--- Log opened Thu Oct 13 00:00:09 2016
20161013 00:04:31-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161013 00:06:19-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 00:09:10< celticminstrel> mattsc: Uhh... that sounds pretty awful...?
20161013 00:09:17< celticminstrel> Unless it's easily suppressed.
20161013 00:10:03-!- tad_carlucci [~lundberg@173.217.65.103] has joined #wesnoth-dev
20161013 00:12:36-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161013 00:12:38< mattsc> It’s a message to the scenario designer.  Pops up, remove shroud=yes from sceanrio. will never pop up again
20161013 00:12:46-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 00:13:11< mattsc> celticminstrel: How is that different from a deprecation message?
20161013 00:13:13< celticminstrel> I don't quite get it.
20161013 00:13:26< celticminstrel> Well, a deprecation message pops up every time.
20161013 00:13:50< mattsc> Until the scenario code is fixed, yes.
20161013 00:14:19< mattsc> What’s bad about having a message that say “this scenario will not work as you imagine it” pop up when you put it together?
20161013 00:14:47< mattsc> It could have saved the poster from that thread a lot of headaches.
20161013 00:20:50< mattsc> Conceptually that
20161013 00:21:08< mattsc> ’s the same type of message as “this macro is not defined” or “you forgot the closing tag”.
20161013 00:23:28< tad_carlucci> editor/action/mouse/mouse_action.hpp:265:17: error: 'click_left' overrides a member function but is not marked 'override' [-Werror,-Winconsistent-missing-override]
20161013 00:23:42< tad_carlucci> + several more near and the same
20161013 00:24:30< tad_carlucci> I got that using scons strict with clang
20161013 00:25:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161013 00:32:05< celticminstrel> Maybe I misunderstood what you meant. If it's exactly the same as a deprecation message (other than the content), then I guess it would be fine.
20161013 00:33:29< celticminstrel> I'm getting a warning on the pragma-fix for a warning. :/
20161013 00:33:53< irker333> wesnoth: Celtic Minstrel wesnoth:master 1007f5da61fd / src/gui/dialogs/multiplayer/ (mp_join_game.cpp mp_staging.cpp): Fix XCode 4 build https://github.com/wesnoth/wesnoth/commit/1007f5da61fdd1b8c66c3b90bfe739b2dc1648d9
20161013 00:37:44< mattsc> celticminstrel: I am probably not expressing myself very well.
20161013 00:38:35< mattsc> What I mean is: add something to the MAI setup code (not the AI code itself) that pops up a warning saying: You are trying to use an MAI for a shrouded side.  That’s not going to work.  Either turn off shroud or do it differently.
20161013 00:38:41< mattsc> Paraphrased, of course.
20161013 00:38:59< celticminstrel> "pops up a warning"?
20161013 00:39:10< mattsc> on-screen message
20161013 00:39:52< celticminstrel> My interpretation of that is "a dialog with OK button".
20161013 00:40:26-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161013 00:40:33< mattsc> Okay.  Not what I meant.  Sorry for being sloppy in my formulation.
20161013 00:41:03< tad_carlucci> Not that there is anything wrong with a true popup other than it's a lot of work
20161013 00:41:05< mattsc> message = wesnoth.message() in my mind, but I see how that is not clear at all.
20161013 00:41:20< mattsc> right
20161013 00:41:56-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161013 00:49:06-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has quit [Read error: Connection reset by peer]
20161013 00:58:48-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161013 01:01:57< tad_carlucci> celticminstrel, static_assert failed "I == is_placeholder<T>::value"
20161013 01:01:58< tad_carlucci> src/gui/dialogs/preferences_dialog.cpp:402:52: note: in instantiation of template class 'std::_Bind<detail::apply<bool, bool> (boost::arg<1>)>' requested here
20161013 01:02:40-!- gfgtdf_ [~chatzilla@x4e363e64.dyn.telefonica.de] has joined #wesnoth-dev
20161013 01:04:57-!- gfgtdf [~chatzilla@x4e363a0d.dyn.telefonica.de] has quit [Ping timeout: 260 seconds]
20161013 01:05:05-!- gfgtdf_ is now known as gfgtdf
20161013 01:12:41< tad_carlucci> Oh, geez. Ticket #12397 (new Bugs) static_assert in arg.hpp failing when using boost::placeholder with std::bind Opened 8 weeks ago
20161013 01:27:27< tad_carlucci> Well, whadayaknow it worked. Created src/boost/bind/arg.hpp and applied the patch and all fixed.
20161013 01:42:02< celticminstrel> tad_carlucci: That sounds like the same error that other people have gotten (eg ancestral) and fixed by editing the Boost header.
20161013 01:42:35< tad_carlucci> celticminstrel, I dislike editing the stock headers if I can override them.
20161013 01:43:33< tad_carlucci> celticminstrel, putting it on master is a bad idea but I just want past the error. I'll move it into usr/local/include when I am done
20161013 01:57:04-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 02:00:23< mattsc> Are deprecation messages shown in debug mode only?
20161013 02:01:35-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 258 seconds]
20161013 02:05:07< celticminstrel> Not sure, maybe?
20161013 02:15:28< tad_carlucci> I seem to recall a recent discussion and that seemed to be what I heard. Could be wrong.
20161013 02:16:45-!- vultraz [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20161013 02:16:45-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20161013 02:16:45-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161013 02:18:25< gfgtdf> "tad_carlucci:	celticminstrel, putting it on master is a bad idea but I just want past the error"
20161013 02:19:01< gfgtdf> tad_carlucci: i didnt think about this eariler but why not? if it fixes compilation fo wesnoth for those systems we shodul consider that
20161013 02:19:34< tad_carlucci> Because if we forget to remove it when Boost updates and they in the future change boost/bind/arg.h we're hosed.
20161013 02:20:06< tad_carlucci> Who worked on multimenu?
20161013 02:20:34< gfgtdf> tad_carlucci: we'd have to put it in some boost version check anyway to make sure not to break older version, so i dont think this ia reason against
20161013 02:21:29< tad_carlucci> gfgtdf, I don't see how that would work.
20161013 02:22:23< tad_carlucci> vultraz, celticminstrel Const-correctness issue with multimenu in clang. A virtual override drops the const qualifier.
20161013 02:22:38< vultraz> multimenu?
20161013 02:22:40< vultraz> eh?
20161013 02:23:17< gfgtdf> tad_carlucci: we cudol just have some header included somehere before the boostargs header and sdo  "#if boost version = 1.16 then <boost 1.16 arg.hpp code>" where "<boost 1.16 arg.hpp code>" inclides the "#define BOOST_BIND_ARG_HPP_INCLUDED"
20161013 02:23:28< gfgtdf> s/sdo/do
20161013 02:23:40< gfgtdf> includes
20161013 02:23:47< vultraz> (also for the record apparently the boost issue was fixed in 1.62, i hear)
20161013 02:23:59< tad_carlucci> line 25 of multimenu.cpp SDL_Rect &rect vs base saying const
20161013 02:24:00< gfgtdf> vultraz: right i mean the boost 1.62 coide then
20161013 02:24:05< gfgtdf> code*
20161013 02:24:57< tad_carlucci> gfgtdf, Right. OK. I'll look into it. That should work. Only needed in one file that I know of so do it early and it should do the trick.
20161013 02:25:03< celticminstrel> tad_carlucci: Oh, I assumed that was a hiding warning...
20161013 02:25:21< vultraz> I don't recall anyone touching that
20161013 02:25:24< vultraz> it's gui1
20161013 02:25:26< celticminstrel> Oh wait, it is a hiding warning.
20161013 02:26:59< tad_carlucci> celticminstrel, vultraz If I change it to const the error pops up again for a call to SDL_BlitSurface who demands it not be const. I whacked the error by copy-to-temp but if the code expects the destination rect to be updated, per SDL_BlitSurface, I did the WRONG thing.
20161013 02:28:30< celticminstrel> But it's not an erroneous hiding warning like the one in menu_events.
20161013 02:28:38< tad_carlucci> Of all the clang issues .. this is the one which concerns me.
20161013 02:28:52< vultraz> oh huh
20161013 02:28:57< vultraz> I did change a const thing
20161013 02:29:01< vultraz> in that file recently
20161013 02:29:01< tad_carlucci> The menu_events were missing override qualified. Noise message.
20161013 02:29:25< celticminstrel> No, I think it was something to do with a using statement.
20161013 02:29:49< tad_carlucci> I pasted it a while ago? That the one?
20161013 02:29:51< celticminstrel> It seems that I'm the only one who gets the warning, so I'm not too concerned about it.
20161013 02:30:06< celticminstrel> Probably not?
20161013 02:32:14< tad_carlucci> I got missing override qualifiers on mouse_action and mouse-action_select today. The Boost fix. And this const thing. The const thing seems important. The rest are noise, or I have the Boost patch and will look at doing what gfgtdf suggested to work past it.
20161013 02:32:43< celticminstrel> I get a warning about hiding in menu_events.
20161013 02:32:57< celticminstrel> But the function that is supposedly being hidden is also imported with using IIRC.
20161013 02:32:58< tad_carlucci> I am still watching scons strict using clang++ instead of g++
20161013 02:33:22< celticminstrel> Yeah, looks like it is.
20161013 02:33:58< tad_carlucci> celticminstrel, I seems to recall several messages about hiding when I compile Visual Studion 2015 for Debug.
20161013 02:34:43< celticminstrel> menu_events.cpp:1075 - "events::console_handler::is_enabled" hides overloaded virtual function
20161013 02:35:02< celticminstrel> Pretty sure that's actually a lie,
20161013 02:35:51< tad_carlucci> celticminstrel, add override to the end clear it?
20161013 02:36:49< tad_carlucci> s/^/does /
20161013 02:39:12< vultraz> so, sadly, we cannot remove the multimenu until after the relase
20161013 02:39:35< vultraz> when i drop the gui1 mp stuff
20161013 02:39:46< celticminstrel> Is it an override? I didn't think it was.
20161013 02:40:31< tad_carlucci> 'virtual' is optional if you're defining an override. Pedantic compilers like clang++ will note that you forgot override, though.
20161013 02:41:41< celticminstrel> But they're not warning about it there, are they?
20161013 02:41:58< tad_carlucci> Let me check.
20161013 02:45:11< tad_carlucci> VS Code says is_enabled is virtual. Add override to the end in menu_events:1075 and see if that clears your message.
20161013 02:45:28-!- gfgtdf [~chatzilla@x4e363e64.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 49.0.1/20160922113459]]
20161013 02:48:19< celticminstrel> Nope, that gives an error.
20161013 02:48:50< celticminstrel> I'm pretty much convinced it's a compiler bug. This is clang ~3.2.
20161013 02:49:47< celticminstrel> mattsc and ancestral (the other Mac people) do not get the warning.
20161013 02:50:40< tad_carlucci> Ah. I'm running 3.8.1
20161013 02:51:18< tad_carlucci> Of course, it's possible I just haven't hit it yet.
20161013 02:51:22< celticminstrel> The "~" is because it's the Apple customized version.
20161013 02:52:19< tad_carlucci> I'm running Arch linux. I'd have to check to be positive, but Arch tried to run un-modified upstream as much as possible.
20161013 02:54:30< tad_carlucci> vultraz, So you're going to revert whatever it was which is causing the const-correctness issue I hit with multimenu.?pp
20161013 02:54:38< vultraz> No
20161013 02:55:13< vultraz> Though I suppose I could, since we're dropping it next week anyway...
20161013 02:55:25< vultraz> so it doesn't matter what it uses
20161013 03:00:52< tad_carlucci> If it behaves correctly, we could leave it in. The issue starts with an override dropping the const qualification. So it's only a problem if someone things they're using the base and expects it to be const when it isn't
20161013 03:02:30< tad_carlucci> this must be a typo in SConscipt somewhere: gcc: error: unrecognized command line option ‘-Qunused-arguments’; did you mean ‘-Wunused-argument’?
20161013 03:03:05< tad_carlucci> Or I missed something seting to use clang. New to me so that's possible.
20161013 03:20:25< mattsc> celticminstrel: Sorry, I’ve not been following along so I don’t know if this is what you’re talking about, but I’ve been getting a small number (3 or 4 or so) of warnings about overloaded virtual functions (or something along those lines) recently.
20161013 03:22:22< tad_carlucci> mattsc, I found some in mouse events, but not the one he mentioned. Checked the code and it does need override, I think.
20161013 03:25:14< mattsc> tad_carlucci, celticminstrel: http://pastebin.com/j2K52FF5
20161013 03:25:29< mattsc> That’s what I am getting from compiling sometime earlier today.
20161013 03:25:51< mattsc> Can’t stay aroudn right now though, this is always “heavy duty family time”.
20161013 03:26:06< mattsc> This ToD, I mean.
20161013 03:27:19< tad_carlucci> That is the issue vulraz and I have been discussing.
20161013 03:27:50< tad_carlucci> If I add const there it causes a problem later where it requires non-const.
20161013 03:33:54-!- irker333 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20161013 03:46:01-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20161013 03:46:07-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20161013 04:16:02-!- irker828 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20161013 04:16:02< irker828> wesnoth: Charles Dang wesnoth:master 7748822248a4 / / (3 files in 2 dirs): Removed unused GUI1 progressbar widget https://github.com/wesnoth/wesnoth/commit/7748822248a4fcac11fc419ba5b805e6f91f1790
20161013 04:23:32< irker828> wesnoth: Charles Dang wesnoth:master 80d2f0a0abff / src/widgets/ (multimenu.cpp multimenu.hpp): Revert GUI1 multimenu part of 07e993ce3e322 https://github.com/wesnoth/wesnoth/commit/80d2f0a0abffb4a4af188c339a2e8d978da5c072
20161013 04:23:33< vultraz> tad_carlucci: ^
20161013 04:23:38< vultraz> see if that fixes it.
20161013 04:25:00< vultraz> i was hesitant about reverting it since it brings back a case of blit_surface, but since it will be dropped post-release it doesn't matter.
20161013 04:25:47< tad_carlucci> Cool. Patch looks good and will certainly clear the message. I'll check once the current batch finishes. I'm trying to check the boost::bind arg.hpp bugfix.
20161013 04:26:20< vultraz> don't you need to update to boost 1.62 to do that
20161013 04:31:11< tad_carlucci> The bug is in 1.56 through 1.62 inclusive.
20161013 04:42:00-!- travis-ci [~travis-ci@ec2-54-147-58-26.compute-1.amazonaws.com] has joined #wesnoth-dev
20161013 04:42:01< travis-ci> wesnoth/wesnoth#11491 (master - 7748822 : Charles Dang): The build was broken.
20161013 04:42:01< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/167248547
20161013 04:42:01-!- travis-ci [~travis-ci@ec2-54-147-58-26.compute-1.amazonaws.com] has left #wesnoth-dev []
20161013 04:45:01-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 260 seconds]
20161013 04:49:26-!- travis-ci [~travis-ci@ec2-54-242-19-170.compute-1.amazonaws.com] has joined #wesnoth-dev
20161013 04:49:27< travis-ci> wesnoth/wesnoth#11492 (master - 80d2f0a : Charles Dang): The build was broken.
20161013 04:49:27< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/167249396
20161013 04:49:27-!- travis-ci [~travis-ci@ec2-54-242-19-170.compute-1.amazonaws.com] has left #wesnoth-dev []
20161013 05:00:34< vultraz> ...where in the recipe was that, I didn't see it
20161013 05:01:04< vultraz> ohh
20161013 05:01:12< vultraz> stupid me
20161013 05:02:10< irker828> wesnoth: Charles Dang wesnoth:master 2f05be684923 / src/ (CMakeLists.txt SConscript): Updated Scons and CMake for 7748822248a4 https://github.com/wesnoth/wesnoth/commit/2f05be68492382a52fe81c0cda7444db39dc077f
20161013 05:02:24< vultraz> (also, curses upon XCode 4)
20161013 05:03:44< vultraz> i hope celmin abandons it soon
20161013 05:08:25-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161013 05:11:03-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 268 seconds]
20161013 05:11:03-!- wedge010 is now known as wedge009
20161013 05:13:56-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 05:18:44-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20161013 05:32:27-!- Kwandulin [~Miranda@p200300760F2C712BF882C606E50F2DCB.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161013 05:38:35-!- travis-ci [~travis-ci@ec2-54-205-4-72.compute-1.amazonaws.com] has joined #wesnoth-dev
20161013 05:38:37< travis-ci> wesnoth/wesnoth#11493 (master - 2f05be6 : Charles Dang): The build was fixed.
20161013 05:38:37< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/167253361
20161013 05:38:37-!- travis-ci [~travis-ci@ec2-54-205-4-72.compute-1.amazonaws.com] has left #wesnoth-dev []
20161013 05:41:53< tad_carlucci> Well darn. I'm gonna have to finish this tomorrow. I fixed one typo and it's doing a rebuild of everything.
20161013 05:42:36< tad_carlucci> vultraz, Who knows about src/SDL_SavePNG/savepng.c ???
20161013 05:43:11< vultraz> it's an external lib, added by uh.. iceiceice I think
20161013 05:43:13< vultraz> why?
20161013 05:43:35< tad_carlucci> scons won't run because it's a C file, not C++
20161013 05:44:44< tad_carlucci> I'm going to try simply renaming it. But that will have to wait until tomorrow because I need some sleep and scons is doing a 'make all'.  *sigh**
20161013 05:44:46< shadowm> What scons version is that?
20161013 05:44:58< shadowm> And what compiler?
20161013 05:45:05< shadowm> And what's the actual output?
20161013 05:45:14< tad_carlucci> clang
20161013 05:45:32< tad_carlucci> gcc: error: unrecognized command line option ‘-Qunused-arguments’; did you mean ‘-Wunused-argument’?
20161013 05:46:09< shadowm> Okay, sounds like SCons is trying to run gcc with clang++'s arguments.
20161013 05:46:21< shadowm> Can you please tell loonycyborg about this? He's supposed to take care of the SCons recipe.
20161013 05:46:24< tad_carlucci> $ yaourt -Q scons
20161013 05:46:24< tad_carlucci> extra/scons 2.4.1-1
20161013 05:47:15< vultraz> we deliberately changed it to .c recently
20161013 05:47:18< shadowm> It probably just requires some changes so that it's aware that clang++ is the C++ compiler and clang the C compiler, or maybe a new ctool option in addition to cxxtool, I don't know.
20161013 05:47:30< shadowm> The point is that renaming it would defeat the whole point of your PR wrt Lua.
20161013 05:47:48< tad_carlucci> Why is that?
20161013 05:48:01< shadowm> Yes, never mind, just remembered that Lua must be compiled as C++.
20161013 05:48:06 * tad_carlucci nods.
20161013 05:49:05< tad_carlucci> And clang does not support forcing a .c file as .cpp from what an old commit days. If that's not true, I want to rename the kit back to .c and fix the builds instead to make it easier in the future.
20161013 05:50:44< shadowm> From the clang 3.6.2 man page: "-x language   Treat subsequent input files as having type language."
20161013 05:51:17< shadowm> I've never used it personally, though.
20161013 05:52:36< tad_carlucci> Tonights the first time I ever used clang. I just wanted to check Lua compiled cleanly with it and can't get a clean compile. I have a PR being built for most of the issues.
20161013 05:54:01< tad_carlucci> $ yaourt -Q clang
20161013 05:54:01< tad_carlucci> extra/clang 3.8.1-1
20161013 05:55:49< tad_carlucci> And that old commit message is correct: the '-x' option to select the dialect C++ for a .c file is no more.
20161013 05:56:30< vultraz> do we need savepng compiled as c++...?
20161013 05:56:32< shadowm> Then it's probably not that old.
20161013 05:57:19< shadowm> Not that I can find it. :\
20161013 05:57:56< tad_carlucci> Everything else is C++. Unless something has drastically changed, compiling a C program using C++ should work fine.
20161013 05:58:18< shadowm> But if you saw some old commit message referencing this then you probably know why savepng.c is C again?
20161013 05:59:07< tad_carlucci> The commit was the rename of Lua from .c to .cpp .. don't know about savepng.c
20161013 05:59:48< shadowm> Well, that was like half a decade ago. I'm curious why clang 3.6 and 3.8 would differ in that regard since they both postdate the commit.
20161013 05:59:54< shadowm> Anyway: https://github.com/wesnoth/wesnoth/commit/0180a72573a482683485821c8d43a82a546a86c6
20161013 06:01:47< shadowm> `clang -x c++ -c -o test1.o savepng.c -I/usr/include/SDL` seems to work for me.
20161013 06:02:23< tad_carlucci> But you're running 3.6 do we want to require everyone using clang to roll back to there?
20161013 06:02:51< tad_carlucci> That commit makes sense, btw.
20161013 06:03:38< shadowm> `clang-3.8 -x c++ -c -o test1.o savepng.c -I/usr/include/SDL`
20161013 06:03:40< shadowm> Success as well.
20161013 06:04:01< tad_carlucci> clang.org does not show -x as an option.
20161013 06:05:19< shadowm> Have you actually tried it?
20161013 06:06:03-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Remote host closed the connection]
20161013 06:07:16< shadowm> As I said, the manual page (for 3.6, and 3.8, and the online version at http://clang.llvm.org/docs/CommandGuide/clang.html ) lists it.
20161013 06:07:38< shadowm> And it seems to work for me.
20161013 06:08:39< tad_carlucci> clang-3.8: error: language not recognized: '-c'
20161013 06:08:48< shadowm> -x c++
20161013 06:08:59< shadowm> My command line follows -x c++ with -c, not -x with -c.
20161013 06:09:37< tad_carlucci> my typo
20161013 06:12:42< irker828> wesnoth: Wedge009 wesnoth:master e25a44bc726e / projectfiles/VC12/ (wesnoth.vcxproj wesnoth.vcxproj.filters): Removing progress bar from VC project files (from commit 77488222). https://github.com/wesnoth/wesnoth/commit/e25a44bc726ebc83af6747a0fbee3466b0deecbe
20161013 06:13:33< tad_carlucci> clang-3.8: warning: treating 'c' input as 'c++' when in C++ mode, this behavior is deprecated
20161013 06:13:51-!- yaiyan is now known as Yaiyan
20161013 06:14:42< tad_carlucci> If you remove the -x option. I did find it on the clang web site .. says it's for PCH generation. But not listed in the options list.
20161013 06:14:53< shadowm> http://clang.llvm.org/docs/CommandGuide/clang.html
20161013 06:16:57-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161013 06:18:47< tad_carlucci> So the problem is their documentation is inconsistent.
20161013 06:19:10< shadowm> I'd call it incomplete, but sure.
20161013 06:19:15< tad_carlucci> It's on the man pages but not in the user's manual.
20161013 06:20:30< tad_carlucci> So the issue is simply using clang++ and adding '-x c' for that file. If nobody fixes it overnight I'll take care of it tomorrow.
20161013 06:20:53< tad_carlucci> Meanwhile I guess I can push up the changes I have for consideration.
20161013 06:21:07-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20161013 06:21:24< shadowm> I mean, I'm not a fan of dropping people's names repeatedly, but I mentioned loonycyborg in case he has a different notion of how it would be est done with SCons.
20161013 06:21:27< shadowm> *best
20161013 06:21:38< shadowm> You know, the "not stepping on other people's toes" principle.
20161013 06:24:15< tad_carlucci> I'll do it as a separate PR. If he disappoves, don't merge it and I'll delete it.
20161013 06:32:15< vultraz> tad_carlucci: you mean the boost fix is not in 1.62? :/
20161013 06:33:43< tad_carlucci> vultraz, The commit is not tagged for 1.62 and I don't have 1.62 for arch showing as an update. I could hand check the download but I don't expect it to be there. The Ticket was 8 weeks ago. The patch is dated August.
20161013 06:34:26< vultraz> 1.62 came out sept 8 :/
20161013 06:34:36< vultraz> th
20161013 06:34:39< tad_carlucci> Missed it by that much.
20161013 06:34:48< vultraz> 28*th
20161013 06:34:51< vultraz> >_>
20161013 06:35:20< vultraz> ie, 2 weeks ago
20161013 06:36:45< tad_carlucci> Hang on. Downloading 1.62.  Don't understand why Arch doesn't have 1.62 up yet. Usually takes only 24-48 hours to appear after an upstream release.
20161013 06:38:28< vultraz> .. you're right, the patch doesn't seem to be there
20161013 06:38:30< vultraz> whaaa??
20161013 06:39:15< tad_carlucci> Nope.
20161013 06:39:31< tad_carlucci> Well, the RM must have frozen before the commit went it.
20161013 06:39:32< vultraz> How is the patch from August and the release in September but it not in it :|
20161013 06:39:52< vultraz> a month beforehand?
20161013 06:40:01 * tad_carlucci shrugs
20161013 06:40:09< tad_carlucci> It is what it is.
20161013 06:42:00< vultraz> anyway, is there any reason your patched file cannot be included in utils/functional.hpp
20161013 06:43:08< tad_carlucci> No reason at all other than it documents the files it actually effects and keeps the clutter down for everywhere else.
20161013 06:43:26< vultraz> (I'm also slightly curious whether boost ever intents to drop stuff like bind that is now in the standard library)
20161013 06:43:59< tad_carlucci> I doubt it. Probably afraid of breaking legacies.
20161013 06:44:09 * vultraz rolls eyes
20161013 06:44:16< tad_carlucci> A **LOT** of code uses Boost.
20161013 06:44:27< vultraz> granted
20161013 06:44:41< vultraz> I suppose getting rid of stuff is harder when you write stuff a lot of people use
20161013 06:44:48< vultraz> maybe they'll drop it in 10 years :P
20161013 06:44:52< tad_carlucci> I wish they'd fix that darned deprecation note too
20161013 06:45:35< tad_carlucci> vultraz, There's a reddit about that today about Java and deprecation.
20161013 06:47:04< tad_carlucci> Anyway, I'll check for loonycyborg and speak with him about the savepng thing if he's around tomorrow otherwise I'll just do a PR to have a hand up that it need attention.
20161013 06:47:29< tad_carlucci> Oh, and your commit did clear the error, vultraz.
20161013 06:47:42< vultraz> ah
20161013 06:47:45< vultraz> good
20161013 06:47:54 * tad_carlucci waves
20161013 06:47:57-!- tad_carlucci [~lundberg@173.217.65.103] has quit [Quit: Leaving]
20161013 06:49:39-!- mjs-de [~mjs-de@x4e319152.dyn.telefonica.de] has joined #wesnoth-dev
20161013 06:53:35-!- mjs-de [~mjs-de@x4e319152.dyn.telefonica.de] has quit [Remote host closed the connection]
20161013 06:56:50-!- Kwandulin [~Miranda@p200300760F2C712BF882C606E50F2DCB.dip0.t-ipconnect.de] has quit [Ping timeout: 258 seconds]
20161013 07:02:14-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 07:06:53-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20161013 07:26:12-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 260 seconds]
20161013 07:29:40< celticminstrel> [Oct 13@01:02:24am] vultraz: (also, curses upon XCode 4)
20161013 07:29:41< celticminstrel> [Oct 13@01:03:44am] vultraz: i hope celmin abandons it soon
20161013 07:29:42< celticminstrel> So you'd prefer I do all Wesnoth dev work on Windows?
20161013 07:30:17-!- boucman_work [~boucman@159.195.204.77.rev.sfr.net] has joined #wesnoth-dev
20161013 07:30:33< vultraz> well, I'd prefer if {} worked for everyone
20161013 07:34:21< JyrkiVesterinen> celticminstrel: I'd advise you to stop using OS X 10.7 anyway. It's a dead end, even major web browsers have stopped supporting it.
20161013 07:34:30< JyrkiVesterinen> GNU/Linux is free.
20161013 07:34:58< celticminstrel> Yeah, I noticed Firefox stopped supporting it... that's why I downgraded back to 47.
20161013 07:35:35< celticminstrel> Almost all my stuff is on the Mac though. It doesn't really feel worthwhile to move it (and in some cases it can't be done).
20161013 07:38:44< vultraz> You could put in a request for a new mac.
20161013 07:39:07-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161013 07:39:16< celticminstrel> Uh. Why?
20161013 07:40:46< vultraz> Since your mac is from 2006 or something?
20161013 07:40:57< celticminstrel> It is from 2006, yes.
20161013 07:42:30< vultraz> Given that you're one of our most active (and few remaining) code contributors, I'm sure the rest of the board would approve such an expenditure.
20161013 07:42:52< celticminstrel> But it's not as if I need a Mac for developing.
20161013 07:43:30< celticminstrel> Actually I could probably try using CB or something on the Mac with a newer compiler, even.
20161013 07:43:59-!- Greywhind [~Greywhind@c-71-232-29-126.hsd1.ma.comcast.net] has quit [Ping timeout: 258 seconds]
20161013 07:46:37< vultraz> building anything on a decade-old machine is probably a pain
20161013 07:47:27< vultraz> also, sure, you could use your windows rig, but we need an active dev on mac os
20161013 07:51:51< vultraz> ironically, i think right now we have more devs on windows than any other platform
20161013 07:52:44-!- Greywhind [~Greywhind@c-71-232-29-126.hsd1.ma.comcast.net] has joined #wesnoth-dev
20161013 07:53:45< shadowm> Having a C++ coder on OS X is extremely important indeed. That was one of the things we missed the most for 1.11.x and early 1.13.x.
20161013 07:55:39< vultraz> ^
20161013 07:56:27< vultraz> Given that, I certainly think we'd approve the purchase of a new mac to ensure you can continue developing on the platform.
20161013 07:57:26< vultraz> Especially important since I don't know for how much longer ancestral will be helping with packaging.
20161013 07:57:47< shadowm> Emphasis on C++. I think we've always had devs on OS X, but they were generally unable to fix or debug things on their own without help from non-Mac devs. Walking people on Mac through debugging procedures when you don't know how to use a Mac yourself isn't the most optimal thing to do.
20161013 07:58:07< shadowm> (Just reminding vultraz that not all devs can C++.)
20161013 07:58:40< vultraz> I am aware.
20161013 07:58:59< celticminstrel> Well, unless Wesnoth gets hardware acceleration requiring 3.x, I should still be able to use this Mac for it. If it becomes a problem to be using an ancient compiler, that can be worked around.
20161013 07:59:12< celticminstrel> By simply not using XCode.
20161013 07:59:18< celticminstrel> For example, I could build with scons instead.
20161013 08:00:36< vultraz> Well, up to you.
20161013 08:04:31-!- celticminstrel is now known as celmin|sleep
20161013 08:12:45-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20161013 08:26:25< irker828> wesnoth: Ignacio R. Morelle wesnoth:1.12 7c56e8179c78 / data/core/about.cfg: Fix credits https://github.com/wesnoth/wesnoth/commit/7c56e8179c7827c3f01be095a84fff065478596f
20161013 08:26:28< irker828> wesnoth: Ignacio R. Morelle wesnoth:master 4c5b7d1f8726 / data/core/about.cfg: Fix credits https://github.com/wesnoth/wesnoth/commit/4c5b7d1f8726544514a00a55844c9454bbd81463
20161013 08:26:49< vultraz> wha?
20161013 08:27:14< vultraz> wrong person or misspelling or..
20161013 08:30:33< irker828> wesnoth: Ignacio R. Morelle wesnoth:master 66915f4b9370 / data/tools/about_cfg_to_wiki: Update about_cfg_to_wiki to handle U+2022 BULLET correctly https://github.com/wesnoth/wesnoth/commit/66915f4b93705094b8db4f4610cfbb9e4d85813d
20161013 08:32:24< JyrkiVesterinen> vultraz: I found the breaking commit with GitHub blame: https://github.com/wesnoth/wesnoth/commit/5f3f46e0dae14d03d665a4ae01f667b64c02f3b4
20161013 08:32:42< vultraz> huh
20161013 08:33:07< irker828> wesnoth: Charles Dang wesnoth:master b19443483cdb / data/multiplayer/scenarios/2p_Aethermaw.cfg: Aethermaw: removed some excessive newlines in starting message https://github.com/wesnoth/wesnoth/commit/b19443483cdb5a506d534766cccdba6d08d3a488
20161013 08:33:56< shadowm> vultraz: Next time you or your pet technician of the day make a release, please, update Credits in the wiki. The instructions are in ReleasingWesnoth.
20161013 08:34:09< vultraz> OK
20161013 08:34:30< shadowm> The previous edit was on December 2015 when I released 1.13.2.
20161013 08:34:46< shadowm> I refuse to believe that there haven't been credit edits in any development releases since then.
20161013 08:35:31< shadowm> Also, maybe poke me if you ever decide to change the credits file format again like you did with the bullets.
20161013 08:35:54< vultraz> Ok
20161013 08:38:19< shadowm> When you finish working on the GUI2 Add-ons Manager, you should consider adding a user interface for the change-passphrase command.
20161013 08:38:48< vultraz> Why did you tell me months ago to not touch the networking side of that?
20161013 08:38:50< shadowm> <vultraz> change passphrase command?
20161013 08:38:57< shadowm> Yes, there is a change passphrase command.
20161013 08:39:41< shadowm> vultraz: I don't care anymore I guess?
20161013 08:40:04< shadowm> Like, you probably _mostly_ know what you are doing by this point, maybe?
20161013 08:41:17< shadowm> Just as long as you don't end up deploying functionality to turn clients into open proxies, I don't see why you shouldn't be allowed to play around with the second-most frustrating aspect of game development.
20161013 08:41:52< vultraz> Alright
20161013 08:41:54< shadowm> Besides, you are my boss now.
20161013 08:42:22< irker828> wesnoth: Charles Dang wesnoth:master 403c0c624c15 / src/gui/dialogs/multiplayer/mp_create_game.cpp: MP Create Game: skip exit hook if loading saved game https://github.com/wesnoth/wesnoth/commit/403c0c624c15949afe33b295bab9e22d12a16476
20161013 08:42:42< vultraz> Do you have any critiques on the current tooltips?
20161013 08:42:44< shadowm> I can't disallow you to do anything at all, only voice my objections as loud as possible and pray that they won't be entirely ignored.
20161013 08:43:50< shadowm> They look unlike anything else.
20161013 08:44:02< shadowm> Haven't decided how I feel about that.
20161013 08:44:13-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Quit: wedge009]
20161013 08:44:16< vultraz> That's rather the point.
20161013 08:44:27< shadowm> (It can be a good thing, it can be a bad thing. It depends.)
20161013 08:44:32-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161013 08:44:45< vultraz> It felt too visually cluttered to have another black box outlined in gold.
20161013 08:45:29< shadowm> Yes, that wasn't going to work.
20161013 08:46:03< vultraz> Especially once I started using it for longer blocks of text
20161013 08:46:14< vultraz> (the era and mod descriptions in Create are displayed as tooltips)
20161013 08:46:58< shadowm> They need to stand out against the mostly dark UI elements we use, so that's a step in the right direction.
20161013 08:47:27< shadowm> But you used 3D shading for their borders, which seems a bit backwards?
20161013 08:48:38< vultraz> top/right is deliberately subtle. the left ties the entire thing together.
20161013 08:48:40< shadowm> You normally seem to prefer the crappy retro^Wmodern flat aesthetic that's become the norm thanks to mobile platforms.
20161013 08:49:17< shadowm> Which is why I didn't really bother trying to shade the textbox borders.
20161013 08:49:31< vultraz> yeah, no, don't do that
20161013 08:50:32-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 08:51:16< shadowm> Have you considered using {GUI__BORDER_COLOR_DARK} for those [drawing]-based dividers you like so much?
20161013 08:51:58< vultraz> hmmmm
20161013 08:52:13< vultraz> let's see how that looks
20161013 08:52:14< shadowm> The Log In dialog becomes unusuable due to Wesnoth lagging to hell when it is required during the loadscreen previous to joining the MP lobby, by the way.
20161013 08:52:34< shadowm> Increasingly so with each consecutive login attempt, it seems.
20161013 08:52:59< shadowm> Now I can't input anything even though the UI still fires timer events.
20161013 08:53:13< vultraz> Cannot reproduce.
20161013 08:53:23< shadowm> Even if you try to login as "shadowm"?
20161013 08:53:50< shadowm> e.g. Multiplayer → enter some registered name you don't have access to in the Login box above the menu, Join Official Server, etc.
20161013 08:54:13< vultraz> yup, still works fine
20161013 08:54:27< vultraz> I suspect you've hit one of those random, indeterminate GUI2 lag spikes
20161013 08:54:39< shadowm> It's not just a lag spike for me.
20161013 08:54:49-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20161013 08:54:53< shadowm> As I said, GUI2 timer events keep firing with the expected accuracy.
20161013 08:54:54-!- boucman_work [~boucman@159.195.204.77.rev.sfr.net] has quit [Ping timeout: 268 seconds]
20161013 08:55:01< vultraz> For some reason, at random times,GUI2 will becomes unusable due to input taking ages to register
20161013 08:55:05< shadowm> It's mouse and keyboard input that ceases to be processed correctly.
20161013 08:55:30< shadowm> And the thing that makes it suspicious is that the loadscreen dialog is obviously behind.
20161013 08:55:42< shadowm> On the screen. No idea if it's actually running.
20161013 08:56:22< shadowm> Considering how much I tortured GUI2 while on the editor and I didn't get any of this there...
20161013 08:56:31< shadowm> *while working on the file dialog
20161013 08:56:48< vultraz> It's entirely random
20161013 08:56:52-!- boucman_work [~boucman@159.195.204.77.rev.sfr.net] has joined #wesnoth-dev
20161013 08:57:02< irker828> wesnoth: ln-zookeeper wesnoth:master b24477b23378 / data/core/terrain.cfg: Tweaked submerge and unit_height_adjust values for aquatic castle/camp https://github.com/wesnoth/wesnoth/commit/b24477b233785135558b6fe11cf89fe54538e649
20161013 08:57:04< shadowm> It's not random at all for me.
20161013 08:58:15< zookeeper> oh duh, i'm stupid
20161013 08:58:42< zookeeper> of course the camp keep wasn't supposed to have it
20161013 08:58:49< shadowm> vultraz: Is there a bug report for this yet?
20161013 08:58:55< vultraz> No
20161013 08:59:01< vultraz> t that I know of
20161013 08:59:29< shadowm> It's a deal-breaker for me and anyone else who doesn't store their password in preferences.
20161013 09:00:31< shadowm> Also, it's really sad that 1.13.6 will be released with known memory management bugs.
20161013 09:01:24< vultraz> No one has gotten to them yet
20161013 09:01:39< shadowm> Hence sad.
20161013 09:02:02< zookeeper> WRT tooltips, having a box with a very prominent golden left edge is pretty odd. can it be reversed for RTL languages? :p
20161013 09:02:10< shadowm> Especially since this kind of thing interferes with other testing and debugging tasks.
20161013 09:02:33< shadowm> It's essentially like trying to perform a heart surgery in a barn.
20161013 09:02:35< vultraz> it could be reversed with a little work
20161013 09:03:08< shadowm> You should try to encourage people to look at those bugs.
20161013 09:03:16< JyrkiVesterinen> Well, I think it's more important to fix user-visible bugs. It's better to have dozens of unnoticeable memory management issues than even one visible glitch.
20161013 09:03:43< shadowm> I strongly disagree based on past experience.
20161013 09:03:59< shadowm> User-visible bugs are important, sure, but UB tends to introduce user-visible bugs that no-one can reliably reproduce or track down.
20161013 09:04:20< irker828> wesnoth: ln-zookeeper wesnoth:master 6eb6220360e2 / data/core/terrain.cfg: Reverted the addition of submerge for Aquatic Encampment Keep https://github.com/wesnoth/wesnoth/commit/6eb6220360e2204918f3da205ccd238fd2f99f2d
20161013 09:04:29< vultraz> I'll try to get people to work on them for 1.13.7
20161013 09:04:36< JyrkiVesterinen> Most of the time UB is harmless. (Of course it results in highly problematic user-visible bugs sometimes.)
20161013 09:04:40< vultraz> problem is, people's plates are full.
20161013 09:04:52< vultraz> I'm having trouble getting people to look at an actual blocker :|
20161013 09:05:03< vultraz> (flg dialog not showing when joining an mp game)
20161013 09:05:08-!- atarocch [~atarocch@93.56.160.28] has quit [Ping timeout: 260 seconds]
20161013 09:05:11< JyrkiVesterinen> Not that I'm saying that UB shouldn't be fixed. It's just better to focus on other stuff instead, IMHO.
20161013 09:05:16< vultraz> (and wesnoth instead freezing)
20161013 09:05:26-!- atarocch [~atarocch@93.56.160.28] has joined #wesnoth-dev
20161013 09:07:17< shadowm> The other point is that those logs that Matthias collected will quickly become outdated and require everyone involved to try to get new ones.
20161013 09:07:49< JyrkiVesterinen> Getting the logs is not that hard as long as there are steps to reproduce.
20161013 09:08:02< shadowm> Changes on top of bugs can mask (not necessarily fix) previous bugs or create new ones.
20161013 09:08:15< JyrkiVesterinen> Compile the game with ASan/UBSan enabled and follor the STR. Done.
20161013 09:09:21< JyrkiVesterinen> And I think changes on top of bugs aren't that big of a problem either. The Wesnoth codebase is huge. A lot of code remains alone regardless of the flux of changes.
20161013 09:09:57< JyrkiVesterinen> ("Unchanged" is probably better word than "alone", but you probably get my point.)
20161013 09:10:49< shadowm> Oh well.
20161013 09:13:43-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has joined #wesnoth-dev
20161013 09:17:25-!- Netsplit *.net <-> *.split quits: Lohengramm, TheJJ, _laco, Elsi, Ivanovic, kidneb_, boucman_work, shadowm, Sirp_, Soliton,  (+17 more, use /NETSPLIT to show all of them)
20161013 09:20:10-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Ping timeout: 259 seconds]
20161013 09:29:51-!- Nobun [~nobun@5.170.105.207] has joined #wesnoth-dev
20161013 09:30:39< Nobun> little question about pot: I wrote "Content-Type: text/plain; charset=UTF-8\\n" in wmlxgettext, emulating what perl_wmlxgettext did
20161013 09:31:01< Nobun> but I'm asking myself if it is better to switch 'UTF-8' to lower case 'utf-8'
20161013 09:48:53-!- Nobun [~nobun@5.170.105.207] has quit [Ping timeout: 248 seconds]
20161013 10:06:43-!- Duthlet [~Duthlet@dslb-188-104-253-155.188.104.pools.vodafone-ip.de] has joined #wesnoth-dev
20161013 10:07:04-!- Netsplit over, joins: zookeeper, Greywhind, crimson_penguin
20161013 10:08:00-!- Netsplit over, joins: minzbonbon
20161013 10:09:48-!- Netsplit over, joins: boucman_work, Elsi, Sirp_, DeFender1031, Yaiyan, esr, matthiaskrgr
20161013 10:10:10-!- Netsplit over, joins: _laco
20161013 10:10:40-!- Netsplit *.net <-> *.split quits: _laco, Elsi, boucman_work, Sirp_, Yaiyan, DeFender1031, Greywhind, esr, matthiaskrgr, zookeeper,  (+2 more, use /NETSPLIT to show all of them)
20161013 10:27:01-!- Netsplit over, joins: minzbonbon
20161013 10:27:15-!- Netsplit over, joins: _laco
20161013 10:27:24-!- Netsplit over, joins: zookeeper, Greywhind, crimson_penguin
20161013 10:27:56-!- Netsplit over, joins: boucman_work, Elsi, Sirp_, DeFender1031, Yaiyan, esr, matthiaskrgr
20161013 10:28:04-!- Nobun [~nobun@5.170.107.227] has joined #wesnoth-dev
20161013 10:28:05-!- clavii [~clavi@163-172-10-77.rev.poneytelecom.eu] has joined #wesnoth-dev
20161013 10:28:05-!- shadowm [~ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20161013 10:28:05-!- APic [apic@apic.name] has joined #wesnoth-dev
20161013 10:28:05-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20161013 10:28:05-!- kidneb_ [~kidneb@178.79.173.107] has joined #wesnoth-dev
20161013 10:28:05-!- TheJJ [~rofl@ipbcc36896.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20161013 10:28:26-!- tomreyn [~tomreyn@archmage.megaglest.org] has joined #wesnoth-dev
20161013 10:28:26-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has joined #wesnoth-dev
20161013 10:28:26-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20161013 10:28:26-!- EliDupree2 [~quassel@2604:a880:400:d0::9bb:2001] has joined #wesnoth-dev
20161013 10:28:26-!- Soliton [~Soliton@wesnoth/developer/soliton] has joined #wesnoth-dev
20161013 10:28:26-!- Lohengramm [sid1929@gateway/web/irccloud.com/x-dadfvjpayltsiixw] has joined #wesnoth-dev
20161013 10:28:26-!- elias [~allefant@allegro/developer/allefant] has joined #wesnoth-dev
20161013 10:28:26-!- Jetrel [~Jetrel@2001:558:6014:1e:2422:435:dd84:bbf3] has joined #wesnoth-dev
20161013 10:28:26-!- nurupo [~nurupo.ga@unaffiliated/nurupo] has joined #wesnoth-dev
20161013 10:28:38-!- APic [apic@apic.name] has quit [Max SendQ exceeded]
20161013 10:29:07-!- APic [apic@apic.name] has joined #wesnoth-dev
20161013 10:32:14-!- Duthlet [~Duthlet@dslb-188-104-253-155.188.104.pools.vodafone-ip.de] has quit [Quit: leaving]
20161013 10:32:28-!- Duthlet_ [~Duthlet@dslb-188-104-253-155.188.104.pools.vodafone-ip.de] has joined #wesnoth-dev
20161013 10:32:28-!- Duthlet_ [~Duthlet@dslb-188-104-253-155.188.104.pools.vodafone-ip.de] has quit [Client Quit]
20161013 10:32:50-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161013 10:33:37< DeFender1031> vultraz, around?
20161013 10:34:37-!- Duthlet [~Duthlet@dslb-188-104-253-155.188.104.pools.vodafone-ip.de] has joined #wesnoth-dev
20161013 10:38:48-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 10:39:04-!- Nobun1 [~nobun@5.170.109.144] has joined #wesnoth-dev
20161013 10:39:15-!- Nobun1 is now known as Nobun_1
20161013 10:39:31-!- Nobun [~nobun@5.170.107.227] has quit [Disconnected by services]
20161013 10:39:51-!- Nobun_1 [~nobun@5.170.109.144] has quit [Client Quit]
20161013 10:43:46-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20161013 10:45:52-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20161013 10:56:25-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 258 seconds]
20161013 10:56:30-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161013 10:59:36-!- travis-ci [~travis-ci@ec2-54-198-105-107.compute-1.amazonaws.com] has joined #wesnoth-dev
20161013 10:59:37< travis-ci> wesnoth/wesnoth#11501 (master - b24477b : ln-zookeeper): The build has errored.
20161013 10:59:38< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/167291206
20161013 10:59:38-!- travis-ci [~travis-ci@ec2-54-198-105-107.compute-1.amazonaws.com] has left #wesnoth-dev []
20161013 11:04:28-!- Rhonda [~rhonda@anguilla.debian.or.at] has quit [Changing host]
20161013 11:04:28-!- Rhonda [~rhonda@wesnoth/developer/rhonda] has joined #wesnoth-dev
20161013 11:06:01< DeFender1031> Hmm... Apparently [lock_view] only prevents scrolling, not zooming or using the minimap to change the viewing position...
20161013 11:08:14-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20161013 11:30:15< vultraz> DeFender1031: hm?
20161013 11:30:24< vultraz> DeFender1031: yes, this is true
20161013 11:30:56< DeFender1031> Intentional/
20161013 11:30:58< DeFender1031> ?*
20161013 11:31:07< vultraz> yes
20161013 11:31:23< DeFender1031> Why's that?
20161013 11:31:52< vultraz> was never really considered that pertinent to implement it 
20161013 11:33:12< DeFender1031> Okay then. Anyway, the reason I actually pinged you is because I put together a testcase for the ~TC bug I mentioned the other day, and discovered some additional information about the bug's behavior that makes me think I know what the bug is, but not where it is.
20161013 11:33:24< DeFender1031> Anyway, testcase: http://paste.nachsoftware.com/DeFender1031/YSBHL48208c1e1dd7565e2de357fe836a06bbdFL
20161013 11:33:33< DeFender1031> Result: https://drive.google.com/file/d/0B5F6wWfcJTDPZXVCeUJnN3dfclE/view
20161013 11:34:26< vultraz> whaaa...
20161013 11:34:44< DeFender1031> I also noticed that when commenting out the definition for side 7, It ends up coloring side 6 brown.
20161013 11:35:03< DeFender1031> And that if adding a side 8, it colors 7 correctly, but screws up 8
20161013 11:36:07< DeFender1031> This leads me to believe that there's some line of code that looks something like "if(sideNumber < numSides)" that should be "if(sideNum <= numSides)"
20161013 11:36:25< DeFender1031> Though I don't have the first clue of where to begin to look or how to find out.
20161013 11:37:07< DeFender1031> (If I knew where to look for IPF code, I'd have tried to track the actual bug down myself)
20161013 11:37:33< vultraz> here, maybe? https://github.com/wesnoth/wesnoth/blob/master/src/image_modifications.cpp#L623
20161013 11:38:04< DeFender1031> That looks exactly like what I imagined
20161013 11:38:28< vultraz> especially since it then does side_n - 1
20161013 11:38:35< DeFender1031> Wow. I love the incredibly descriptive error message on like 631.
20161013 11:38:40< vultraz> xD
20161013 11:38:58< vultraz> Let me change that line and run your test case
20161013 11:39:08< DeFender1031> and yes. It was clear from what I was seeing that it did the -1 correctly, but only after passing the check.
20161013 11:39:39< DeFender1031> perhaps run the testcase first to make sure that it's still an issue and that some other weirdness elsewhere hasn't masked it somehow?
20161013 11:43:24< DeFender1031> (Though I doubt it)
20161013 11:48:58< irker828> wesnoth: Charles Dang wesnoth:master 36be1ce21ee6 / src/image_modifications.cpp: Fixed incorrect color result when using ~TC() with number of last defined side https://github.com/wesnoth/wesnoth/commit/36be1ce21ee6a95e329ed4fd2acd5a718d687e65
20161013 11:49:00< vultraz> DeFender1031: ^ fixed
20161013 11:49:05< vultraz> you were exactly right
20161013 11:51:02< vultraz> talk about a corner case, though
20161013 11:51:45< DeFender1031> I'd call it more of an off-by-one than a corner case, even though technically it only happens on the corner.
20161013 11:52:04< vultraz> heh
20161013 11:52:20< DeFender1031> And yes, when I saw the block of code you linked, it looked pretty much identical to what I imagined it looked like.
20161013 11:53:44< vultraz> yeah, i had just been looking at the code and then when you mentioned your theory i realized that bit was a very likely possibility
20161013 11:54:59-!- Appleman1234 [~Appleman1@KD106154001050.au-net.ne.jp] has joined #wesnoth-dev
20161013 11:56:07< DeFender1031> Ah. I wondered how you found it so fast. It was like 30 seconds after I stated my theory.
20161013 11:56:36-!- stikonas_ is now known as stikonas
20161013 11:57:27< vultraz> i should clean up the code in that function while im looking at it..
20161013 12:01:58< vultraz> ... huh... is T const& foo different from const T& foo?
20161013 12:03:53< DeFender1031> I seem to recall that it is, but I don't recall the distinction
20161013 12:05:12< vultraz> looks like it's just a stylistic thing
20161013 12:05:39< DeFender1031> anyway, it's also not surprising that there are off-by-one errors in wesnoth, given how there are a lot of things that are converted between 0-indexed and 1-indexed depending on context
20161013 12:06:23< DeFender1031> Hmm... I know with pointers, it changes whether the const applies to the address or the dereference.
20161013 12:10:21< Soliton> T& const would be different but references themselves are const anyway.
20161013 12:11:42< DeFender1031> ohhh
20161013 12:11:46< DeFender1031> right. Of course
20161013 12:14:09-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 258 seconds]
20161013 12:50:50-!- Kwandulin [~Miranda@p200300760F2C7134189C41C098A0B8C9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161013 12:57:10-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20161013 13:07:44-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161013 13:08:01-!- Appleman1234_ [~Appleman1@KD106154005182.au-net.ne.jp] has joined #wesnoth-dev
20161013 13:09:53-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20161013 13:11:13-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161013 13:11:26-!- Appleman1234 [~Appleman1@KD106154001050.au-net.ne.jp] has quit [Ping timeout: 268 seconds]
20161013 13:11:34-!- Appleman1234_ is now known as Appleman1234
20161013 13:12:40-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 268 seconds]
20161013 13:38:15-!- Kwandulin [~Miranda@p200300760F2C7134189C41C098A0B8C9.dip0.t-ipconnect.de] has quit [Read error: Network is unreachable]
20161013 13:42:00-!- Nobun [~nobun@5.170.112.99] has joined #wesnoth-dev
20161013 13:51:35-!- Nobun [~nobun@5.170.112.99] has quit [Quit: Salve a tutti]
20161013 13:52:06-!- Kwandulin [~Miranda@p200300760F2C7134189C41C098A0B8C9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20161013 13:53:22-!- gfgtdf [~chatzilla@x4e363e64.dyn.telefonica.de] has joined #wesnoth-dev
20161013 13:54:45-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 248 seconds]
20161013 13:55:37< vultraz> hmmm
20161013 13:55:57< vultraz> what's a good way to set a sorting option to initially selected in a listbox..
20161013 13:57:32-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161013 13:58:42-!- stikonas_ is now known as stikonas
20161013 14:02:40< vultraz> manually selecting the appropriate widget comes to mind..
20161013 14:03:03< vultraz> or...
20161013 14:03:05< vultraz> hmm
20161013 14:04:09< vultraz> ah, no
20161013 14:11:42-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20161013 14:17:05< vultraz> hm, so you can actually manually sort with tlistbox::order_by
20161013 14:17:46< vultraz> that's public
20161013 14:17:56< vultraz> but that wouldn't actually select the header button
20161013 14:18:01< vultraz> new method!
20161013 14:49:57-!- irker828 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20161013 14:53:33-!- Jetrel [~Jetrel@2001:558:6014:1e:2422:435:dd84:bbf3] has quit [Read error: Connection reset by peer]
20161013 14:54:03-!- Jetrel [~Jetrel@2001:558:6014:1e:2422:435:dd84:bbf3] has joined #wesnoth-dev
20161013 14:57:13-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20161013 15:21:00-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20161013 15:21:06-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20161013 15:43:20-!- Nobun [~nobun@5.170.104.76] has joined #wesnoth-dev
20161013 15:51:22-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20161013 16:04:48-!- Kwandulin [~Miranda@p200300760F2C7134189C41C098A0B8C9.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20161013 16:05:59< celmin|sleep> [Oct 13@08:01:58am] vultraz: ... huh... is T const& foo different from const T& foo?
20161013 16:06:00< celmin|sleep> No, it's the same.
20161013 16:06:01-!- JyrkiVesterinen [~JyrkiVest@87-100-193-57.bb.dnainternet.fi] has joined #wesnoth-dev
20161013 16:06:39-!- celmin|sleep is now known as celticminstrel
20161013 16:14:45-!- boucman_work [~boucman@159.195.204.77.rev.sfr.net] has quit [Ping timeout: 260 seconds]
20161013 16:22:03-!- irker663 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20161013 16:22:03< irker663> wesnoth: Jyrki Vesterinen wesnoth:master 2cc92e881ed0 / src/gui/widgets/listbox.cpp: Don't try to suppress -Wmaybe-uninitialized with Clang https://github.com/wesnoth/wesnoth/commit/2cc92e881ed035e4403daaea5c06f7fd8573c6a5
20161013 16:22:22< mattsc> celticminstrel: I have been thinking about this Micro AI and shroud thing a bit more.  I’ll add a note to the wiki page that MAIs are not guaranteed to work with shroud, that’s not a question.  The question I am having is about adding an on-screen warning.
20161013 16:22:26< celticminstrel> ...huh?
20161013 16:22:36< mattsc> I’ll do that in debug mode only, so that’s one thing.
20161013 16:22:46< celticminstrel> Does clang simply not support that warning at all ever?
20161013 16:23:07< JyrkiVesterinen> I don't know. Hold on, I'll try to find the log from the IRC logs.
20161013 16:23:21< mattsc> But then, somebody _might_ want to use an AI side with an MAI and shroud.  I do not know why, but people have often surprised me …
20161013 16:23:44< JyrkiVesterinen> celticminstrel, here is mattsc's log: http://pastebin.com/j2K52FF5
20161013 16:23:55< mattsc> So, an option to deal with that would be to add a flag suppress_warnings or something to the [micro_ai] tag.
20161013 16:24:16< JyrkiVesterinen> mattsc: do you use a recent compiler?
20161013 16:24:29< mattsc> But it seems silly to add a key to a tag just so that a warning is not shown ...
20161013 16:24:47< mattsc> JyrkiVesterinen: yes, most up to date Xcode
20161013 16:25:18< JyrkiVesterinen> Then that means that Clang likely still doesn't have a -Wmaybe-uninitialized warning.
20161013 16:25:48< JyrkiVesterinen> Surprising to see something where GCC and MSVC are ahead of Clang.
20161013 16:25:53< celticminstrel> Those hiding warnings should probably be fixed too... though I think vultraz said that code is going to be removed soon anyway.
20161013 16:26:36< JyrkiVesterinen> tad_carlucci has fixes for the hiding warnings in PR825.
20161013 16:26:44< celticminstrel> Ah, nice.
20161013 16:27:30-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Read error: No route to host]
20161013 16:27:32< mattsc> celticminstrel: so I am not quite sure what to do.  I’m tempted to just note it on the wiki, and if somebody does not read the instructions and loses a lot of sleep because of that, well, too bad, not my problem.  :P
20161013 16:27:37-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 16:28:27< celticminstrel> I'm not sure either... an unsuppressable warning might not be too bad if it only shows in debug mode...
20161013 16:29:08< mattsc> Maybe.
20161013 16:33:21< gfgtdf> mattsc: a simple reason why people might want to use ais with shroud is that people use shared_shroud=yes on the ai, so that the players see the hexes that the ai has already discovered, (for allies ais)
20161013 16:36:30< irker663> wesnoth: GregoryLundberg wesnoth:master fea3a92e93a3 / src/server/ (server.cpp server.hpp): Suppress NO_FIFO_DIR message on WIN32 https://github.com/wesnoth/wesnoth/commit/fea3a92e93a38fe4b8a7d18003dfeed5de3afa47
20161013 16:36:32< irker663> wesnoth: Jyrki Vesterinen wesnoth:master 2d84467e1150 / src/server/ (server.cpp server.hpp): Merge pull request #823 from GregoryLundberg/GL_No_FIFODIR https://github.com/wesnoth/wesnoth/commit/2d84467e1150261e04c436cb2585971dd4cf83de
20161013 16:37:52-!- Soliton [~Soliton@wesnoth/developer/soliton] has left #wesnoth-dev ["Battle for Wesnoth: http://www.wesnoth.org/"]
20161013 16:37:52< mattsc> gfgtdf: yeah, that’s a good point.  So what’s your take on the warning?
20161013 16:38:19-!- Soliton [~Soliton@wesnoth/developer/soliton] has joined #wesnoth-dev
20161013 16:38:49< gfgtdf> mattsc: not sure which warning you mean, i just read this message "mattsc	But then, somebody _might_ want to use an AI side with an MAI and shroud.  I do not know why, but people have often surprised me …"
20161013 16:39:44< mattsc> gfgtdf: the MAIs don’t necessarily work when the AI side is under shroud (and fixing that is not a realistic option at this point).
20161013 16:40:13< mattsc> So I was thinking about displaying a warning message on the screen (in debug mode) when using an MAI for a shrouded side.
20161013 16:40:31< mattsc> Because if I just put this on the wiki, most people won’t see it.
20161013 16:41:13< mattsc> So the options, as I see them, are:
20161013 16:41:20< gfgtdf> mattsc: what do you mwan by 'not work' ? 1) they cash/error, 2) they fail to execute its given taks, 3) they executed its given task ina suboptimal/supid way.
20161013 16:41:38< mattsc> 2) and/or 3)
20161013 16:41:57< mattsc> 1) hopefully won’t happen
20161013 16:42:03< gfgtdf> mattsc: does thie effect only some or all microais?
20161013 16:43:01< mattsc> It affects not all, but potentially a large number of them.  Potentially because some might work with shroud under some, but not all circumstances.  The answer, as usual, is: it depends.
20161013 16:44:03< mattsc> So, the options I see are:
20161013 16:44:23< mattsc> a. Only put warning on wiki.  Make it the UMC creators responsibility to RTFM.
20161013 16:45:09< mattsc> b. Add warning in debug mode that will then always show, even if one decides that one wants to use an MAI with shroud.
20161013 16:45:36< mattsc> c. Do b. and also add a key to the [micro_ai] tag to allow suppressing the warning.
20161013 16:46:06< mattsc> None of those is ideal, they all have their own set of disadvantages.
20161013 16:46:24< mattsc> The ideal situation (fixing the behavior with shroud) is something I am not going to do.
20161013 16:47:29< gfgtdf> mattsc: hmm since there are sureley cases where you want to use an ai with shrtoudl liek i said above, i am against b.
20161013 16:47:55< gfgtdf> mattsc: i personalyl use debugmode quite rarley but i think thaere are poeple out there that use debug mode even for normal play.
20161013 16:48:07< mattsc> gfgtdf: yeah, makes sense
20161013 16:48:22< mattsc> Yes (on your second comment), I do usually.
20161013 16:48:43< mattsc> Although I rarely ever just do “normal play” any more these days
20161013 16:52:21-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20161013 16:52:25-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20161013 16:53:01-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161013 16:56:10-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 16:57:26-!- tad_carlucci [~lundberg@173.217.65.103] has joined #wesnoth-dev
20161013 16:58:03< mattsc> gfgtdf: I think I’ll go with option c. then.  The problem with a. is that the MAIs may simply silently fail (and that it is essentially impossible to determine from within the MAI whether they should fail or not) and that it is quite difficult for most people to figure out what is going on when an AI does not work as expected.
20161013 16:58:27< gfgtdf> mattsc: ok
20161013 17:02:17-!- mordante [~mordante@2001:984:5786:1:7a24:afff:fe8c:dea8] has joined #wesnoth-dev
20161013 17:02:17-!- mordante [~mordante@2001:984:5786:1:7a24:afff:fe8c:dea8] has quit [Changing host]
20161013 17:02:17-!- mordante [~mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20161013 17:02:33< mordante> servus
20161013 17:02:48< zookeeper> so how does the AI in general treat shroud?
20161013 17:04:03< mattsc> zookeeper: the default AI just ignores it.  That’s possible because the C++ code can distinguish between fog and shroud.
20161013 17:04:15< mattsc> The Lua path finding functions currently cannot.
20161013 17:04:44< mattsc> That could, in principle, be changed, but it seems like a lot of effort for no or not much gain.
20161013 17:05:12< zookeeper> ignores it, as in sees through it or pretends like those locations don't even exist?
20161013 17:05:21< mattsc> it sees through it
20161013 17:05:50< mordante> vultraz, you forgot your goat when you summoned me ;-)
20161013 17:06:01< zookeeper> okay, well in that case any scenario which uses a MAI with shroud already has the rest of the AI see through shroud, right?
20161013 17:06:18< mattsc> yep
20161013 17:07:41< zookeeper> wouldn't it then make sense to make MAI's see through shroud as well (i thought that was an option too)?
20161013 17:08:02< tad_carlucci> loonycyborg, Are you available?
20161013 17:08:30< mattsc> zookeeper: that would be the preferred option, but it is currently not possible (not without a lot of pain)
20161013 17:08:38< zookeeper> ah, okay
20161013 17:08:57< mattsc> I cannot have the MAIs ignore your hidden Avenger on a forest hex while seeing through shroud.
20161013 17:09:31< Soliton> how are the two related?
20161013 17:09:52< zookeeper> in that case, sure a warning seems appropriate, if the MAI is prone to behaving weirdly with shroud.
20161013 17:09:57< Soliton> i thought that's how the default ai works.
20161013 17:10:20< mattsc> Soliton: yes, the default AI does that.  It’s written in C++.  The MAIs are written in Lua.
20161013 17:10:44< Soliton> sure, but you make it sound like that is logically impossible or so.
20161013 17:10:50< loonycyborg> tad_carlucci: yes
20161013 17:11:02< mattsc> Soliton: No, that is absolutely not what I said.
20161013 17:11:14< mattsc> I specifically said that it is possible, but not worth the effort.
20161013 17:11:25< mattsc> Not worth *my* effort.
20161013 17:11:34< Soliton> well, that's how i interpreted "I cannot have the MAIs ignore your hidden Avenger on a forest hex while seeing through shroud."
20161013 17:11:37< mattsc> If somebody else wants to do it, more power to you.
20161013 17:12:01< tad_carlucci> loonycyborg, running scons strict clang  .. about savepng.c .. it crashes trying to run gcc instead of clang. 
20161013 17:12:29< mattsc> Soliton: also read my comment a few lines above that …
20161013 17:12:31< loonycyborg> is it even used for something other than cutter/exploder tools?
20161013 17:12:33< Soliton> i would think adding one more parameter to the path finding function would be that difficult but i'm not arguing that anyway.
20161013 17:12:37< loonycyborg> savepng.c that is
20161013 17:13:14< loonycyborg> tad_carlucci: how exactly did you set it up to use clang?
20161013 17:13:15< Soliton> mattsc: i did read it and explained how i understood it. apparently it's not how you meant it.
20161013 17:13:17< tad_carlucci> loonycyborg, Don't know but can't complete scons due to the crash. And savepng.c must be compiled using C not C++ 
20161013 17:13:45< tad_carlucci> loonycyborg, The command-line toolset option
20161013 17:13:54< tad_carlucci> loonycyborg, per the wiki
20161013 17:14:09< loonycyborg> hmm which wiki?
20161013 17:15:30< tad_carlucci> loonycyborg, "Compiling" subject where it says use the toolset option for clang, ccache if you have it, etc.
20161013 17:15:59< loonycyborg> hmm better list exact command line switch you've used
20161013 17:17:29< tad_carlucci> scons strict=true cxxtool=clang++ ccache=true
20161013 17:17:44< loonycyborg> I've added ctool option too
20161013 17:17:49< loonycyborg> ctool=clang
20161013 17:18:07< mattsc> Soliton: it’s not just adding a parameter to a couple functions.  You then have to go through all the custom AI code and make sure all the different instances of accessing locations/units/paths/reaches/… are dealt with correctly and consistently.
20161013 17:18:29< loonycyborg> tad_carlucci: in fact ctool=clang should be enough
20161013 17:18:39< mattsc> I am currently doing this to make sure fog and petrified units are taken into account correctly and it’s a lot of work.  And, frankly, I am not having fun doing it.
20161013 17:18:45< Soliton> mattsc: left as an exercise to... oh you. :-P
20161013 17:18:53< loonycyborg> because I've added some simple logic to figure out cxxtool if it's set to empty string
20161013 17:18:56< mattsc> Yep, exactly.  :)
20161013 17:19:25< loonycyborg> env['ctool'].rstrip("cc") + "++"
20161013 17:19:35< tad_carlucci> wesnoth->project->developer resources->CompilingWesnoth->4.1.1 Tips needs updating
20161013 17:19:44< celticminstrel> I don't think we really need to support people who use debug mode in normal play.
20161013 17:19:47< loonycyborg> it'll turn gcc into g++ and clang into clang++
20161013 17:19:54< celticminstrel> gfgtdf, mattsc: ^
20161013 17:20:17< celticminstrel> I also think that if the AI ignores shroud then that should be changed eventually.
20161013 17:20:32< loonycyborg> I'll update that wiki if I manage to find that article
20161013 17:21:42< tad_carlucci> loonycyborg, OK I am still doing "Good Morning Arch, what's changed overnight?" system maintenance and will retry the run in a bit. It'll take a long time because the change will 'make all' and the failure is a long way down.
20161013 17:22:05< Soliton> i thought the AI ignores fog but cannot see through shroud or hidden units.
20161013 17:22:06< loonycyborg> I know that feeling
20161013 17:22:32< Soliton> well, except for some special units like leaders.
20161013 17:24:32< tad_carlucci> loonycyborg, So https://wiki.wesnoth.org/CompilingWesnoth#Tips is where it says to use clang++ and needs updating 
20161013 17:24:45< loonycyborg> ok
20161013 17:26:50< loonycyborg> tad_carlucci: I updated it
20161013 17:26:58< tad_carlucci> tyvm
20161013 17:28:20< tad_carlucci> loonycyborg, I'm sure it'll work. I need to adjust some things for the PR before I run scons so I don't have to wade through it all again when I make the changes.
20161013 17:34:15-!- Nobun [~nobun@5.170.104.76] has quit [Quit: Salve a tutti]
20161013 17:34:20< tad_carlucci> loonycyborg, I'm sure it does not matter, but scons shows clang++ set .. is there a way to remove the default setting?
20161013 17:38:39< mattsc> To add to what I said before: adding a parameter to distinguish between fog, shroud and hidden units to wesnoth.find_path() and wesnoth.find_reach() would certainly be a good thing.
20161013 17:39:22< mattsc> Maybe in a couple months, when I have had some ditsance from the current MAI fixing work, I might even be convinced to add that functionality to the Lua AIs then.
20161013 17:41:12< Soliton> might be a candidate for the easy coding page?
20161013 17:42:57< mattsc> probably
20161013 17:44:12< gfgtdf> mattsc: so the problem with shroudl map is basicalyl that te ai has no 'i can't find the way so im going into shroud' mechanims?
20161013 17:44:46< mattsc> gfgtdf: no, that’s not it
20161013 17:45:26< mattsc> gfgtdf: Well, actually, yes, that is one problem.
20161013 17:45:45< mattsc> And I guess that’s why the default AI ignores shroud, yes.
20161013 17:45:58< mattsc> For the MAIs, there are different problems on top of that.
20161013 17:46:49< mattsc> For example, if it cannot find a way to the goto hex, it does not know whether that’s for a “legitimate reason” (such as enemies being in the way), or because it’s hidden under shroud.
20161013 17:47:16< mattsc> Well: there’s _currently_ no _easy_ way to do so.
20161013 17:49:37-!- Duthlet [~Duthlet@dslb-188-104-253-155.188.104.pools.vodafone-ip.de] has quit [Quit: leaving]
20161013 17:49:38< loonycyborg> tad_carlucci: don't understand what you mean. pass cxxtool= if you want that default logic to be used, then cxxtool won't appear in scons-option-cache etc
20161013 17:51:23< tad_carlucci> OK. Did that. I was noting it says deflaut " ... cxxtool='' ..." and was just wondering if there is a way to get back to looking like I'd never run scons before and had just done only what the wiki (now) says.
20161013 17:56:31< loonycyborg> passing cxxtool= will do that
20161013 17:56:39< loonycyborg> because default is empty string
20161013 18:06:12-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Ping timeout: 268 seconds]
20161013 18:06:14-!- celmin [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20161013 18:14:55< tad_carlucci> vultraz, gfgtdf On the Boost::Bind patch you've suggested alternate methods. My preference is to document the effected files by patching each but a central point means less future work. So we need to decide the approach.
20161013 18:16:10< gfgtdf> tad_carlucci: my worrys way mainly that other poeple who want to use boost bind in new files ten have to know this trick if we add it to eah file that uses it.
20161013 18:17:48< tad_carlucci> gfgtdf, I can see that. And using -include avoids changing the victims as I did. vultraz suggested using a commonly-included file .. which would do the same thing and not mean toolset changes, but risks being missed if that file is not included.
20161013 18:25:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161013 18:26:20< mattsc> gfgtdf, celmin, Soliton, zookeeper: I think for now I will settle for simply adding a comment on MAIs and shroud to the wiki page.  In a few weeks, when I am done with the current MAI work (and maybe a little less grumpy), I will look into how much effort it would be to make the MAIs behave in the same way as the default AI.
20161013 18:26:36< zookeeper> fine by me
20161013 18:27:47-!- iwaim__ [~iwaim@rasteenie.alib.jp] has quit [Ping timeout: 268 seconds]
20161013 18:28:39< tad_carlucci> mattsc, Not involved in the discussion but been following it and reading the issue on the forums. Sounds like a good plan to me.
20161013 18:28:56< mattsc> tad_carlucci, zookeeper: thanks
20161013 18:29:23< mattsc> celmin: as for your comment on “the AI should eventually not ignore shroud”, that’s likely not something I will tackle anytime soon.  However, it’s given me an idea for a new “explorer” Micro AI that could be added to some scenarios to achieve some sort of shroud clearing effect.
20161013 18:37:00< Soliton> mattsc: probably better than putting effort into a warning that one needs to be able to disable etc.
20161013 18:40:55-!- mordante [~mordante@wesnoth/developer/mordante] has quit [Quit: Leaving]
20161013 18:43:25-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161013 18:50:31-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Ping timeout: 258 seconds]
20161013 18:51:33-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161013 18:52:16< tad_carlucci> vultraz, calling the question: do you object to using -include to inject the boost patch as gfgtdf suggested?
20161013 18:59:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 18:59:53-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161013 19:00:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 19:09:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161013 19:19:25-!- JyrkiVesterinen [~JyrkiVest@87-100-193-57.bb.dnainternet.fi] has quit [Remote host closed the connection]
20161013 19:33:44-!- tomreyn [~tomreyn@archmage.megaglest.org] has quit [Read error: Connection reset by peer]
20161013 19:36:44-!- irker663 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20161013 19:37:55-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20161013 19:39:04< tad_carlucci> gfgtdf, Just out of curiousity. I am running scons testing using -include and seems fine, so far. Question: since this only effects clang, does CMake support clang and should I update that toolset as well?
20161013 19:45:14-!- iwaim__ [~iwaim@rasteenie.alib.jp] has joined #wesnoth-dev
20161013 19:54:59-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 19:57:10-!- mjs-de [~mjs-de@x4e319152.dyn.telefonica.de] has joined #wesnoth-dev
20161013 20:01:40-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20161013 20:14:57< gfgtdf> tad_carlucci: i think it supports clang, but i don't tink that this onyl effects clang, the rreport on boost https://svn.boost.org/trac/boost/ticket/12397 saidt that is effects also gcc 4.9
20161013 20:16:32< tad_carlucci> gfgtdf, Well I can't get CMake to stop using /usr/bin/c++ which is ccache using gcc .. I guess I'll put it in CMake untested other than to verify the option appears.
20161013 20:18:10< tad_carlucci>  /usr/local/bin/c++ that is
20161013 20:19:15-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 20:44:28-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161013 20:47:54-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 258 seconds]
20161013 20:47:54-!- wedge010 is now known as wedge009
20161013 21:10:20-!- mjs-de [~mjs-de@x4e319152.dyn.telefonica.de] has quit [Remote host closed the connection]
20161013 21:16:24-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161013 21:17:35-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161013 21:39:48< gfgtdf> tad_carlucci: i use neigher gcc nor Cmake so i cannot really help you sry.
20161013 21:41:39< tad_carlucci> gfgtdf, I'm ignoring it and simply checking the -include appears. Might break for not-GCC not-Clang builds but I'm out of my depth so we'll just have to wait for someone to complain.
20161013 21:44:27< Aginor> cmake supports clan
20161013 21:44:45< Aginor> I believe you can override it when you configure cmake by setting CC
20161013 21:44:51< Aginor> s/clan/clang/
20161013 21:49:57< tad_carlucci> Aginor, Tried that. No workie. Probably because I'm using ccache. It's always using /usr/local/bin/c++ instead of honoring my export C and CC settings per Cmake docs
20161013 21:50:09-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 252 seconds]
20161013 21:57:06-!- new_one_ is now known as new_one
20161013 21:57:24-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20161013 21:57:25-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Read error: Connection reset by peer]
20161013 21:59:01< shadowm> CC is the C compiler, CXX is the C++ compiler.
20161013 22:00:09< Aginor> shadowm raises an important point
20161013 22:00:15< Aginor> set both :)
20161013 22:04:59< tad_carlucci> shadowm, Aginor both are required. cmake ignores them,
20161013 22:05:44< shadowm> tad_carlucci: You said you exported C and CC. C doesn't mean anything, CC is the C compiler, and CXX is the C++ compiler.
20161013 22:06:06< tad_carlucci> shadowm, I did CC and CXX and mistyped.
20161013 22:06:53< tad_carlucci> shadowm, Not that it matters. cmake ignored the environment settings and used /usr/local/bin/c++ anyway
20161013 22:07:00< shadowm> Next there's the possibility of using the CMake variables CMAKE_CXX_COMPILER and CMAKE_C_COMPILER with the -D switch (e.g. cmake -DCMAKE_CXX_COMPILER=clang++).
20161013 22:07:49< shadowm> I'm fairly sure this has worked for me in the past before I was forced to ditch clang.
20161013 22:07:50< tad_carlucci> shadowm, I can try that. Runing cmake now to test the change to inject the boost patch everywhere using GCC
20161013 22:09:20-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20161013 22:14:30< vultraz> tad_carlucci: -include is even better
20161013 22:16:40< tad_carlucci> vultraz, Good because that's what I'm doing. BTW, hitting a few gcc messages, all noise, will add changes to the PR for review but all look safe to me.
20161013 22:19:10-!- atarocch [~atarocch@93.56.160.28] has quit [Remote host closed the connection]
20161013 22:20:23< tad_carlucci> src/addon/manager_ui.cpp:104:27: error: no match for ‘operator+’ (operand types are ‘const string {aka const std::__cxx11::basic_string<char>}’ and ‘const bool’)
20161013 22:22:27< shadowm> vultraz: Did you test the divider line coloring I suggested?
20161013 22:23:02< tad_carlucci> Oh .. I see .. missing () .. easy fix, my bad
20161013 22:24:02< vultraz> shadowm: yes, I'm pondering it
20161013 22:24:18< vultraz> im afraid it might make things *too* colorful 
20161013 22:24:59< vultraz> the lines are supposed to be unobtrusive
20161013 22:35:12-!- aidanhs [~aidanhs@81.4.110.234] has quit [Ping timeout: 260 seconds]
20161013 22:35:53< gfgtdf> Soliton: about why 'leave_game' took 10 seconds, i think this could be becasue at the end of the game (when the last palyer leaves) the game +replay is written to disk, i dont know exactly how gic theat file cn be can take long but i think this coudl be an expliation.
20161013 22:38:15< gfgtdf> Soliton: we coudl try to handle this io asyncroniously with another boost::asio::io_server object.
20161013 22:43:57< shadowm> Writing the smallest amount of data to disk tends to take forever when a backup job is running.
20161013 22:44:01-!- aidanhs [~aidanhs@81.4.110.234] has joined #wesnoth-dev
20161013 22:44:34< shadowm> vultraz: Using shades of gray seems far more obstrusive when they are normally used to call attention to stuff the user must read (text).
20161013 22:44:46< shadowm> Here, that is.
20161013 22:46:11< shadowm> It took me that long to realize why slider grooves were bothering me -- it was because they were just as bright as text.
20161013 22:46:39< shadowm> Using the same color further reinforces the notion that two elements somehow belong to the same category.
20161013 22:47:54< tad_carlucci> Oh .. 'slider' .. reminds me: Did anyone take a look at why the zoom slider isn't showing correctly for the minimap?
20161013 22:48:47< gfgtdf> tad_carlucci: maybe you are playing on a small resolution ?
20161013 22:49:19< tad_carlucci> Full screen. 1400x or so.
20161013 22:49:29< gfgtdf> tad_carlucci: hmm  ok
20161013 22:49:52< gfgtdf> tad_carlucci: i asked becsue the ingae theme layout is different on small resolution which also effects the minimap.
20161013 22:50:00< gfgtdf> tad_carlucci: did you file a bugrepotfor the slider issue?
20161013 22:50:54< tad_carlucci> I'll fire it up to check once I'm done clearing the gcc noise from this build. Don't use the tracker. Far too many orphaned web accounts as it is.
20161013 22:52:29< vultraz> zoom slider's been gone for a long time now
20161013 22:52:40< vultraz> dont know why it vanished, but i was going to remove it anyway
20161013 22:53:23< tad_carlucci> Well, I guess it can wait until after Sunday.
20161013 23:06:27< vultraz> ok, we really need to implement auto-hiding scrollbars after n time
20161013 23:14:45-!- matthiaskrgr [matthiaskr@gateway/shell/panicbnc/x-uskkcvdqshamysze] has quit [Ping timeout: 260 seconds]
20161013 23:53:43< tad_carlucci> PR 825 updated PR 820 needed (for me) to get a clean compile using SCons and Cmake and clang and gcc. Unless there are more comments, I and done with them both.
--- Log closed Fri Oct 14 00:00:22 2016