:30:00-!- atarocch_ [~atarocch@93.56.164.28] has quit [Ping timeout: 246 seconds]
20180122 00:54:11< irker250> wesnoth: josteph wesnoth:master a2be9c8f8669 / / (4 files in 3 dirs): Hide 'Rails' and 'Fake Shroud' terrains in the unit help pages when impassable. https://github.com/wesnoth/wesnoth/commit/a2be9c8f86690974d98f7a34c85c54496097185c
20180122 00:57:20< vultraz> josteph's really being helpful
20180122 01:04:45< irker250> wesnoth: pentarctagon wesnoth:master 1192b0f1e5a9 / data/core/units/dunefolk/Khalid.cfg: Rename the Khalid's shield bash to pommel strike. https://github.com/wesnoth/wesnoth/commit/1192b0f1e5a962ee95f6b58bd4a58450b7d06edd
20180122 01:28:25< irker250> wesnoth/wesnoth:master pentarctagon ebc0e6f7b8 Rename the Khalid's shield bash to pomme AppVeyor: All builds passed
20180122 01:33:45< vultraz> and vgaming
20180122 01:40:25< celticminstrel> Yup.
20180122 01:40:33< celticminstrel> Are they in credits yet?
20180122 02:01:40< irker250> wesnoth/wesnoth:master pentarctagon 0093086581 Rename the Khalid's shield bash to pomme AppVeyor: All builds passed
20180122 02:02:39< vultraz> don't think so
20180122 02:05:54< celticminstrel> Maybe add them?
20180122 02:06:35-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20180122 02:22:05-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20180122 02:22:13-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20180122 02:26:47-!- ElFerna [~ElFerna@pc-128-111-160-190.cm.vtr.net] has joined #wesnoth-dev
20180122 02:27:05-!- ElFerna [~ElFerna@pc-128-111-160-190.cm.vtr.net] has left #wesnoth-dev []
20180122 03:22:49< celticminstrel> Incidentally, the forum thread I linked goes some way to showing what I would prefer GUI2 markup to look like.
20180122 03:23:38< celticminstrel> Where rows and columns don't need to be explicitly specified by a tag, but are implicitly gleaned from the attributes of the widget tags.
20180122 03:24:06< celticminstrel> Though it may be too much effort to actually change the GUI2 markup, we could at least use something similar for the new ThemeWML.
20180122 03:39:42 * celticminstrel also wonders if DeFender1031 ever plans to actually do that parenthetical_split stuff he was talking about awhile back.
20180122 03:52:30-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20180122 03:52:39-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20180122 04:01:47< vultraz> *pokes DeFender1031 *
20180122 04:04:55< celticminstrel> But I just did.
20180122 04:05:34< vultraz> twice! :P
20180122 04:05:57< celticminstrel> No, I only poked him once.
20180122 04:06:09< vultraz> No, now we've poked him twice
20180122 04:08:57< celticminstrel> Your words are true but your denial is incorrect.
20180122 04:09:19< vultraz> True
20180122 04:18:37< irker250> wesnoth/wesnoth:master josteph 694b7993e8 HttT S18: Rearrange the merman's message AppVeyor: All builds passed
20180122 04:30:49< irker250> wesnoth/wesnoth:master josteph 838e5d380c Hide 'Rails' and 'Fake Shroud' terrains  AppVeyor: All builds passed
20180122 04:34:58< irker250> wesnoth/wesnoth:master pentarctagon d7aa7a8aa3 Have cmake not copy over everything for  AppVeyor: vs2017/Release Failed
20180122 04:34:59< irker250> Details: https://ci.appveyor.com/project/wesnoth/wesnoth-605wt/build/Wesnoth-VS2017-master-817
20180122 04:46:13< vultraz> celticminstrel: where is [if] implemented
20180122 04:46:46< vultraz> ah, here
20180122 04:47:19< celticminstrel> wml-conditionals.lua
20180122 04:47:46< vultraz> it's wml-flow.lua
20180122 04:49:22< celticminstrel> Yeah that.
20180122 04:49:25< vultraz> celticminstrel: what would I need to do to add a `condition=` key that takes Lua under [if]?
20180122 04:49:36< celticminstrel> Don't.
20180122 04:49:39< vultraz> why
20180122 04:49:49< celticminstrel> Instead add wesnoth.wml_conditionals.lua
20180122 04:49:59< celticminstrel> Which would be very similar to wesnoth.wml_actions.lua
20180122 04:50:07< vultraz> so..[if][lua]code=?
20180122 04:50:16< celticminstrel> Yes.
20180122 04:50:33< celticminstrel> But note that the code would need to end with returning true or false.
20180122 04:50:48< vultraz> hmm
20180122 04:50:51< vultraz> ok, how do I add such a thing..
20180122 04:51:00< celticminstrel> Look at wesnoth.wml_actions.lua
20180122 04:51:53< vultraz> oh, I just need to register it as a tag from the lua?
20180122 04:52:10< celticminstrel> Right, but under wml_conditionals rather than wml_actions.
20180122 04:52:25< vultraz> is there a dedicated file for coniditons
20180122 04:52:27< vultraz> conditions
20180122 04:52:42< irker250> wesnoth: Wedge009 wesnoth:master ff3412b67ad4 / src/buffered_istream.hpp: Avoid type mismatch warning. https://github.com/wesnoth/wesnoth/commit/ff3412b67ad42e45623e438b67f0271fdf9fc23e
20180122 04:52:57< celticminstrel> Not ATM. AFAIK there's only one custom condition, which is defined in wml/object.lua
20180122 04:53:22< vultraz> there's also one in wml-tags.lua
20180122 04:53:56< vultraz> I'll move both into a file
20180122 04:53:59< vultraz> what should I call it?
20180122 04:54:05< celticminstrel> Don't.
20180122 04:54:11< vultraz> mein gott
20180122 04:54:17< celticminstrel> The one in object.lua relies on local variables in that file.
20180122 04:54:26< celticminstrel> You can move the one in wml-tags if you want.
20180122 04:54:29< vultraz> ok, I'll move *one* into a file
20180122 04:54:33< celticminstrel> wml-conditionals.lua seems like a good name.
20180122 04:55:29< vultraz> and you made it so these files are loaded automatically?
20180122 04:56:03< celticminstrel> Well, you'd need to add a require for wml-conditionals somewhere.
20180122 04:56:11< celticminstrel> Anything under wml is automatically loaded though.
20180122 04:56:19< vultraz> I see
20180122 04:56:59< vultraz> should they be loaded before or after?
20180122 04:57:15< vultraz> and by the core, or by wml-tags
20180122 05:00:22< vultraz> hmmm
20180122 05:00:38< vultraz> hmmmmmmmmmmmm
20180122 05:01:46< vultraz> celticminstrel: how would I ensure the lua code returned something?
20180122 05:03:42< celticminstrel> Well, wml/ and wml-flow are loaded from wml-tags.
20180122 05:03:50< celticminstrel> You shouldn't load anything from core, ever.
20180122 05:03:59< celticminstrel> wml-tags is loaded from _main.cfg or something.
20180122 05:04:05< vultraz> in the core
20180122 05:04:20< vultraz> :P
20180122 05:04:24< celticminstrel> ?
20180122 05:04:39< vultraz> it's loaded from data/core/_main.cfg not data/main.cfg
20180122 05:04:44< celticminstrel> Yeah, sure.
20180122 05:05:02< vultraz> along with two other things
20180122 05:05:13< vultraz> I should probably just make it load a main.lua file
20180122 05:05:17< celticminstrel> Anyway, as to ensuring the Lua code returns something, it's impossible; just ~~ the function's return value to coerce to boolean.
20180122 05:05:30< vultraz> ~~?
20180122 05:05:31< vultraz> what is ~~
20180122 05:05:31< celticminstrel> I don't like the name main.lua.
20180122 05:05:41< celticminstrel> It's the unary negation operator applied twice.
20180122 05:05:46< celticminstrel> IIRC
20180122 05:06:03< vultraz> so just `return ~~bytecode()`?
20180122 05:06:23< celticminstrel> Yeah, but uh, you do need to pass the cfg as parameter too.
20180122 05:06:33< celticminstrel> The [args] cfg or something, I forget.
20180122 05:07:35< vultraz> yeah
20180122 05:07:50< vultraz> hmmmmmmmmm
20180122 05:07:54< vultraz> something occurs to me
20180122 05:08:08< vultraz> there's nothing stopping this lua from having side effects
20180122 05:08:15< celticminstrel> True.
20180122 05:08:45< vultraz> and [lua] is actionwml
20180122 05:08:47< vultraz> too
20180122 05:08:51< celticminstrel> Yeah?
20180122 05:09:21< vultraz> so perhaps one could just replace the entire [if] block with a [lua] block
20180122 05:09:44< celticminstrel> Probably. Or whoever wanted a custom Lua condition could add their own tag under wml_conditional.
20180122 05:09:56< celticminstrel> ^+s
20180122 05:10:15< vultraz> still worth adding this you think?
20180122 05:10:37< celticminstrel> FTR, I considered adding this ages ago. I didn't do it due to things people said.
20180122 05:10:47< celticminstrel> I don't remember any details though.
20180122 05:10:57< celticminstrel> Personally I think it kinda makes sense to have it.
20180122 05:11:06< vultraz> ok
20180122 05:11:12< celticminstrel> But at the same time, in a way it is a bit redundant.
20180122 05:12:40< celticminstrel> Then again, you could argue the existing ActionWML [lua] is equally redundant (assuming the tag is still supported at global and scenario level).
20180122 05:13:44< vultraz> we did just add a wesnoth.variables table did we not
20180122 05:13:56< celticminstrel> wml.variables IIRC
20180122 05:13:58< vultraz> oh
20180122 05:14:47< celticminstrel> BTW, the wml module needs official documentation. Feel free to add to this draft page if you have time: https://wiki.wesnoth.org/LuaAPI
20180122 05:15:04< celticminstrel> (Side note: The wesnoth module coverage on that page deliberately omits most things.)
20180122 05:15:07< vultraz> well, then for some reason the condition is returning true even in the case of a syntax error
20180122 05:15:09< vultraz> :(
20180122 05:15:19< celticminstrel> Show code or something?
20180122 05:16:32< vultraz> https://pastebin.com/t6QC5btm
20180122 05:17:23< celticminstrel> That's not a syntax error.
20180122 05:17:51< vultraz> I mean, it says "variables" is a nil field, yet the condition still returns true
20180122 05:18:00< vultraz> also true in case of syntax error
20180122 05:19:41< celticminstrel> Maybe something in the implementation of wml_conditionals itself?
20180122 05:19:51< vultraz> possibly
20180122 05:20:01< vultraz> [if] always passes if the coniditional tag doesn't exist
20180122 05:20:03< celticminstrel> But a syntax error should enter the if, not the else...
20180122 05:20:03< vultraz> for example
20180122 05:20:05< vultraz> conditional
20180122 05:20:13< celticminstrel> Which returns false.
20180122 05:20:49< celticminstrel> The runtime error "variables is nil", on the other hand, enters the else but then raises an error, which is presumably caught somewhere in C++ and printed to the screen.
20180122 05:23:19< vultraz> syntax error went away now for some reason...
20180122 05:23:38< vultraz> ah, perhaps since I'm testing again *after* adding the `return false`
20180122 05:23:40< vultraz> ok
20180122 05:23:51< vultraz> celticminstrel: so, how do I catch runtime errors?
20180122 05:23:59< celticminstrel> With pcall.
20180122 05:24:12< vultraz> oh my
20180122 05:24:17< celticminstrel> Yeah, it's ugly.
20180122 05:24:28< vultraz> t'is getting into stuff I don't know
20180122 05:24:34< celticminstrel> IIRC, you do pcall(fcn, args) instead of fcn(args), but I can't quite remember.
20180122 05:25:53< vultraz> ok, that makes the condition fail, but also makes it so the error doesn't appear on-screen
20180122 05:25:54< vultraz> ;_;
20180122 05:26:28< celticminstrel> ...oh yeah, duh. It explicitly called error() in the syntax error case, so your return false is never actually reached.
20180122 05:27:28< celticminstrel> How to use pcall(): https://stackoverflow.com/questions/38796108/lua-error-handling-pcall
20180122 05:28:29< celticminstrel> IOW if you're checking ~~pcall(bytecode, cfg), that's not right.
20180122 05:28:44< vultraz> return pcall(bytecode, helper.get_child(cfg, "args"))
20180122 05:28:51< celticminstrel> Yeah, exactly, that's wrong.
20180122 05:29:05< vultraz> but it does fail to pass
20180122 05:29:11< celticminstrel> That returns true if there were no errors, ignoring the code's actual return value.
20180122 05:30:12< vultraz> but the problem is the errors (about variables being nil) don't get printed
20180122 05:31:01< celticminstrel> Uh, yeah, because you caught the error.
20180122 05:31:10< vultraz> gahhh
20180122 05:31:14< vultraz> I'm so confused ;|
20180122 05:31:17< celticminstrel> Just look at this example already: https://stackoverflow.com/a/38796450/1502810
20180122 05:31:26< celticminstrel> pcall() returns multiple values.
20180122 05:31:33< celticminstrel> The first value is a boolean.
20180122 05:31:43< celticminstrel> Subsequent values depend on the first.
20180122 05:31:54< celticminstrel> If the first is true, all subsequent values are whatever the function returned.
20180122 05:32:04< celticminstrel> If the first is false, then the second value is the error message.
20180122 05:32:44< celticminstrel> If you want to log the error without actually raising an error, you probably need to do something like wesnoth.log('wml', msg)
20180122 05:33:02< vultraz> I see
20180122 05:36:55< vultraz> celticminstrel: does `error` exit the function?
20180122 05:37:00< celticminstrel> Yes.
20180122 05:37:07< vultraz> so it would return true?
20180122 05:37:20< celticminstrel> Uh. It would do whatever the C++ does when the Lua code raises an error.
20180122 05:37:33< celticminstrel> If that's returning true, then sure.
20180122 05:39:02< vultraz> ok
20180122 05:39:09< vultraz> now i get the error on-screen
20180122 05:39:12< vultraz> (variables nil)
20180122 05:39:16< vultraz> no traceback tho
20180122 05:39:29< celticminstrel> Ah, yeah, that's actually rather a hard problem.
20180122 05:39:47< celticminstrel> (Getting a traceback after using pcall, I mean.)
20180122 05:39:52< vultraz> if only there were a try block in lua
20180122 05:40:04< celticminstrel> At least, I recall looking into it at one point and finding it to be hard.
20180122 05:40:12< celticminstrel> You could probably make a try block.
20180122 05:40:34< celticminstrel> Or rather a try function.
20180122 05:41:46< celticminstrel> TBH I'm not fully convinced that we want to catch this error. Maybe it's better to just let the error be raised normally.
20180122 05:42:15< vultraz> right, but I want the function to return false
20180122 05:42:20< vultraz> if there's such an error
20180122 05:42:29< celticminstrel> That could be done on the C++ side though.
20180122 05:43:03< celticminstrel> Of course, that means other custom conditions will also return false in the event of an error, though.
20180122 05:43:09< celticminstrel> I dunno if that's what we'd want.
20180122 05:44:30< vultraz> I'd say yes
20180122 05:45:03< celticminstrel> Okay, then go for it!
20180122 05:48:59< vultraz> gaaaaahhhh dammit
20180122 05:49:46< vultraz> ok I must be doing this wrong
20180122 05:50:00< vultraz> I assume I don't want to immediately exit if `luaW_pcall` returns false?
20180122 05:50:07< celticminstrel> Hm?
20180122 05:50:10< celticminstrel> Immediately exit?
20180122 05:50:39< vultraz> the conditional code is executed in the lua kernel
20180122 05:50:40< vultraz> result = resources::lua_kernel->run_wml_conditional(key, it.get_child());
20180122 05:50:49< vultraz> run_wml_conditional has a call to luaW_pcall
20180122 05:51:05< vultraz> which returns bool
20180122 05:51:30< vultraz> I thought you meant returning false if luaW_pcall returned false..
20180122 05:52:45< celticminstrel> Huh? Does it really return bool?
20180122 05:52:48< vultraz> yes
20180122 05:52:57< celticminstrel> That doesn't make sense?
20180122 05:53:12< celticminstrel> lua_pcall returns an integer representing the number of arguments returned by the invoked function.
20180122 05:53:16< vultraz> i mean, it's our function, someone made it so
20180122 05:53:18< celticminstrel> In the event of error, it returns 0.
20180122 05:53:37< celticminstrel> So, it's weird if the Wesnoth wrapper changes that to a bool, because it loses important information.
20180122 05:53:59< vultraz> luaW_pcall wraps luaW_pcall_internal which calls lua_pcall
20180122 05:54:11< vultraz> luaW_pcall_internal returns an int
20180122 05:54:37< vultraz> luaW_pcall returns true if res == 0
20180122 05:54:41< vultraz> else false
20180122 05:54:53< celticminstrel> ...
20180122 05:55:42< vultraz> hm?
20180122 05:55:44< celticminstrel> I'm so confused right now.
20180122 05:55:55< vultraz> that makes two of us
20180122 05:56:08< celticminstrel> We use Lua 5.3 right?
20180122 05:58:08< vultraz> .4 yes
20180122 05:58:08< celticminstrel> Oh hmm, maybe I misread this.
20180122 05:58:08< celticminstrel> There doesn't seem to be a 5.4 manual.
20180122 05:58:08< celticminstrel> https://www.lua.org/manual/5.3/manual.html#lua_pcall
20180122 05:58:08< vultraz> 5.3.4
20180122 05:58:08< vultraz> not 5.
20180122 05:58:08< vultraz> 4
20180122 05:58:08< celticminstrel> Ahh, right, you specify the number of return values in the call, right.
20180122 05:58:08< celticminstrel> Okay, so yeah, ==0 means success, indeed (though really we should probably use the LUA_OK constant instead).
20180122 05:58:08< vultraz> we specify via luaW_pcall's use in run_wml_conditional that we want 1 return value
20180122 05:58:34< vultraz> well, it doesn't check against == 0
20180122 05:58:38< celticminstrel> So basically, if it returns false, return false, right?
20180122 05:58:38< vultraz> it does if(errorcode)
20180122 05:58:50< celticminstrel> if(errorcode != LUA_OK) then.
20180122 06:00:53< celticminstrel> Anyway, I need to get to bed.
20180122 06:01:02< vultraz> gah, ok, I reverted my changes and now the stacktrace isn't appearing anymore!
20180122 06:01:05< vultraz> :(
20180122 06:01:14< celticminstrel> <_<
20180122 06:01:36< celticminstrel> Good night.
20180122 06:01:46-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
20180122 06:53:30< irker250> wesnoth/wesnoth:master pentarctagon d7aa7a8aa3 Have cmake not copy over everything for  AppVeyor: 1/6 builds failed
20180122 06:53:31< irker250> Details vs2017/Release: https://ci.appveyor.com/project/wesnoth/wesnoth-605wt/build/Wesnoth-VS2017-master-817
20180122 06:55:04-!- Bhoren [~Bhoren_wh@2a01:e0a:c:2150:f56e:9ee4:6b02:2778] has joined #wesnoth-dev
20180122 07:26:09-!- JyrkiVesterinen [~JyrkiVest@195.192.251.124] has joined #wesnoth-dev
20180122 08:06:02-!- atarocch [~atarocch@93.56.164.28] has joined #wesnoth-dev
20180122 09:08:39< irker250> wesnoth/wesnoth:master Wedge009 ff3412b67a Avoid type mismatch warning. AppVeyor: 1/6 builds failed
20180122 09:08:40< irker250> Details vs2017/Release: https://ci.appveyor.com/project/wesnoth/wesnoth-605wt/build/Wesnoth-VS2017-master-821
20180122 09:26:12-!- vladimirslavik [~vslavik@95.82.135.184] has joined #wesnoth-dev
20180122 10:11:45-!- galegosimpatico [~uprego@141.red-83-46-196.dynamicip.rima-tde.net] has quit [Quit: poweroff]
20180122 10:13:44-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20180122 10:53:07-!- vladimirslavik [~vslavik@95.82.135.184] has quit [Quit: Leaving]
20180122 10:55:30-!- JyrkiVesterinen [~JyrkiVest@195.192.251.124] has quit [Quit: .]
20180122 11:23:33-!- Bhoren [~Bhoren_wh@2a01:e0a:c:2150:f56e:9ee4:6b02:2778] has quit [Remote host closed the connection]
20180122 11:28:21-!- JyrkiVesterinen [~JyrkiVest@195-192-251-124.s1networks.fi] has joined #wesnoth-dev
20180122 11:29:20-!- vladimirslavik [vslavik@nat/redhat/x-laupgbaizpfbvzne] has joined #wesnoth-dev
20180122 11:35:18< vn971> Maybe add a "Blocker" badge so we don't forget to fix it before 1.14 release? https://github.com/wesnoth/wesnoth/issues/2372
20180122 11:36:13< JyrkiVesterinen> Does the bug occur every time you try to stop a replay?
20180122 11:37:31< vultraz> I could not repro that
20180122 11:41:07< vn971> I'll try to make a reproducable case then, see if it reproduces on my system at least. The backtrace should have a lot of info though, does it?
20180122 11:43:24< vn971> I tried to use Stop button right now, had a SIGSEGV again. wesnoth version is  v1.13.10+dev (bbdb5ad8-Clean)  maybe a bit outdated.
20180122 11:43:51< JyrkiVesterinen> According to your call stack, Wesnoth crashes when executing this line:
20180122 11:43:53< JyrkiVesterinen> lua_getfield(L, LUA_REGISTRYINDEX, aisKey);
20180122 11:43:57< vn971> Just opened arbitrary save file, checked the "Show replay" checkbox, then clicked "play side turn" / "stop" multiple times at random.
20180122 11:44:12< JyrkiVesterinen> The problem is that lua_getfield() should never cause a segfault, ever.
20180122 11:44:43< JyrkiVesterinen> Even if the AI context were to not exist in the Lua state, the function should just throw an exception, not segfault.
20180122 11:47:10< AI0867> which would indicate that your stack is already corrupted
20180122 11:47:37< AI0867> try to get a few more stacktraces. They should be very different if this is the case
20180122 11:48:22< vn971> AI0867: OK, np, will do. Also, I'm re-building wesnoth now from master.
20180122 11:48:36< vn971> *from master now
20180122 11:50:05-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20180122 11:57:05-!- travis-ci [~travis-ci@ec2-54-158-116-62.compute-1.amazonaws.com] has joined #wesnoth-dev
20180122 11:57:06< travis-ci> wesnoth/wesnoth#16185 (master - ff3412b : Wedge009): The build passed.
20180122 11:57:06< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/331658953
20180122 11:57:06-!- travis-ci [~travis-ci@ec2-54-158-116-62.compute-1.amazonaws.com] has left #wesnoth-dev []
20180122 12:03:35< irker250> wesnoth: Charles Dang wesnoth:master da3a0ed7cf45 / / (4 files in 3 dirs): Added [lua] conditional tag https://github.com/wesnoth/wesnoth/commit/da3a0ed7cf457eabe14a3c57bfa3c07aed075933
20180122 12:11:16< irker250> wesnoth: V N wesnoth:master f4e9ac3f8f2a / data/core/images/units/dwarves/stalwart/ (9 files): remove executable flag from images https://github.com/wesnoth/wesnoth/commit/f4e9ac3f8f2a02bee447d294c4066aab9bad013a
20180122 12:12:51< vn971> hmm, I could not repro on one file, but I managed to repro on the other. So it may be file-specific. Maybe it's even the `_` tag that we fixed lately, have to check.
20180122 12:19:53< vn971> I managed to repro it on a file that does not have the [_] tag. Maybe it has something else strangely defined, IDK. I'll attach the troublesome file on the GH issue then.
20180122 12:22:55< JyrkiVesterinen> Might be a race condition of some kind. It occurs when a lot of AI context is being teared down.
20180122 13:03:00< vn971> got distracted offline, forgot to write. Here is another bt: https://github.com/wesnoth/wesnoth/issues/2372#issuecomment-359408462 (AI0867) And here's the save file where the problem definitely does repro: https://github.com/wesnoth/wesnoth/issues/2372#issuecomment-359409062
20180122 13:04:11< JyrkiVesterinen> Hmm, same backtrace.
20180122 13:04:32< JyrkiVesterinen> That points to something other than memory corruption. Lua bug?
20180122 13:23:44< TadCarlucci> It's in a dtor my first guess would be a use-after-free
20180122 13:34:04-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20180122 13:38:12< AI0867> which is still memory corruption, but a more deterministic kind
20180122 13:38:21< AI0867> which should make it easier to debug
20180122 13:38:52< AI0867> I won't have time to take a look until at least tomorrow, so if someone else wants to, please go ahead
20180122 14:26:51< irker250> wesnoth: Charles Dang wesnoth:master f23305f7e247 / src/ (100 files in 21 dirs): Formatting cleanup: T const& -> const T& https://github.com/wesnoth/wesnoth/commit/f23305f7e247ba1cd9a8a96ab53f03c8d3ab0b04
20180122 15:09:57-!- vladimirslavik [vslavik@nat/redhat/x-laupgbaizpfbvzne] has quit [Changing host]
20180122 15:09:58-!- vladimirslavik [vslavik@wesnoth/translator/VladimirSlavik] has joined #wesnoth-dev
20180122 15:40:42-!- Netsplit *.net <-> *.split quits: syrma[m], matthiaskrgr, loonycyborg
20180122 15:41:01-!- matthiaskrgr [matthiaskr@gateway/shell/panicbnc/x-nliupunlbxslrqlf] has joined #wesnoth-dev
20180122 15:41:14-!- Netsplit over, joins: loonycyborg
20180122 15:41:15-!- JyrkiVesterinen [~JyrkiVest@195-192-251-124.s1networks.fi] has quit [Quit: .]
20180122 15:41:25-!- matthiaskrgr is now known as Guest41639
20180122 15:43:50-!- irker250 [~irker@uruz.ai0867.net] has quit [Ping timeout: 260 seconds]
20180122 15:44:43-!- syrma[m] [syrmamatri@gateway/shell/matrix.org/x-yuukgfyfjsgskrgc] has joined #wesnoth-dev
20180122 15:51:46-!- Guest41639 [matthiaskr@gateway/shell/panicbnc/x-nliupunlbxslrqlf] has quit [Changing host]
20180122 15:51:46-!- Guest41639 [matthiaskr@unaffiliated/matthiaskrgr] has joined #wesnoth-dev
20180122 15:51:46-!- Guest41639 [matthiaskr@unaffiliated/matthiaskrgr] has quit [Changing host]
20180122 15:51:46-!- Guest41639 [matthiaskr@gateway/shell/panicbnc/x-nliupunlbxslrqlf] has joined #wesnoth-dev
20180122 15:51:56-!- Guest41639 is now known as matthiaskrgr
20180122 15:52:20-!- irker962 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20180122 15:52:20< irker962> wesnoth: Gregory A Lundberg wesnoth:master 84da344dbe2d / src/ (3 files in 3 dirs): Fix errors in  f23305f7e247 https://github.com/wesnoth/wesnoth/commit/84da344dbe2d630dfb51bbf40a5f9be0d2ada94e
20180122 15:58:41-!- sigurdfd [sigurdfd@dynamic-acs-72-23-110-196.zoominternet.net] has joined #wesnoth-dev
20180122 16:29:38< irker962> wesnoth: Alexander van Gessel wesnoth:master 32119a5a05a3 / src/terrain/terrain.cpp: Initialize all members https://github.com/wesnoth/wesnoth/commit/32119a5a05a35ba7d6ec09c53b2b5023f83ec1e3
20180122 16:33:07-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20180122 16:41:29-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20180122 16:41:39-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20180122 16:50:09-!- vultraz [uid24821@wesnoth/developer/vultraz] has quit [Quit: Connection closed for inactivity]
20180122 17:08:44-!- Oebele [~quassel@143.177.58.202] has joined #wesnoth-dev
20180122 17:40:38< irker962> wesnoth/wesnoth:master V N 32a6500d5a review fixes AppVeyor: All builds passed
20180122 17:47:17-!- sigurdfd [sigurdfd@dynamic-acs-72-23-110-196.zoominternet.net] has quit []
20180122 17:48:23-!- JyrkiVesterinen [~JyrkiVest@85-23-197-3.bb.dnainternet.fi] has joined #wesnoth-dev
20180122 18:12:28-!- fabi_ [~fabi@2a02:810c:c840:2e65:4dfd:3890:2492:124f] has joined #wesnoth-dev
20180122 18:16:46-!- fabi [~fabi@wesnoth/developer/fendrin] has quit [Ping timeout: 256 seconds]
20180122 18:46:43-!- Bhoren [~Bhoren_wh@2a01:e0a:c:2150:785d:4e2c:486c:96e1] has joined #wesnoth-dev
20180122 19:01:33-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20180122 19:01:44-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20180122 19:31:49-!- vultraz [uid24821@wesnoth/developer/vultraz] has joined #wesnoth-dev
20180122 19:55:09-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 246 seconds]
20180122 20:08:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20180122 20:10:34-!- vladimirslavik [vslavik@wesnoth/translator/VladimirSlavik] has quit [Quit: Leaving]
20180122 20:24:15< irker962> wesnoth/wesnoth:master Alexander van Gessel 32119a5a05 Initialize all members AppVeyor: All builds passed
20180122 21:13:38-!- Bhoren [~Bhoren_wh@2a01:e0a:c:2150:785d:4e2c:486c:96e1] has quit [Remote host closed the connection]
20180122 21:50:55-!- vultraz [uid24821@wesnoth/developer/vultraz] has quit [Quit: Connection closed for inactivity]
20180122 22:04:13-!- JyrkiVesterinen [~JyrkiVest@85-23-197-3.bb.dnainternet.fi] has quit [Quit: .]
20180122 22:31:05-!- Oebele [~quassel@143.177.58.202] has quit [Remote host closed the connection]
20180122 22:37:54-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 246 seconds]
20180122 22:45:26-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20180122 22:46:17-!- vultraz [uid24821@wesnoth/developer/vultraz] has joined #wesnoth-dev
20180122 23:28:03-!- fabi [~fabi@2a02:810c:c840:2e65:4dfd:3890:2492:124f] has joined #wesnoth-dev
20180122 23:28:03-!- fabi [~fabi@2a02:810c:c840:2e65:4dfd:3890:2492:124f] has quit [Changing host]
20180122 23:28:03-!- fabi [~fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20180122 23:28:29-!- irker962 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20180122 23:32:24-!- fabi_ [~fabi@2a02:810c:c840:2e65:4dfd:3890:2492:124f] has quit [Ping timeout: 256 seconds]
20180122 23:39:54-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20180122 23:44:05-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Remote host closed the connection]
20180122 23:44:22-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
--- Log closed Tue Jan 23 00:00:04 2018