--- Log opened Wed Aug 24 00:00:45 2016
20160824 00:02:06-!- travis-ci [~travis-ci@ec2-54-92-146-1.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 00:02:07< travis-ci> wesnoth/wesnoth#10526 (master - 44219de : Celtic Minstrel): The build failed.
20160824 00:02:07< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154607253
20160824 00:02:07-!- travis-ci [~travis-ci@ec2-54-92-146-1.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 00:02:25< mattsc> hmm, there’s even a warning on the wiki page that that might interfere with AI code ...
20160824 00:02:35< celmin> Really?
20160824 00:02:37< celmin> Huh.
20160824 00:02:58< celmin> I'm not sure why it would've before all my work on it though.
20160824 00:09:55< mattsc> If I were more motivated, I’d look into who made that comment and when, but I think I am too lazy for that.
20160824 00:10:36< mattsc> As for the right-click-option / hotkey double execution, that only happens when I use the hotkey, not from the right-click menu
20160824 00:10:59< mattsc> I’ll look into that some more later …
20160824 00:16:25< celmin> What happens if you try to assign a unit to a side that doesn't exist?
20160824 00:17:21< mattsc> is that a general question, or one for me?
20160824 00:18:08< celmin> General.
20160824 00:20:00< mattsc> “ error engine: wrong side in [unit] tag - no such side: 3 ( number of teams :2) “
20160824 00:20:20< celmin> And then the unit doesn't appear?
20160824 00:20:24< mattsc> correct
20160824 00:21:28< celmin> I noticed that team::get_side_color_index(i) just returns teams()[i].color(), which renders it redundant in certain places where the code already has a team object (one place even does team::get_side_color_index(t.side()) which is the same as t.color()). I think this might fix an editor crash when trying to add a new side.
20160824 00:24:38-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160824 00:34:23< celmin> …okay, there is one difference between them, but I kinda hope it doesn't matter...
20160824 00:54:44< celmin> I just noticed that TC does seem to be working in my build, but only on flags.
20160824 00:55:01< celmin> Ellipses, units, orbs, and terrain icons are still magenta.
20160824 01:01:00-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has joined #wesnoth-dev
20160824 01:03:21< celmin> RatArmy: I heard female_message isn't working properly?
20160824 01:03:37< RatArmy> yes
20160824 01:03:40< RatArmy> male is same
20160824 01:04:44< celmin> Is that on latest master?
20160824 01:04:48< celmin> Or in 1.13.5?
20160824 01:04:59< RatArmy> latest
20160824 01:06:28< celmin> I thought that was recently fixed, in https://github.com/wesnoth/wesnoth/commit/9d44a45f88e4a1044fbb02ea3ef3b572da5a70ca
20160824 01:08:02< RatArmy> oh.. I will pull and rebuild
20160824 01:08:29< celmin> It's a Lua fix, so rebuild not required.
20160824 01:08:39-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 01:13:16-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 265 seconds]
20160824 01:13:19< RatArmy> sorry, it works!
20160824 01:13:29< RatArmy> thx to fix
20160824 01:20:06-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160824 01:20:14-!- vultraz [~chatzilla@124.109.10.167] has quit [Read error: Connection reset by peer]
20160824 01:20:29-!- VultCave is now known as vultraz
20160824 01:33:04-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has quit [Ping timeout: 250 seconds]
20160824 01:39:08-!- un214 [~un214@104.220.56.173] has joined #wesnoth-dev
20160824 01:39:18-!- louis94 [~~louis94@91.178.241.1] has quit [Ping timeout: 276 seconds]
20160824 01:50:29< vultraz> interesting
20160824 01:50:37< vultraz> he fixed the crash quite easily
20160824 01:56:53< celmin> ?
20160824 01:57:01< celmin> Oh, that one.
20160824 01:57:19< celmin> vultraz: So, what should we do about PR 744?
20160824 01:57:39< irker566> wesnoth: Celtic Minstrel wesnoth:master d1604e56a192 / src/ (display.cpp gui/dialogs/unit_recruit.cpp reports.cpp team.cpp): Fix editor crash when adding a new side https://github.com/wesnoth/wesnoth/commit/d1604e56a19273c6f27c54c4931ccde26442de88
20160824 01:57:48< vultraz> merge, of course
20160824 01:59:45-!- Shiki [~Shiki@141.39.226.227] has quit [Remote host closed the connection]
20160824 02:00:29< vultraz> shadowm: so it looks jyrki fixed the crash I was having much easier than expected by utilizing the handled/halt parameters (https://github.com/wesnoth/wesnoth/commit/af2303df778225cb837461cc9943470835d01e1f) I dunno if it will be useful in the filechooser dialog, but worth a shot.
20160824 02:05:29< vultraz> when you have a chance let me know if that works
20160824 02:06:31< celmin> vultraz: Merge now or more testing first?
20160824 02:07:03< vultraz> the only way such a thing can be sufficiently tested is during normal use load
20160824 02:07:21< celmin> 'kay, will do.
20160824 02:08:04< vultraz> (perhaps not the "only way", but it's the most practical for us right now)
20160824 02:08:10< vultraz> (and you've fixed the crash it had)
20160824 02:11:55< celmin> Well, it wasn't a crash, but yeah.
20160824 02:27:49-!- un214 [~un214@104.220.56.173] has quit [Remote host closed the connection]
20160824 02:34:58-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 02:34:59< travis-ci> wesnoth/wesnoth#10527 (master - d1604e5 : Celtic Minstrel): The build is still failing.
20160824 02:34:59< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154629341
20160824 02:34:59-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 02:35:12-!- oldlaptop [~quassel@162.247.150.37] has quit [Ping timeout: 276 seconds]
20160824 02:35:31-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160824 02:37:14< tad_> celmin: say your changes for filter_recall and am compiling them now. I'm anxious to see how clean it makes my proof-of-concept changes. If it works there's a LOT of store/unstore work which can be avoided and by avoiding makes choose_level and other things work SO much better. 
20160824 02:37:32< celmin> Okay.
20160824 02:38:02< tad_> Got a set-but-not-used warning in /home/lundberg/wesnoth/src/scripting/lua_common.cpp:758:59: 
20160824 02:39:11< tad_> And my thanks to whoever told me about ccache .. makes it much faster to make the engine.
20160824 02:43:37< celmin> L758 is a function definition...
20160824 02:46:00< tad_> Let me find the commit and add a line comment ...
20160824 02:46:20< tad_> Oh and your changes look great.
20160824 02:46:23< celmin> I'll check GH notifications then, I guess.
20160824 02:48:19-!- oldlaptop [~quassel@162.247.150.37] has joined #wesnoth-dev
20160824 02:49:18< tad_> https://github.com/wesnoth/wesnoth/commit/1e575480a409a3b377bd85052a6aaef741e861fd
20160824 02:50:18< celmin> What a useless warning.
20160824 02:50:25< tad_> Yep.
20160824 02:50:56< celmin> The whole point of the function is to set that variable.
20160824 02:51:16< celmin> Though, I'm not sure why I made it a pointer instead of a reference...
20160824 02:51:37< tad_> That may be the point if the warning, your know.
20160824 02:51:53< celmin> Yeah, okay, there actually is a problem there.
20160824 02:51:59< celmin> It should've been *vcfg = ...
20160824 02:53:28< celmin> I'm just trying to decide whether or not to change it to a reference parameter, and if so, whether to use vconfig& or vconfig*&
20160824 02:54:08< tad_> So, what I'm doing in TSG for the proof of concept is running through the entire recall list to clear a [unit][variables] (I should optimize that) at the start of the scenario then running though all units on the map at the end to set it. Is there a faster way to [foreach] all units on the map other than [store_unit] all of them?
20160824 02:54:38< tad_> *&? A reference to a pointer? You sure? Seems 'odd'
20160824 02:54:47< celmin> It is indeed odd.
20160824 02:54:56< celmin> I've done it before somewhere.
20160824 02:55:00< celmin> Decided not to do it here though.
20160824 02:55:37< tad_> I mean, I can see doing it, but it is rather unusual and seems 'smelly' but I guess it have valid uses.
20160824 02:56:04< celmin> Hmm...
20160824 02:56:46< celmin> Based on the use sites, *& actually makes more sense, but maybe the use sites could be changed...
20160824 02:56:57-!- oldlaptop [~quassel@162.247.150.37] has quit [Ping timeout: 258 seconds]
20160824 02:57:06< celmin> For example, game_lua_kernel.cpp:2517
20160824 02:57:29< tad_> 'He has a pointer to the object I want to change and passed me the pointer." Seems a reference might make more sense in virtually evrey case
20160824 02:57:59< celmin> That's not exactly what's going on here though.
20160824 02:58:16< celmin> It's "I want to get back a config, but if it was originally a vconfig, then I want that too."
20160824 02:58:47< tad_> Maybe 'const* &'?
20160824 02:59:19< tad_> Like I said, a reference to a pointer seems 'smelly' ..
20160824 02:59:44< celmin> This situation was already kind of weird even before this commit.
20160824 03:05:57-!- oldlaptop [~quassel@162.247.150.37] has joined #wesnoth-dev
20160824 03:11:24< tad_> Oh, I'm am really liking this. Keeping all the units in the recall list is SO much cleaner!
20160824 03:11:52< tad_> No more checking for variables and checking inside them for units to clean up. Just clean up and that's that.
20160824 03:12:51< tad_> And all those bugs about units recalling poisoned or unable to move or whatever .. POOF! .. no more bug
20160824 03:13:10-!- oldlaptop [~quassel@162.247.150.37] has quit [Ping timeout: 250 seconds]
20160824 03:17:23< tad_> What was the latest word about "'~RC(magenta>1)' not found" errors. I seem to have done something which is causing them and wonder what I should do to fix it.
20160824 03:17:42< tad_> I think it's related to changing leaders.
20160824 03:17:55< celmin> I'm not really sure what causes that...
20160824 03:19:02< tad_> Do ellipse or overlays (hero and loyal) use it? It could try resetting the ellipse.
20160824 03:23:14< celmin> Units, ellipses, flags, orbs, terrain icons. I don't think it could be caused by the latter two though since the target colour isn't based on a side.
20160824 03:23:23< celmin> Overlays do not use it.
20160824 03:23:33< celmin> (Unless specifically requested, of course.)
20160824 03:25:11< tad_> OK. Well what's changing is the unit and that implies ellipse. The side flag is constant. No icons on the map. So it's either the ellipse or the unit or its orb. I'll try to localize it next time I'm in that area.
20160824 03:25:33< celmin> Probably not its orb, like I said.
20160824 03:26:04< celmin> The target colours for orbs are set by preferences, not based on the side number, but in certain cases the side number is returned when requesting a side's colour.
20160824 03:26:15-!- oldlaptop [~quassel@162.247.150.37] has joined #wesnoth-dev
20160824 03:26:47< celmin> I changed something recently, not sure if that would affect it (though if it did it would probably be to drop the 1 from the message).
20160824 03:28:20< tad_> I've seen that, too. But right now its >1 and fairly repeatable .. and related to something I changes because when I stash and re-test it goes away
20160824 03:29:16< celmin> I assume side 1 has a color set in its config?
20160824 03:29:40< celmin> I'm not sure what the game does if you don't set a colour.
20160824 03:29:51< celmin> I think it probably chooses a default though.
20160824 03:30:22< celmin> On the other hand if you set the colour to empty via Lua or [modify_side], it might just set it to empty, which might be a way to get this error (not quite sure).
20160824 03:33:25< tad_> Hah. I think it's this. "Move a unit to the target, it's moves are now 0. During the moveto event, remove the unit from the map." I just added, "And put it back on the map after you made your changes.
20160824 03:33:26-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160824 03:33:49< celmin> ?
20160824 03:33:56< tad_>  and the message went away. So the orb wants to change from team-color to red and there's no orb to change .. I **think**
20160824 03:34:12< celmin> You mean the ellipse?
20160824 03:34:35< tad_> The little moves-left color indicator over the leader-band.
20160824 03:34:58< celmin> Eh? Is it really caused by the orb?
20160824 03:35:14< celmin> Why do you think it's the orb and not the ellipse or unit?
20160824 03:35:37< tad_> Dunno. But if I leave the unit on the recall list instead of the map I get the error.
20160824 03:35:48< tad_> It's a leader, btw.
20160824 03:37:02< celmin> So, moving a unit to the recall list in a moveto event appears to trigger it?
20160824 03:37:38< tad_> I think any removal from the map (like leaving in a variable). I can retest. Gimme a couple minutes.
20160824 03:38:06< celmin> Leaving it in a variable sounds even less likely… since a unit in a variable isn't actually a unit from the engine's viewpoint (it's just a config).
20160824 03:38:51-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160824 03:39:27< tad_> First test: I simply left it in the variable and got the error.
20160824 03:40:07-!- JyrkiVesterinen [~JyrkiVest@87-92-32-74.bb.dnainternet.fi] has joined #wesnoth-dev
20160824 03:40:44-!- vultraz [~chatzilla@124.109.10.167] has quit [Ping timeout: 250 seconds]
20160824 03:40:58-!- VultCave is now known as vultraz
20160824 03:41:12-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20160824 03:41:12-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160824 03:41:13< tad_> Second test: unstored to recall list and I get the error
20160824 03:42:44< celmin> Weird.
20160824 03:42:56< celmin> But it sounds like the [store_unit] is the cause then...
20160824 03:43:04< tad_> Third test .. now I'm confused. It did NOT occur last time, but now, when I unstore it to the map I **DO** get the error
20160824 03:43:26< tad_> As I said, this error is beating me up. I cannot get it to reliably repeat ...
20160824 03:44:43< tad_> It sure seems to have something to do with storing/unstoring the unit during a moveto event but MAN do I wish it would be consistently repeatable.
20160824 03:46:42-!- celmin [~celticmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The minstrel departs, to spread the music to the masses!]
20160824 03:48:26-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20160824 03:48:37-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160824 03:53:51-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 276 seconds]
20160824 03:55:14-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160824 04:04:22-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160824 04:04:52-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20160824 04:09:27-!- VultCave [~chatzilla@124.109.10.167] has quit [Ping timeout: 276 seconds]
20160824 04:13:45-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160824 04:14:58-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160824 04:29:31-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 04:34:54-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 265 seconds]
20160824 04:37:37-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has joined #wesnoth-dev
20160824 04:43:25-!- Kwandulin [~Miranda@p200300760F35BF8DC0763A3E0109D5F8.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160824 04:44:43-!- ancestral [~ancestral@67-4-224-82.mpls.qwest.net] has quit [Ping timeout: 244 seconds]
20160824 04:55:07-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160824 04:57:59-!- irker566 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160824 04:59:35-!- JyrkiVesterinen [~JyrkiVest@87-92-32-74.bb.dnainternet.fi] has quit [Quit: .]
20160824 06:07:33-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160824 06:07:49-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Client Quit]
20160824 06:10:42-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160824 06:19:25-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 06:23:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 258 seconds]
20160824 06:28:02-!- irker973 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160824 06:28:02< irker973> wesnoth: Celtic Minstrel wesnoth:master f6974f01f694 / src/scripting/ (game_lua_kernel.cpp lua_common.cpp lua_common.hpp): fixup! When creating units in Lua from a vconfig, don't parse any contained [fil https://github.com/wesnoth/wesnoth/commit/f6974f01f694c5d72f2be63c3f2dfac4a504e6b1
20160824 06:28:04< irker973> wesnoth: Celtic Minstrel wesnoth:master 8699ad0a0514 / / (4 files in 3 dirs): Use GUI2 for game menus https://github.com/wesnoth/wesnoth/commit/8699ad0a05144d2971a8b8d3cfd585349b5db255
20160824 06:36:39-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160824 06:36:50-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160824 06:40:54-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160824 06:48:25< irker973> wesnoth: Charles Dang wesnoth:master fbfabfba50b8 / src/ (5 files in 3 dirs): Mp Lobby: set up plugins context https://github.com/wesnoth/wesnoth/commit/fbfabfba50b8b4dd38354183b525ec1a5b7b42d3
20160824 06:48:28< irker973> wesnoth: Charles Dang wesnoth:master 88d0619d5806 / src/help/help_topic_generators.cpp: Cleaned up another use of snprintf https://github.com/wesnoth/wesnoth/commit/88d0619d5806c14e70d919221bd73494e51fd0d7
20160824 06:48:57< celticminstrel> Ugh, your timing is the worst. I was just about to push the lua_refactor merge.
20160824 06:49:04< vultraz> oh deer :|
20160824 06:49:15< vultraz> can't you just press the merge button
20160824 06:49:20< celticminstrel> Well, I can rebase with preserve-merges, so it's just a bit more of a pain than usual.
20160824 06:49:39< celticminstrel> It has conflicts, so I elected to merge manually rather than rebasing the branch.
20160824 06:49:44< vultraz> ah
20160824 06:49:55< celticminstrel> So what do you think of the new menus?
20160824 06:50:20< vultraz> new menus?
20160824 06:50:25< celticminstrel> New menus.
20160824 06:50:29< vultraz> what new menus
20160824 06:51:00 * vultraz checks logs
20160824 06:51:05< celticminstrel> The ones you pulled just now that I pushed about ten minutes ago.
20160824 06:51:38< vultraz> ooo
20160824 06:52:09< vultraz> why... did you add another label
20160824 06:52:14< celticminstrel> For hotkeys.
20160824 06:52:24< vultraz> oh
20160824 06:52:29< vultraz> right
20160824 06:52:40< vultraz> let me build and take a look
20160824 06:53:17< celticminstrel> Some of the editor menus were especially troublesome.
20160824 06:53:57< vultraz> the selection problem is due to them being a listbox
20160824 06:54:10< celticminstrel> Even so, it's probably solvable, technically.
20160824 06:54:21< vultraz> I suppose we could try that untested 'minimum selection of 0' feature
20160824 06:54:50< celticminstrel> We could.
20160824 06:55:15< celticminstrel> I was thinking just hacking the toggle-panel into a deselected state.
20160824 06:55:23< celticminstrel> But something not hacky is probably better.
20160824 06:56:09< vultraz> that'd probably break the listbox
20160824 06:56:32< celticminstrel> I don't see why it would, but who knows.
20160824 06:59:51< vultraz> hm
20160824 07:00:00< vultraz> why do you need to create an sdl_rect here?
20160824 07:00:25< celticminstrel> The drop down list was designed to be triggered by a button.
20160824 07:00:35< celticminstrel> So, it takes the rect of that button as a parameter.
20160824 07:00:49< celticminstrel> That's how it can position itself above or below the button, as space allows.
20160824 07:01:03< vultraz> hmmm
20160824 07:01:09< celticminstrel> Unfortunately, we don't know the width of the button here (or even if there is one).
20160824 07:01:11< vultraz> somehow it looks rather odd without padding
20160824 07:01:33< vultraz> I designed that window to have no padding at all
20160824 07:02:11< celticminstrel> BTW, there's also a 800x600 MP Create layout, if you missed that.
20160824 07:02:12-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has joined #wesnoth-dev
20160824 07:02:20< vultraz> but it's a good start
20160824 07:02:21< vultraz> good work :)
20160824 07:02:30< celticminstrel> Though it's a bit glitchy. I discovered it boots to menu if you click Create when on the settings tab.
20160824 07:02:46-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has quit [Client Quit]
20160824 07:02:46< vultraz> what's this 'fugly markup' you speak of?
20160824 07:03:19< celticminstrel> Menus used to use &x=x markup for images and columns and such.
20160824 07:03:46< vultraz> and....
20160824 07:03:49< celticminstrel> This is hardcoded into some of the generated items in the editor (groups and times of day), and I didn't see an easy way to get around it.
20160824 07:03:58< vultraz> hardcoded :|
20160824 07:04:00< vultraz> well then
20160824 07:04:05< vultraz> we'll have to un-hardcode it
20160824 07:04:17< celticminstrel> Well, hardcoded might be a slight exaggeration.
20160824 07:04:38< vultraz> huh you managed a layout
20160824 07:04:40< vultraz> I'm impressed
20160824 07:05:06< celticminstrel> The key point is that a menu is a list of strings, so it's pretty hard to encode anything complicated.
20160824 07:05:26< vultraz> but it's not a list of strings anymore :/
20160824 07:05:29< celticminstrel> And it probably should stay a list of strings, anyway - these are item IDs it's working with.
20160824 07:05:41-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 07:05:42< travis-ci> wesnoth/wesnoth#10528 (master - 8699ad0 : Celtic Minstrel): The build is still failing.
20160824 07:05:42< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154660677
20160824 07:05:42-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 07:05:53< celticminstrel> No, I'm not talking about the drop down list, I'm talking about command_executor::show_menu.
20160824 07:06:19< celticminstrel> I think the layout actually doesn't quite fit in 800x600 due to the settings tab though. Still needs some work.
20160824 07:13:06< vultraz> for some reason I can't understand the wml here..
20160824 07:15:21< vultraz> also don't see thehorizontalspacer line :/
20160824 07:16:15< vultraz> oh
20160824 07:16:17< vultraz> duh
20160824 07:16:22< vultraz> I'm looking at the wrong file :|
20160824 07:19:03< irker973> wesnoth: Celtic Minstrel wesnoth:master 797613f760d0 / src/scripting/ (lua_api.cpp lua_kernel_base.cpp): Refactor some duplicate code in Lua API https://github.com/wesnoth/wesnoth/commit/797613f760d0c84afefa39b305bb512f4a574479
20160824 07:19:05< irker973> wesnoth: Celtic Minstrel wesnoth:master 1521f326df7a / / (17 files in 4 dirs): Move Lua unit wrapper to own file https://github.com/wesnoth/wesnoth/commit/1521f326df7a04feae7b1f7f56ffbce736a58c02
20160824 07:19:07< irker973> wesnoth: Celtic Minstrel wesnoth:master 4dd99b881f3d / src/scripting/ (8 files): Implement Lua placement new https://github.com/wesnoth/wesnoth/commit/4dd99b881f3de9a31af86214eb9d1d23a675c594
20160824 07:19:09< irker973> wesnoth: Celtic Minstrel wesnoth:master 6d38c77aab3e / src/ (5 files in 2 dirs): Extract Lua unit push code into a function https://github.com/wesnoth/wesnoth/commit/6d38c77aab3ed3e52e158b548e2e1a826c7ce8be
20160824 07:19:11< irker973> wesnoth: Celtic Minstrel wesnoth:master 3804027d100f / src/scripting/ (game_lua_kernel.cpp game_lua_kernel.hpp lua_unit.cpp lua_unit.hpp): Avoid luaW_hasmetatable when checking for units https://github.com/wesnoth/wesnoth/commit/3804027d100fc02328a6829997792b8ef0587303
20160824 07:19:13< irker973> wesnoth: Celtic Minstrel wesnoth:master 4dcba0c3cdec / src/scripting/game_lua_kernel.cpp: Fix bug in wesnoth.put_unit https://github.com/wesnoth/wesnoth/commit/4dcba0c3cdec6cb535206f8c23f0beb866db737f
20160824 07:19:15< irker973> wesnoth: Celtic Minstrel wesnoth:master fa0e63e3a346 / src/scripting/ (game_lua_kernel.cpp lua_unit.cpp lua_unit.hpp): Move unit metatable stuff into lua_unit file https://github.com/wesnoth/wesnoth/commit/fa0e63e3a3462dd3dabcdff5c2aecbc1fa213dbc
20160824 07:19:17< irker973> wesnoth: Celtic Minstrel wesnoth:master 22a39d0bb4d9 / / (6 files in 2 dirs): Move Lua unit attacks metatable to own file https://github.com/wesnoth/wesnoth/commit/22a39d0bb4d9f1b756cfdc43e7ff63565b580c04
20160824 07:19:19< irker973> wesnoth: Celtic Minstrel wesnoth:master 2dd19b7d15a5 / src/scripting/lua_unit_attacks.cpp: Fix Lua attacks metatable not working for attacks obtained from unit_type https://github.com/wesnoth/wesnoth/commit/2dd19b7d15a5f2997907c0b48649b28d0910b46c
20160824 07:19:21< irker973> wesnoth: Celtic Minstrel wesnoth:master bb43fa29c46b / src/scripting/game_lua_kernel.cpp: Use luaW_pushunit when getting lists of units https://github.com/wesnoth/wesnoth/commit/bb43fa29c46b41e52c3169f72a2dcaad94618778
20160824 07:19:23< irker973> wesnoth: Celtic Minstrel wesnoth:master 30b7ab8b2dc7 / / (25 files in 9 dirs): Reduce use of const_cast https://github.com/wesnoth/wesnoth/commit/30b7ab8b2dc7a3a7d19372216307255327ce5b19
20160824 07:19:25< irker973> wesnoth: Celtic Minstrel wesnoth:master 39477b40a727 / src/ (6 files in 2 dirs): Fix Lua metatables by using lauxlib to manage them https://github.com/wesnoth/wesnoth/commit/39477b40a727ed4aa4d717475331553e7176f7e1
20160824 07:19:27< irker973> wesnoth: Celtic Minstrel wesnoth:master e821e508c3b9 / / (37 files in 10 dirs): Merge pull request #744 from lua_refactor https://github.com/wesnoth/wesnoth/commit/e821e508c3b9504564b769bda649433833fa13ae
20160824 07:24:38< celticminstrel> Updating some branches now...
20160824 07:28:06< celticminstrel> Or maybe a bit later...
20160824 07:29:01-!- boucman_work [~boucman@229.29.205.77.rev.sfr.net] has joined #wesnoth-dev
20160824 07:29:32-!- Kwandulin [~Miranda@p200300760F35BF8DC0763A3E0109D5F8.dip0.t-ipconnect.de] has quit [Ping timeout: 250 seconds]
20160824 07:30:11-!- travis-ci [~travis-ci@ec2-54-92-146-1.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 07:30:12< travis-ci> wesnoth/wesnoth#10529 (master - 88d0619 : Charles Dang): The build is still failing.
20160824 07:30:12< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154663534
20160824 07:30:12-!- travis-ci [~travis-ci@ec2-54-92-146-1.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 07:32:55-!- celticminstrel is now known as celmin|sleep
20160824 07:33:04< vultraz> whyyy is there this random space here
20160824 07:33:15< celmin|sleep> ?
20160824 07:33:18< celmin|sleep> Where?
20160824 07:34:01< vultraz> in the tiny res
20160824 07:34:08< celmin|sleep> I assumed that much.
20160824 07:34:18< vultraz> above the stuff in the general tab
20160824 07:34:36< celmin|sleep> Like, above the game name field?
20160824 07:35:04< vultraz> yes
20160824 07:35:07< vultraz> finally, fixed >_>
20160824 07:35:16< celmin|sleep> I think the settings tab is actually too tall; I was thinking of putting the two slider banks side-by-side instead. Not sure if that's related though.
20160824 07:35:31< celmin|sleep> Also, yay fix.
20160824 07:43:28-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 07:43:29< travis-ci> wesnoth/wesnoth#10530 (master - e821e50 : Celtic Minstrel): The build is still failing.
20160824 07:43:29< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154669226
20160824 07:43:29-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 07:43:59< irker973> wesnoth: Wedge009 wesnoth:master 9b9873440087 / projectfiles/VC12/ (wesnoth.vcxproj wesnoth.vcxproj.filters): Updating VC project files for merge of PR #744. https://github.com/wesnoth/wesnoth/commit/9b987344008780fe9327e43eaed7b931c26b9fed
20160824 07:54:01< irker973> wesnoth: Charles Dang wesnoth:master deaae3f382df / data/gui/window/mp_create_game/ (small_resolution.cfg tiny_resolution.cfg): MP Create: layout improvements for tiny version https://github.com/wesnoth/wesnoth/commit/deaae3f382dff97b8aa14c813e991034eefffdcc
20160824 07:54:03< vultraz> celmin|sleep: ^
20160824 08:00:55-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160824 08:03:05-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 08:03:06< travis-ci> wesnoth/wesnoth#10531 (master - 9b98734 : Wedge009): The build is still failing.
20160824 08:03:06< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154673387
20160824 08:03:06-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 08:04:34-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160824 08:07:39-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 08:12:10-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 252 seconds]
20160824 08:12:12-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 08:12:13< travis-ci> wesnoth/wesnoth#10532 (master - deaae3f : Charles Dang): The build is still failing.
20160824 08:12:13< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154675227
20160824 08:12:13-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 08:27:42-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160824 08:45:13-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160824 09:01:59-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20160824 09:06:25< irker973> wesnoth: Charles Dang wesnoth:master 7d7d3f9508bc / data/gui/window/drop_down_list.cfg: Dropdown List: Few layout tweaks https://github.com/wesnoth/wesnoth/commit/7d7d3f9508bca48634757fe2e2bcbeca9fae9b1c
20160824 09:08:53< vultraz> celmin|sleep: ^ added padding on the right to make the borders consistent
20160824 09:08:54< vultraz> looks fine now
20160824 09:10:51< vultraz> i'll work at de-harcoding the offensive formatting later
20160824 09:23:28-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 09:23:29< travis-ci> wesnoth/wesnoth#10533 (master - 7d7d3f9 : Charles Dang): The build is still failing.
20160824 09:23:29< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154689366
20160824 09:23:29-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 09:25:03-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160824 09:39:54-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20160824 09:53:57-!- Kwandulin [~Miranda@p200300760F35BFF844C4C972FF268998.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160824 09:55:56-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 10:00:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160824 10:39:05-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160824 10:58:53< Kwandulin> Has there been talk about a [chapter] tag for [campaign]?
20160824 10:59:12< Kwandulin> (https://forums.wesnoth.org/viewtopic.php?p=599351#p599351)
20160824 10:59:52< vultraz> not that I've heard
20160824 11:40:07-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160824 11:44:14-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 11:44:27-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 276 seconds]
20160824 11:44:27-!- wedge010 is now known as wedge009
20160824 11:49:00-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160824 12:01:26-!- yaiyan_ is now known as Yaiyan
20160824 12:06:47-!- irker973 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160824 12:19:20-!- Kwandulin [~Miranda@p200300760F35BFF844C4C972FF268998.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160824 12:36:21-!- Kwandulin [~Miranda@p200300760F35BFF834A3A5DC8AE7AC29.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160824 12:42:34-!- irker813 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160824 12:42:34< irker813> wesnoth: Charles Dang wesnoth:master 30fe0cecd8c0 / data/lua/wml-tags.lua: Fixed bug in [unstore_unit] when unstoring from variable https://github.com/wesnoth/wesnoth/commit/30fe0cecd8c0f55460adcc7215f07bb7b3389e5c
20160824 12:42:46< vultraz> celmin|sleep: ^
20160824 12:46:25< vultraz> i guess my commit message is a little inaccurate, actually
20160824 12:47:00< vultraz> more like bad interaction with check_passability... 
20160824 12:47:50< vultraz> perhaps not even the 'right' fix
20160824 12:47:53< vultraz> I'll let celmin|sleep decide
20160824 12:50:03< irker813> wesnoth: Charles Dang wesnoth:master ab4a9cf80ac8 / src/save_index.cpp: Attempt to fix console warnings about attempting to draw '~RC(magenta>1)' https://github.com/wesnoth/wesnoth/commit/ab4a9cf80ac86b2ba8277a53b09ae03695bc114b
20160824 12:59:57-!- travis-ci [~travis-ci@ec2-54-224-50-50.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 12:59:58< travis-ci> wesnoth/wesnoth#10534 (master - 30fe0ce : Charles Dang): The build is still failing.
20160824 12:59:58< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154735297
20160824 12:59:58-!- travis-ci [~travis-ci@ec2-54-224-50-50.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 13:05:45< irker813> wesnoth: Charles Dang wesnoth:master 7d76d486286c / data/gui/window/drop_down_list.cfg: Dropdown List: further layout improvements https://github.com/wesnoth/wesnoth/commit/7d76d486286c39f7433a6190f51bc11bd905abf8
20160824 13:08:18-!- travis-ci [~travis-ci@ec2-54-224-50-50.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 13:08:19< travis-ci> wesnoth/wesnoth#10535 (master - ab4a9cf : Charles Dang): The build is still failing.
20160824 13:08:19< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154737347
20160824 13:08:19-!- travis-ci [~travis-ci@ec2-54-224-50-50.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 13:13:04-!- Shiki [~Shiki@141.57.58.245] has joined #wesnoth-dev
20160824 13:24:43< Shiki> Building master fails for me: Source `src/scripting/lua_api.cpp' not found, needed by target `build/debug/scripting/lua_api.o'
20160824 13:24:45-!- travis-ci [~travis-ci@ec2-54-158-184-207.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 13:24:46< travis-ci> wesnoth/wesnoth#10536 (master - 7d76d48 : Charles Dang): The build is still failing.
20160824 13:24:46< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154741289
20160824 13:24:46-!- travis-ci [~travis-ci@ec2-54-158-184-207.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 13:29:49< vultraz> Shiki: will fix
20160824 13:33:35-!- Kwandulin [~Miranda@p200300760F35BFF834A3A5DC8AE7AC29.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160824 13:34:10-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160824 13:46:15< irker813> wesnoth: Charles Dang wesnoth:master 6d39cffe14df / data/gui/widget/menu_button_default.cfg: Menu Button: increase min width https://github.com/wesnoth/wesnoth/commit/6d39cffe14df8164ed1db0df36466d2f7135b5fa
20160824 13:46:18< irker813> wesnoth: Charles Dang wesnoth:master 9579b89522fc / / (3 files in 3 dirs): Dropdown List: renamed label_right to details and made it align left https://github.com/wesnoth/wesnoth/commit/9579b89522fc1c9bee507e7590d80bff0777ab6a
20160824 13:46:21< irker813> wesnoth: Charles Dang wesnoth:master 39e4ace187de / src/gui/dialogs/preferences_dialog.cpp: Preferences: display resolution aspect ratio in details column https://github.com/wesnoth/wesnoth/commit/39e4ace187decb17204f81af07c51d20bfe80882
20160824 13:46:24< irker813> wesnoth: Charles Dang wesnoth:master b570c5635482 / projectfiles/CodeBlocks/wesnoth.cbp src/CMakeLists.txt src/SConscript: Updated build files for PR 744 https://github.com/wesnoth/wesnoth/commit/b570c5635482ea8539315b2982893323f5fd80f8
20160824 13:46:31< vultraz> Shiki: ^ should be fixed
20160824 13:50:57< Shiki> Vultraz: It is, thanks
20160824 13:51:28< fabi_> hi
20160824 13:52:40< irker813> wesnoth: Celtic Minstrel wesnoth:spirit_po 3cdbd4a5edf1 / src/ (10 files in 2 dirs): Add new translation library https://github.com/wesnoth/wesnoth/commit/3cdbd4a5edf1efc54d431dab2c6e379800d7ddb2
20160824 13:52:42< irker813> wesnoth: Celtic Minstrel wesnoth:spirit_po 255519e33f52 / src/ (CMakeLists.txt SConscript gettext.cpp): Remove option to link libintl instead of Boost.Locale https://github.com/wesnoth/wesnoth/commit/255519e33f527c6d0ee8a08e8e2ea924a1950998
20160824 13:52:44< irker813> wesnoth: Celtic Minstrel wesnoth:spirit_po 0792e1bb1fe9 / / (4 files in 3 dirs): Use spirit-po instead of Boost.Locale for translations https://github.com/wesnoth/wesnoth/commit/0792e1bb1fe94a4c194fdd72e5d32c1272a731fb
20160824 13:52:46< irker813> wesnoth: Celtic Minstrel wesnoth:spirit_po 5b5cece246b7 / src/ (gettext.hpp gettext_boost.cpp wesnoth.cpp): Remove obsolete functions from translations framework https://github.com/wesnoth/wesnoth/commit/5b5cece246b729934d0f7b901e5d4ee09ca90425
20160824 13:52:48< irker813> wesnoth: Celtic Minstrel wesnoth:spirit_po 8af3527c2817 / po/.rsync-filter projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Update XCode project https://github.com/wesnoth/wesnoth/commit/8af3527c281714b3b762015072ea370876c02b21
20160824 13:52:50< irker813> wesnoth: loonycyborg wesnoth:spirit_po 1145534bf82e / src/wesnoth.cpp: Fix compile on non-apple non-windows platforms https://github.com/wesnoth/wesnoth/commit/1145534bf82eed13f59532f82e52d85fc7394d69
20160824 13:52:52< irker813> wesnoth: Celtic Minstrel wesnoth:spirit_po 2f97e55145d9 / src/gettext_boost.cpp: Initialize po catalog from memory instead of file https://github.com/wesnoth/wesnoth/commit/2f97e55145d98a812d9e2e5f6952f9a8956fd303
20160824 13:52:54< irker813> wesnoth: Celtic Minstrel wesnoth:spirit_po c2034ec5a760 / src/gui/dialogs/ (language_selection.cpp loadscreen.cpp): Show loading screen when switching languages https://github.com/wesnoth/wesnoth/commit/c2034ec5a760ef6ab5df10fe9cffe38a41b657b1
20160824 13:52:56< irker813> wesnoth: Celtic Minstrel wesnoth:spirit_po e426580a5af8 / src/wesnoth.cpp: Fix indentation https://github.com/wesnoth/wesnoth/commit/e426580a5af88c10972197f723597a65dad0dc99
20160824 13:52:58< irker813> wesnoth: Celtic Minstrel wesnoth:spirit_po 450f7cd71eba / src/tests/utils/game_config_manager.cpp: Fix compile errors in unit tests https://github.com/wesnoth/wesnoth/commit/450f7cd71eba0ed587d020d4242bd26ec8ce5adf
20160824 13:53:00< irker813> wesnoth: Celtic Minstrel wesnoth:spirit_po c290dc23f11b / src/gettext_boost.cpp: Suppress "language missing" errors for en_US https://github.com/wesnoth/wesnoth/commit/c290dc23f11b6414710370d6fbbce9c696e03d1b
20160824 13:56:59-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20160824 13:59:25< irker813> wesnoth: Celtic Minstrel wesnoth:team_index_refactor a1d6628c89ef / src/ (76 files in 18 dirs): Refactor team indexing to avoid using 0-based indices as much as possible https://github.com/wesnoth/wesnoth/commit/a1d6628c89ef5990735127a7a15c766c12fcba42
20160824 14:03:35< mattsc> Hi — who here was it who worked on the hotkey system recently?
20160824 14:04:02< fabi_> mattsc: It was Aginor iirc.
20160824 14:04:22< mattsc> fabi_: thanks
20160824 14:04:36< mattsc> Aginor: when I put this code into a scenario: http://pastebin.com/S00cJ197
20160824 14:05:07< mattsc> and press the hotkey, it gets executed twice for me (this happens with any Lua code, this is just a simple example)
20160824 14:05:21< mattsc> By contrast, when I click on the menu option, it only gets executed once.
20160824 14:05:42< irker813> wesnoth: Celtic Minstrel wesnoth:wml_tag_porting aab32a5a04cc / / (9 files in 4 dirs): Properly port [modify_side] to Lua https://github.com/wesnoth/wesnoth/commit/aab32a5a04cc111d608f41393d417eb640895667
20160824 14:05:44< irker813> wesnoth: Celtic Minstrel wesnoth:wml_tag_porting 6b0a437b2616 / data/lua/wml-tags.lua src/ai/manager.cpp src/scripting/game_lua_kernel.cpp: Properly port [modify_ai] to Lua https://github.com/wesnoth/wesnoth/commit/6b0a437b261644e020c179f7fbb1f1e453ea8971
20160824 14:05:46< irker813> wesnoth: Celtic Minstrel wesnoth:wml_tag_porting c26ef962383a / / (4 files in 4 dirs): Properly port [heal_unit] to Lua https://github.com/wesnoth/wesnoth/commit/c26ef962383a2160f5d1f5cd51e31434ef4c13b0
20160824 14:05:57< mattsc> I checked in the inspector, the menu item is only registered once and I have looked for other potential causes, I cannot find anything.
20160824 14:06:37< mattsc> This does not happen when I put a [message] tag in there, then the respective message only appears once with both menu click and hotkey.
20160824 14:07:33< mattsc> This is with master compiled yesterday and on OS X.
20160824 14:15:25< celmin|sleep> "De-hardcoding" the offensive formatting will likely involve some way of accessing the group definitions from the command executor. (The time-of-day manager is also needed but likely already accessible from there.)
20160824 14:15:26< celmin|sleep> I'd deliberately not linked the name group in the dropdown list, because I felt that a long hotkey on a short menuitem shouldn't stretch the whole menu. Also, border=7, weird. :P
20160824 14:15:27< celmin|sleep> The hotkeys should align right in the dropdown menu, not left. Admittely that makes the time-of-day menu a bit weird, but I didn't feel that was a major issue.
20160824 14:15:28< celmin|sleep> vultraz: ^
20160824 14:16:15< vultraz> I felt border=5 was a little too small
20160824 14:16:27< celmin|sleep> Speaking of the inspector, do you like the new design, mattsc?
20160824 14:16:45< vultraz> " because I felt that a long hotkey on a short menuitem shouldn't stretch the whole menu. "
20160824 14:16:50< vultraz> you mean a long name?
20160824 14:17:00< vultraz> a long hotkey would stretch it in any case
20160824 14:17:17< celmin|sleep> Well, if they're not linked, the borders don't have to line up, do they?
20160824 14:17:23< vultraz> ... what?
20160824 14:17:32< celmin|sleep> I thought that's what linked groups did.
20160824 14:17:44< celmin|sleep> Suppose you have two menuitems.
20160824 14:17:50< vultraz> the entire dropdown list will always be the size of its longest member
20160824 14:17:57< celmin|sleep> One has a long name but short hotkey, the other a long hotkey but short name.
20160824 14:17:58< vultraz> if you mean that you want
20160824 14:18:02< vultraz> longgggg short
20160824 14:18:05< vultraz> and
20160824 14:18:07< vultraz> long s
20160824 14:18:10< vultraz> to not line up
20160824 14:18:13< vultraz> then no, that looks ugly
20160824 14:18:16< celmin|sleep> Without linked groups, the long hotkey could overlap the long name, right?
20160824 14:18:21< vultraz> right
20160824 14:18:23< vultraz> but that looks ugly
20160824 14:18:26< vultraz> in practice
20160824 14:18:56< celmin|sleep> Left-aligning it though is definitely wrong.
20160824 14:19:02< mattsc> celmin|sleep: yes, I definitely like how AI information is now split up into categories.  Makes it much easier to find things (once you get used to it)
20160824 14:19:07< vultraz> the gui1 menus did that
20160824 14:21:09< vultraz> now, where is this offensive formatting?
20160824 14:21:17< celmin|sleep> That's how hotkeys are universally done.
20160824 14:21:26< celmin|sleep> Right-aligned, I mean.
20160824 14:21:39-!- celmin|sleep is now known as celticminstrel
20160824 14:21:45< vultraz> hmmmm
20160824 14:21:53-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 14:21:54< travis-ci> wesnoth/wesnoth#10538 (spirit_po - c290dc2 : Celtic Minstrel): The build is still failing.
20160824 14:21:54< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154753606
20160824 14:21:54-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 14:21:54< vultraz> chrome right-aligns then, so you must be right
20160824 14:22:01< celticminstrel> The offensive formatting is in the editor. I don't remember exactly where.
20160824 14:22:07< vultraz> :|
20160824 14:22:57< celticminstrel> Context manager
20160824 14:23:18< celticminstrel> expand_time_menu, expand_local_time_menu
20160824 14:23:50< celticminstrel> Also search for expand_paette_groups_menu, but that shouldn't be in the context manager.
20160824 14:24:22< celticminstrel> Context manager is in src/editor/map
20160824 14:24:49< celticminstrel> The other should be in src/editor/palette somewhere, and there are several versions of the function for different palettes.
20160824 14:25:05< celticminstrel> Of course, you've already seen where the markup is handled on the other end.
20160824 14:25:14-!- Shiki [~Shiki@141.57.58.245] has quit [Remote host closed the connection]
20160824 14:25:19< vultraz> yeah, this is evil gui1 markup
20160824 14:25:31< celticminstrel> Yeah, it's absolutely GUI1 markup.
20160824 14:25:55< vultraz> hmm
20160824 14:26:01< vultraz> this will require some slight refactoring
20160824 14:27:20< celticminstrel> Obviously. :P
20160824 14:28:26< celticminstrel> My thought was to assign some consistent generated ID instead - "editor-time-of-day-" + time.id, or "editor-group-" + group_type + "-" + group_name, and then expand that in get_menu_images to fetch the appropriate image.
20160824 14:28:43< celticminstrel> However, I don't know a way to access the editor palettes from there.
20160824 14:28:59< vultraz> I don't follow
20160824 14:29:16< celticminstrel> Which part?
20160824 14:29:25< vultraz> consistent generated ids?
20160824 14:29:28< vultraz> whatever for
20160824 14:29:39< celticminstrel> Well, show_menu is passed a list of menuitem IDs.
20160824 14:30:04< celticminstrel> Except that for these two uses it instead gets the actual menuitem content.
20160824 14:30:10< vultraz> menuitem ids.. and not labels directly...
20160824 14:30:12< vultraz> hm
20160824 14:30:16< vultraz> this is unfortunate
20160824 14:30:26< celticminstrel> It's the IDs usedi n ThemeWML.
20160824 14:30:29< celticminstrel> &used in
20160824 14:30:30< vultraz> ugh
20160824 14:30:33< celticminstrel> ^ not&
20160824 14:30:37 * vultraz curses profusely
20160824 14:30:54< celticminstrel> Changing to not use IDs would probably be bad.
20160824 14:31:04< vultraz> theme WML is the worst thing in wesnoth
20160824 14:31:08< vultraz> it's worse than gui1
20160824 14:31:38< celticminstrel> It's not something we can just drop though.
20160824 14:31:42< vultraz> we need to experiment with gui2 for the in-game UI
20160824 14:32:16< vultraz> I believe it might be possible
20160824 14:32:54< vultraz> anyway, I decided not to do this now :|
20160824 14:33:09< vultraz> the theme code is pissing me off
20160824 14:34:35< vultraz> (using gui2 for the in-game UI will of course take some work)
20160824 14:35:23-!- travis-ci [~travis-ci@ec2-54-224-50-50.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 14:35:24< travis-ci> wesnoth/wesnoth#10540 (team_index_refactor - a1d6628 : Celtic Minstrel): The build is still failing.
20160824 14:35:24< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154756109
20160824 14:35:24-!- travis-ci [~travis-ci@ec2-54-224-50-50.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 14:35:29< celticminstrel> I think a GUI2 dialog could possibly be constructed on-the-fly from the ThemeWML, but I didn't want to try that.
20160824 14:35:36< vultraz> no
20160824 14:35:46< vultraz> we must destroy theme wml in its current state
20160824 14:36:10< vultraz> not make it compatible :|
20160824 14:36:46< celticminstrel> "in its current state"
20160824 14:36:56< celticminstrel> That sounds like you want to improve ThemeWML, not destroy it.
20160824 14:37:32< vultraz> obviously we will need some kind of layout control
20160824 14:37:39< vultraz> possibly with gui2 dirctly 
20160824 14:37:41< vultraz> directly
20160824 14:38:22< vultraz> but the current syntax is just horrible
20160824 14:38:24< vultraz> and must go
20160824 14:38:47< vultraz> relative placing isn't inherently bad, but themewml's design is just... so bad
20160824 14:40:53< vultraz> I believe the viewport widget was intended to hold the game map
20160824 14:40:56< vultraz> but I have no idea
20160824 14:41:00-!- Kwandulin [~Miranda@p200300760F35BFF80D4B447DCF2B98F9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160824 14:43:56< vultraz> the biggest problem is gui2 is kinda....
20160824 14:44:00< vultraz> well...
20160824 14:44:28< vultraz> its design speaks more of popup windows than game screens.
20160824 14:44:50< vultraz> something I've been worried about when constructing these big fullscreen dialogs is they might allow other dialogs to exist underneath...
20160824 14:45:19< vultraz> I *think* set_restore(true) ensures they don't, but I'm not sure..
20160824 14:47:37< celticminstrel> What do you mean?
20160824 14:48:15< vultraz> try setting the background opacity of the loadscreen to something other than transparent
20160824 14:48:19< vultraz> and load a game
20160824 14:48:27< vultraz> you'll see the main menu underneath
20160824 14:50:28< vultraz> hm
20160824 14:50:39< vultraz> even WITH set_restore(true) it's still to be seen
20160824 14:50:55-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 14:50:56< travis-ci> wesnoth/wesnoth#10541 (wml_tag_porting - c26ef96 : Celtic Minstrel): The build failed.
20160824 14:50:56< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154758869
20160824 14:50:56-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 14:51:51< celticminstrel> The Lua console can appear on top of the inspector and vice versa, so it's definitely possible to have layers of fullscreen dialogs open.
20160824 14:52:04< vultraz> yes
20160824 14:52:33< vultraz> and in fact, we want that
20160824 14:53:01-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20160824 14:53:12< celticminstrel> As long as the hidden dialogs don't redraw or receive any events, I don't think there'd be a problem.
20160824 14:54:10< vultraz> I hope so
20160824 14:55:00< vultraz> now, if we do port the gamescreen to gui2, that would mean making the map a widget.
20160824 14:55:13< celticminstrel> Apart from the left-aligned hotkeys, the menus still look good. I don't have a screenshot handy to prove that it's an improvement.
20160824 14:55:38< vultraz> this poses two major problems
20160824 14:55:42< vultraz> Scrolling and Events
20160824 14:56:04< vultraz> scrolling might be less of a problem
20160824 14:56:12< vultraz> events more
20160824 14:56:54< vultraz> the widget should optimally know nothing about the gamestate.
20160824 14:57:04< vultraz> and just display what its told
20160824 14:57:55< vultraz> the question here is how to evaluate the gamestate
20160824 14:57:58< celticminstrel> Mini-create is better too.
20160824 14:57:58-!- boucman_work [~boucman@229.29.205.77.rev.sfr.net] has quit [Read error: Connection reset by peer]
20160824 14:58:31< vultraz> gui2's event pipeline is separate from the "main" one.
20160824 14:59:11-!- boucman_work [~boucman@229.29.205.77.rev.sfr.net] has joined #wesnoth-dev
20160824 14:59:18< vultraz> so how will we pass events back and forth to and from the gamestate?
20160824 14:59:22< celticminstrel> Does the map really have to be a widget?
20160824 14:59:48< vultraz> how will we simplify the million bloody classes we have for displaying stuff on the map?
20160824 14:59:53< vultraz> celticminstrel: it really should be
20160824 14:59:56< celticminstrel> What?
20160824 15:00:27< vultraz> game_map, gameboard, game_display... 
20160824 15:00:39< celticminstrel> Oh.
20160824 15:00:56< celticminstrel> Is there any reason the game interface couldn't be a non-modal dialog though?
20160824 15:01:07< celticminstrel> So that you can interact either with the map or with the dialog.
20160824 15:01:13< vultraz> a non-modal dialog drawn over the map?
20160824 15:01:17< celticminstrel> Yeah.
20160824 15:01:18< vultraz> I don't really like that
20160824 15:01:27< vultraz> A. how will we make it non-rectangular 
20160824 15:01:46< celticminstrel> What? Why do you want it non-rectangular?
20160824 15:01:53< vultraz> B. it still leaves us with the handling of the game map display managed outside gui2
20160824 15:02:11< vultraz> celticminstrel: sidebar + menu bar = non rectangular :|
20160824 15:02:37< celticminstrel> It's totally rectangular.
20160824 15:02:49< vultraz> it's TWO rectangles 
20160824 15:03:00< celticminstrel> What?
20160824 15:03:09< vultraz> the sidebar is a rectangle
20160824 15:03:12< vultraz> and the menu bar is another
20160824 15:03:14< celticminstrel> Oh, you mean making the dialog non-rectangular.
20160824 15:03:17< celticminstrel> Not the map.
20160824 15:03:21< vultraz> yes
20160824 15:03:28< vultraz> the dialog cannot be non-rectangular
20160824 15:03:33< vultraz> ergo, the game map must be a widget
20160824 15:03:48< celticminstrel> Hmm, maybe.
20160824 15:06:28< vultraz> so what we'd need to do is plug in the  game controller to the widget
20160824 15:07:13< vultraz> (I use game controller loosely)
20160824 15:07:33< vultraz> so, for example, unit moves from hex a to b
20160824 15:07:39< vultraz> widget notifies 
20160824 15:07:53< vultraz> appropriate action is calculated 
20160824 15:08:12< vultraz> any resulting visual changes are sent to the widget
20160824 15:08:16< vultraz> make sense?
20160824 15:08:35< celticminstrel> Maybe?
20160824 15:09:13< vultraz> my point is the gamestate and its visual representation should be seperate 
20160824 15:09:17< vultraz> separate 
20160824 15:09:21< celticminstrel> Of course.
20160824 15:10:00< vultraz> I think we'll need Aginor's global event context possibly...
20160824 15:11:27< vultraz> anyway, the map itself is really just a giant surface, I think
20160824 15:11:36< vultraz> drawing it and scrolling it should be easy
20160824 15:11:50< vultraz> one question is how to distinguish hexes
20160824 15:13:04-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160824 15:19:10< celticminstrel> I'm not sure selecting a game by index is very useful...
20160824 15:26:10< celticminstrel> So find_widget(w, ...) is about equivalent to w->find plus dynamic_cast.
20160824 15:28:31< vultraz> oh?
20160824 15:28:39< vultraz> w->find is not more efficient?
20160824 15:29:20< celticminstrel> Probably not, I guess?
20160824 15:31:43< celticminstrel> mp_create_game_set_password.cfg and mp_create_game_choose_mods.cfg - are these needed?
20160824 15:33:23-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160824 15:35:06< vultraz> what?
20160824 15:35:33< vultraz> I don't know what you're talking about
20160824 15:35:39< celticminstrel> In data/gui/window
20160824 15:35:59< vultraz> hmm
20160824 15:36:24< vultraz> I likely not..
20160824 15:36:26 * celticminstrel assumes there are also corresponding source files in src/gui/dialogs
20160824 15:36:27< vultraz> likely*
20160824 15:36:29< vultraz> but let me see
20160824 15:37:17< vultraz> doesn't seem to be used
20160824 15:37:26< vultraz> can't even find a file for the former
20160824 15:37:28< vultraz> er
20160824 15:37:29< vultraz> latter
20160824 15:37:36< vultraz> former I integrated directly into Create
20160824 15:39:56< vultraz> the file was added in 2012..
20160824 15:40:03< vultraz> but I don't see it anywhere :|
20160824 15:40:26< vultraz> ah
20160824 15:40:30< vultraz> it was deleted in 2013
20160824 15:40:51< vultraz> I guess after mod selection was integrated into gui1?
20160824 15:41:42< vultraz> it can be removed
20160824 15:41:46-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Ping timeout: 244 seconds]
20160824 15:42:09< vultraz> ah, the depcheck dialog..
20160824 15:42:21< vultraz> is that supposed to show dependencies of games too>?
20160824 15:42:51< vultraz> I could have sworn that was a feature
20160824 15:43:03-!- boucman_work [~boucman@229.29.205.77.rev.sfr.net] has quit [Ping timeout: 240 seconds]
20160824 15:43:11< vultraz> seems not for that, at least..
20160824 15:43:23< vultraz> but i could still have sworn it was a thing..
20160824 15:45:58-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160824 15:49:43< celticminstrel> Did I miss anything important?
20160824 15:51:10-!- JyrkiVesterinen [~JyrkiVest@87-100-204-68.bb.dnainternet.fi] has joined #wesnoth-dev
20160824 15:57:06-!- Kwandulin [~Miranda@p200300760F35BFF80D4B447DCF2B98F9.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160824 15:58:35-!- Shiki [~Shiki@141.39.226.227] has joined #wesnoth-dev
20160824 16:01:05< celticminstrel> I'll take that as a no. :P
20160824 16:01:17-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 16:13:56< vultraz> celticminstrel: btw, was that the correct fix to [unstore_unit[
20160824 16:13:57< vultraz> ]
20160824 16:15:35< celticminstrel> I looked, but I'm not sure.
20160824 16:16:08< celticminstrel> What was the bug?
20160824 16:16:19< vultraz> i was getting errors about the unit not being on the map when unstoring from a variable 
20160824 16:18:10< celticminstrel> I guess it could be right.
20160824 16:19:33-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160824 16:21:04-!- Shiki is now known as shiki|afk
20160824 16:22:55< irker813> wesnoth: Charles Dang wesnoth:master 379d4ea17b26 / data/gui/window/mp_create_game_choose_mods.cfg: Cleaned up remaining file for removed dialog https://github.com/wesnoth/wesnoth/commit/379d4ea17b2690b893740923602445a258ad69ee
20160824 16:23:50< vultraz> did you see my comment in the commit about the use of has_minimum?
20160824 16:24:12< celticminstrel> Yes, I was going to do something with that.
20160824 16:24:32< vultraz> that being?
20160824 16:24:46< celticminstrel> Enabling it?
20160824 16:24:56< vultraz> but then comobox selections are not shown
20160824 16:24:57< celticminstrel> And testing it out and seeing if it works and if so committing it.
20160824 16:25:03< vultraz> er, menu buttons
20160824 16:25:06< vultraz> I already tested it
20160824 16:25:07< vultraz> it works
20160824 16:25:36< vultraz> just then the dropdown menus... (in hindsight, menu_button was a bad name for the widget)'s initial selection is not shown
20160824 16:28:34< vultraz> what do you propose to do?
20160824 16:28:47< celticminstrel> Something. Not 100% sure yet.
20160824 16:30:03< celticminstrel> It's after I fix MP Create tiny-res.
20160824 16:30:28< vultraz> keep a pointer
20160824 16:30:33< celticminstrel> ?
20160824 16:30:47< vultraz> i assume the problem is it cannot find the password/game name boxes?
20160824 16:30:56< celticminstrel> That, and the scrollbar.
20160824 16:31:25< celticminstrel> My assumption is that the problem lies in the lack of overridden tstacked_widget::find.
20160824 16:31:29< celticminstrel> But not quite sure.
20160824 16:31:54< celticminstrel> Widgets in stacked layers are still logically in the dialog even when the layer is hidden, right? So I'd expect find_widget to find them.
20160824 16:32:05< vultraz> yes
20160824 16:32:15< vultraz> but they cannot be found if they're not in the visible layer
20160824 16:32:20< vultraz> if you can fix this it'd be great
20160824 16:32:43< celticminstrel> I mean, there's no compelling reason for this behaviour, right?
20160824 16:32:51< celticminstrel> And it's not relied upon anywhere that you know of?
20160824 16:33:43< celticminstrel> Would be good if we can prove it's not relied upon anywhere.
20160824 16:33:53< vultraz> not that I know of
20160824 16:33:59< vultraz> in fact, I've had to work around it :|
20160824 16:38:03< vultraz> I just love when wesnoth decides to crash for no reason
20160824 16:39:35< celticminstrel> Loading screen?
20160824 16:39:49< vultraz> after
20160824 16:40:05< vultraz> but possibly still in the same area
20160824 16:40:32< vultraz> anyway, let me fix my fixup to the RC(magenta>1) bug
20160824 16:40:34< vultraz> I screwed up :|
20160824 16:42:02< vultraz> ya know
20160824 16:42:16< vultraz> why is the save_index the only file we save in non-human readable format
20160824 16:42:36< celticminstrel> No idea.
20160824 16:42:44< celticminstrel> Is it a binary file?
20160824 16:42:58< vultraz> I think so
20160824 16:44:26< vultraz> (we could save savefiles as binary files and save space :P )
20160824 16:44:33< vultraz> (I jest)
20160824 16:45:03< celticminstrel> Given that we compress them, it probably wouldn't even save space.
20160824 16:47:32< vultraz> well I seem to have totally misfixed this fix and now don't know how to fix the fix
20160824 16:48:24< celticminstrel> Hmm.
20160824 16:48:45-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 16:48:46< travis-ci> wesnoth/wesnoth#10542 (master - 379d4ea : Charles Dang): The build is still failing.
20160824 16:48:46< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154805259
20160824 16:48:46-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 16:48:54-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160824 16:50:53< tad_> I'm getting a runtime error killing the scenario when unstoring a unit from a variable to the recall list. Does that ring a bell on a recent change?
20160824 16:51:33< celticminstrel> Guessing that's vultraz's attempt to fix the RC error messages which apparently turned out to be wrong.
20160824 16:51:39< vultraz> uhh
20160824 16:51:41< vultraz> what???
20160824 16:51:43< vultraz> celticminstrel: ??
20160824 16:51:56< celticminstrel> Well, that was the context originally producing the errors, after all?
20160824 16:51:58< vultraz> he said unstoring unit :|
20160824 16:52:03< vultraz> not rc fixes
20160824 16:52:11< tad_> I don't see how that could be. I'm leaning toward the refactoring CM did
20160824 16:52:20< celticminstrel> Could be.
20160824 16:53:59< tad_> I could bisect, I guess.
20160824 16:54:08< vultraz> current master?
20160824 16:54:12< tad_> Yes.
20160824 16:54:25< vultraz> what is the error?
20160824 16:55:13< tad_> HAng on. It's not on screen any more. Line 1050 in wml_tags, iirc
20160824 16:55:39< tad_> Soemthing about unstoring to recall and it was already on recall. 
20160824 16:55:40< vultraz> sounds like where I fixed that unstore_unit thing..
20160824 16:55:42< vultraz> huh
20160824 16:55:45< vultraz> ok, different
20160824 16:56:02< celticminstrel> Ah, that sounds like something probably related to my refactoring then.
20160824 16:56:04-!- shiki|afk is now known as Shiki
20160824 16:56:06< tad_> And, of course, I checked and it was not on recall, but the variable had x,y=recall,recall
20160824 16:56:28< irker813> wesnoth: Charles Dang wesnoth:master e3f397ca8cba / src/save_index.cpp: Fixup ab4a9cf80ac86b https://github.com/wesnoth/wesnoth/commit/e3f397ca8cbad80c69e06cd6550a516a1ad4e516
20160824 16:56:34< vultraz> ^ that should properly fix the RC errors
20160824 16:56:52-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160824 16:56:52-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20160824 16:58:04< tad_> error scripting/lua: lua/wml-tags.lua:1053: calling 'to_recall' on bad self (unit already on recall list)
20160824 16:58:23< tad_> Line s/b 1050 but I have a logger to get the unit name so I know where to look
20160824 16:59:11< celticminstrel> There were conflicts on the relevant lines, so it's possible I resolved them incorrectly.
20160824 16:59:26< celticminstrel> Or wait, that's not true.
20160824 16:59:37< celticminstrel> Those conflicts were on the team_index_refactor branch.
20160824 16:59:41< celticminstrel> Hmm.
20160824 17:01:27-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160824 17:02:38< vultraz> I was going to make the load game dialog show all human leaders
20160824 17:04:37< zookeeper> tad_, so you're going to convert the TSG recall list switching to just use [filter_recall] and avoid all storing/unstoring?
20160824 17:05:57< tad_> zookeeper: I'm looking that way. It's working great in my proof-of-concept. Makes it much easier to manage who is recallable when
20160824 17:06:31< zookeeper> sure, sounds good.
20160824 17:08:33< nore> Hm, how would you test multiplayer locally?
20160824 17:08:43< vultraz> Local Game?
20160824 17:08:45< nore> I mean, hosting a server and several clients locally
20160824 17:08:48< tad_> I am very releaved you think so, zookeeper 
20160824 17:09:06< nore> Local Game does not do what I want it to do, since I want several wesnoth instances
20160824 17:09:22< nore> (I am modifying the engine and I want to test my changes)
20160824 17:09:50< vultraz> you need to launch several instances of wesnoth, then
20160824 17:09:50< tad_> nore: do you have vmware or virtualbox? I would fire up a box for the server and a couple for the clients.
20160824 17:09:52< celticminstrel> Apparently this was an intentional feature of the "independent" placement policy.
20160824 17:10:01< celticminstrel> That it only searches visible items.
20160824 17:10:06< nore> vultraz: yes, but it doesn't work
20160824 17:10:11-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160824 17:10:22< nore> only one instance is actually connected to the server
20160824 17:10:23< vultraz> nore: are you on 1.13?
20160824 17:10:25< nore> yes
20160824 17:10:37< vultraz> might as well just connect to the official server
20160824 17:10:47< vultraz> and host a game there
20160824 17:10:51< vultraz> no one's on it anyway
20160824 17:10:55< nore> vultraz: well, I modified the server code
20160824 17:10:58< nore> that's the problem
20160824 17:10:58< celticminstrel> You could choose the "host a game" option.
20160824 17:10:59< vultraz> oh
20160824 17:11:05< celticminstrel> Or run the server from the command-line.
20160824 17:11:13< vultraz> you need to do as celticminstrel said, then
20160824 17:11:17< nore> celticminstrel: lauching the server *does* work
20160824 17:11:26< nore> the problem is that only one client will connect to it
20160824 17:11:33< celticminstrel> Huh?
20160824 17:11:34< zookeeper> tad_, yeah it's not that i think the current storing/unstoring is somehow bad, but the filtering method seems a bit more straightforward and not particularly likely to end up being buggy
20160824 17:11:34-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 17:11:37< vultraz> you need to do Connect to Server
20160824 17:11:40< tad_> zookeeper: The huge payoff for using the recall list instead of all those store/unstore will be in DM where it's heavily used
20160824 17:11:42< nore> more precisely:
20160824 17:11:54< nore> I can launch a second client
20160824 17:11:54< vultraz> and put in an address
20160824 17:12:01< nore> tell it to connect to localhost server
20160824 17:12:12< tad_> zookeeper: plus all those 'returns in poor shape' issues go away because you just leave the units on the map and they get fixed as normal
20160824 17:12:19< nore> it works, but when I try to start a game, this client does not enter the game
20160824 17:12:23< vultraz> hm
20160824 17:12:27< vultraz> I don't know, then
20160824 17:12:29< vultraz> ask loonycyborg 
20160824 17:12:34< zookeeper> tad_, sure, i think DM has traditionally had a lot of issues with that because the storing/unstoring code has been poorly done
20160824 17:12:37< nore> if then I disconnect, *no client at all* is able to connect to the server
20160824 17:12:54< nore> (except the one hosting)
20160824 17:13:01< vultraz> celticminstrel: the independent placing policy, eh
20160824 17:13:12< zookeeper> tad_, or at least that's been my impression, i've very rarely even looked at DM's code
20160824 17:13:13< nore> hm, will do then, do you know when they are around?
20160824 17:13:23< celticminstrel> vultraz: Currently trying to figure out where that's used.
20160824 17:13:34< vultraz> celticminstrel: somewhere in the generator?
20160824 17:13:38< tad_> zookeeper: I sat down to clean up my proof-of-concept and merge it to TGS-elven branch but hit this show-stopper about unstoring and need to clear that error, first
20160824 17:13:45< vultraz> nore: he's in the chat now, dunno if he's 'around', though
20160824 17:13:51< celticminstrel> vultraz: BTW, if someone implemented the matrix placement policy, you could probably do card-based listboxes.
20160824 17:14:03< loonycyborg> I'm here, though a bit distracted
20160824 17:14:15< nore> oh, nice
20160824 17:14:41< loonycyborg> so what's the original question?
20160824 17:14:53< nore> problem is that I can't test multiplayer locally
20160824 17:15:06< nore> by that, I mean launching wesnothd and several wesnoth instances
20160824 17:15:19< nore> (or using "host game")
20160824 17:15:37< nore> after hosting a game with one wesnoth instance
20160824 17:15:46< celticminstrel> vultraz: So it's used only by multipage and by stacked widget.
20160824 17:15:55< nore> I can launch a second client, and tell it to connect to localhost
20160824 17:16:10< nore> it works, but this client does not enter the games that are created
20160824 17:16:17< nore> though, it gets to choose its side
20160824 17:16:32< celticminstrel> What's multipage about anyway that's different from stacked widget...
20160824 17:16:39< nore> after it disconnects, no other client at all is able to connect to localhost server
20160824 17:18:04< tad_> celticminstrel: I'm getting lost in all the conversations. I think I'll drop out and try to bisect to find the error
20160824 17:18:11< loonycyborg> what do you mean it doesn't enter the games?
20160824 17:18:21< celticminstrel> tad_: I think I have an idea what the error is already.
20160824 17:18:36< tad_> celticminstrel: OK. I'll wait .. bisect is a pain
20160824 17:18:37< loonycyborg> can you run wesnothd from command line maybe?
20160824 17:18:42< loonycyborg> to see its output
20160824 17:18:49< nore> it stays in the game lobby, with the screen showing each player's faction
20160824 17:18:53< celticminstrel> Ah, I guess multipage has a single definition, like a listbox...
20160824 17:19:05< nore> I get a message about using the same IP as well
20160824 17:19:13< loonycyborg> that is to be expected
20160824 17:19:23< loonycyborg> same ip that is
20160824 17:19:53< celticminstrel> vultraz: For future reference, I think all we need to do to enable find_widget to find things in hidden tabs is to call select_layer(-1) in post_show.
20160824 17:19:59< celticminstrel> Trying that now,
20160824 17:20:13< tad_> celticminstrel: To tickle the bug I use TGS S06a .. de-fog, kill the queen, long move gerrick to target
20160824 17:20:38< loonycyborg> nore: it stays there after there is enough players?
20160824 17:20:46< nore> loonycyborg: yes
20160824 17:21:03< loonycyborg> generally there is no option to continue until all network slots are filled
20160824 17:21:23< loonycyborg> after there's enough players there's a button to start game on the host
20160824 17:21:27< nore> hmmm
20160824 17:21:43< loonycyborg> meaning the client that created the game
20160824 17:21:49< nore> wesnothd segfaulted, but the "Host Game" didn't show anything
20160824 17:22:08 * nore tries to see if this is due to his changes or not
20160824 17:22:26< loonycyborg> ah so it segfaulted
20160824 17:22:34< loonycyborg> which version you're using?
20160824 17:22:41< loonycyborg> git master?
20160824 17:22:44-!- travis-ci [~travis-ci@ec2-54-224-50-50.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 17:22:46< travis-ci> wesnoth/wesnoth#10543 (master - e3f397c : Charles Dang): The build is still failing.
20160824 17:22:46< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154813692
20160824 17:22:46-!- travis-ci [~travis-ci@ec2-54-224-50-50.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 17:23:31< nore> ok, my fault
20160824 17:24:15< nore> strange thing though, that the client that hosted the game didn't report the crash of wesnothd
20160824 17:24:23< nore> (I mean, the client that started wesnothd)
20160824 17:25:01< celticminstrel> Yeah, that seems like a thing that should happen, doesn't it...
20160824 17:25:51-!- Kwandulin [~Miranda@p200300760F35BFF8748F2530F77E8BD2.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160824 17:26:54< loonycyborg> nope, the only thing client will see
20160824 17:26:58< loonycyborg> if server crashes
20160824 17:27:04< celticminstrel> tad_: So the error is "unit already on recall list", right?
20160824 17:27:06< loonycyborg> is tcp reset
20160824 17:27:33< vultraz> celticminstrel: does add_child return the added child?
20160824 17:27:34< nore> hm, sorry to have bothered you then
20160824 17:27:46< tad_> celticminstrel:  calling 'to_recall' on bad self (unit already on recall list)
20160824 17:27:57< nore> oh, and why is the "control" command not available, btw?
20160824 17:28:00< celticminstrel> vultraz: You could always check yourself? But if you mean a tree view, yes.
20160824 17:28:06< vultraz> no, I mean config::
20160824 17:28:10< celticminstrel> Ah.
20160824 17:28:14< loonycyborg> anyway
20160824 17:28:22< celticminstrel> I think it does.
20160824 17:28:22< loonycyborg> I only worked on serverside atm
20160824 17:28:24< vultraz> it does seem to
20160824 17:28:28< tad_> celticminstrel: note that the unit is in a variable (x,y=recall,recall) and does not really appear in the recall list at that point
20160824 17:28:34< loonycyborg> so I don't know much about client's error handling
20160824 17:28:45< celticminstrel> vultraz: What would I do to make a grid align to the top?
20160824 17:28:53< vultraz> I cannot say
20160824 17:28:56< celticminstrel> So that any extra space is filled in at the bottom.
20160824 17:28:57< vultraz> without le context
20160824 17:29:08< vultraz> but in general
20160824 17:29:21< vultraz> the column above it should have vertical_grow = "true"
20160824 17:31:33< celticminstrel> That doesn't seem to work...
20160824 17:31:45< vultraz> it's finnicky 
20160824 17:31:49< vultraz> let me see the code
20160824 17:33:19< celticminstrel> http://celmin.pwcsite.com/wesnoth/paste.txt
20160824 17:33:27< celticminstrel> Of course that goes in a stacked widget.
20160824 17:33:55< vultraz> didn't I fix this..?
20160824 17:34:25< vultraz> ok, first off, you need grow_factor = 0 in the first toplevel row and = 1 in the second
20160824 17:35:08< celticminstrel> I don't think it was an issue before.
20160824 17:35:17< celticminstrel> So no, you didn't fix it.
20160824 17:36:06< vultraz> I wonder if load game should display every leader..
20160824 17:36:14< vultraz> thoughts?
20160824 17:37:03< tad_> vultraz: there are times when some campaigns have several canrecruit=yes units on the map at once. I think LoW maxes that at 5
20160824 17:37:11< celticminstrel> tad_: I think I found the issue.
20160824 17:37:32< vultraz> tad_: yes, my question is whether the load game dialog should display them all
20160824 17:38:08< celticminstrel> Okay, so...
20160824 17:38:13< tad_> vultraz: My opinion: all or none. I like the consistency of 'none' myself.
20160824 17:38:22< vultraz> ...eh?
20160824 17:38:28< vultraz> right now it shows the sprite of the human leader
20160824 17:38:30< vultraz> and that's it
20160824 17:39:05< tad_> vultraz: thats what I meant: LoW can have 5 human-controlled leaders
20160824 17:39:08< celticminstrel> Previously, that error would be triggered if the unit was on any side's recall list.
20160824 17:39:18< celticminstrel> I changed it to only trigger if it was on the same side's recall list.
20160824 17:39:22< vultraz> ...and?
20160824 17:39:25< vultraz> should it show all 5 :|
20160824 17:39:28< vultraz> or just the first
20160824 17:39:37< vultraz> I'm leaning to 'all'
20160824 17:39:48< tad_> and I'm leaning to 'none'
20160824 17:39:53 * tad_ chuckles
20160824 17:40:06< vultraz> oh well :P
20160824 17:40:21< celticminstrel> None is certainly simpler.
20160824 17:40:45< celticminstrel> I don't really have an opinion on what's better though.
20160824 17:41:02< tad_> When it's showing them, it's not consistent. So when I'm arrowing up and down my eye is drawn to the changes. That can be a good thing, I guess.
20160824 17:41:21< celticminstrel> So, anyway, the question is: should wesnoth.put_recall_unit be able to move a unit from side N's recall list to side M's recall list?
20160824 17:41:26< vultraz> I don't know why it doesn't show them sometimes
20160824 17:41:31< vultraz> celticminstrel: yes
20160824 17:41:37< celticminstrel> Okay, then...
20160824 17:42:13< tad_> celticminstrel: If the creator said do it, then do it. It's only a problem if it's inadvertent
20160824 17:42:42< celticminstrel> I think this should fix it, then.
20160824 17:42:54< tad_> celticminstrel: but I can see requiring modify_unit or set_variable to change the side, first, if it presents problems
20160824 17:42:57< celticminstrel> I'll repeat my test case in the Lua console, of course, just to be sure.
20160824 17:43:27< zookeeper> vultraz, eh, _how_ would you display X leaders in the load dialog anyway? stacking them vertically or what?
20160824 17:43:36< vultraz> indeed
20160824 17:43:59< celticminstrel> Or a multi-page widget with navigation arrows?
20160824 17:44:20< zookeeper> you can't display an arbitrary number of leaders nor can you really display even 5, i'd think.
20160824 17:44:27< vultraz> says who?
20160824 17:44:32< tad_> zookeeper: make a campaign with 32 human leaders and have vultraz shrink-to-fit to 1x1 images .. muhahaha
20160824 17:44:36< celticminstrel> Well, that's why I said none is simpler.
20160824 17:44:49< zookeeper> well sure, if you shrink-to-fit then you can have an arbitrary number, i guess
20160824 17:44:57 * vultraz will now do this since you told him it's impossible
20160824 17:45:04 * vultraz feels very 'murican right now
20160824 17:45:38< zookeeper> is there any mainline case where the side 1 leader (or the leader of the numerically first human side) wouldn't make sense as the only displayed leader..?
20160824 17:45:52< zookeeper> err
20160824 17:45:53< zookeeper> i mean
20160824 17:46:06< zookeeper> yeah nevermind, i got confused for a moment
20160824 17:47:09< celticminstrel> vultraz: Adding the grow factors helped.
20160824 17:47:12< celticminstrel> Still not quite right though.
20160824 17:47:18< vultraz> celticminstrel: screenshot!
20160824 17:47:22< celticminstrel> Oddly, only in the left column. Right column is fine.
20160824 17:47:25< tad_> zookeeper: what I noticed, before this last rc>1 fix was the leader for TSG S06a changes as I arrowed over the saves .. Deoran .. Gerrick .. Gerrick .. finally Deoran on the winnig move
20160824 17:47:28< vultraz> giff screenshot
20160824 17:47:38< vultraz> plox
20160824 17:48:58< tad_> vultraz: just don't move the clickable UI elements when you resize. It's a pain to be clicking the scroll down button and suddenly have it be a savegame and loading it by accident :P
20160824 17:49:23< vultraz> say what now?
20160824 17:49:35< celticminstrel> http://celmin/pwcsite.com/wesnoth/mp_create_unbalanced.png
20160824 17:49:35< vultraz> the dialog doesn't resize :/
20160824 17:50:00< vultraz> celmin’s server DNS address could not be found.
20160824 17:50:08< celticminstrel> http://celmin.pwcsite.com/wesnoth/mp_create_unbalanced.png
20160824 17:50:10< tad_> I've seen that on other programs. You're scolling and something wants to resize and suddenly it's not a scroll-down button but a load-this-file area ...
20160824 17:50:28< celticminstrel> Sorry, typo.
20160824 17:50:38< vultraz> what... is going on here O_O
20160824 17:50:38< celticminstrel> I don't have a browser open right now.
20160824 17:50:56< vultraz> celticminstrel: i need all the code
20160824 17:51:02< celticminstrel> The only change from the code I showed is adding grow factors.
20160824 17:51:08< celticminstrel> By all the code you mean...?
20160824 17:51:32< vultraz> er..
20160824 17:51:33< vultraz> hm
20160824 17:51:35< vultraz> wait
20160824 17:51:55< vultraz> celticminstrel: ok
20160824 17:52:05< vultraz> in both columns of the second row, add horizontal_grow = "true"
20160824 17:52:55< vultraz> er
20160824 17:52:57< vultraz> hmm..
20160824 17:53:03< celticminstrel> Trying that now.
20160824 17:53:09< vultraz> celticminstrel: report back
20160824 17:53:36< celticminstrel> I wonder if I should prune old screenshots from that directory...
20160824 17:55:04< vultraz> does _GUI_MP_CREATE_MAP_SETTINGS start with a [row]?
20160824 17:55:19< celticminstrel> Yeah.
20160824 17:55:24< celticminstrel> It contains several rows.
20160824 17:55:37< tad_> You could be like my wife and insist on a 4T external drive because she never deletes a thing and loves to make copies because she's afraid she will lose them if she works on the originals.
20160824 17:55:57< celticminstrel> I wish I had a 4T drive.
20160824 17:55:58-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160824 17:56:29< celticminstrel> Though my total disk space is probably close to 3TB actually..
20160824 17:56:34< celticminstrel> If I recall correctly.
20160824 17:57:48< vultraz> celticminstrel: try this code http://pastebin.com/gtwTqnPC
20160824 17:57:50< tad_> I got one on close-out at Sam's Club (WD MyCloud) and should have gotten the other they had on the shelf. I ssh in and kill all the cloud junk and it's a very nice 4T NAS I got for only $50 because Sam's wanted to get it off the shelf.
20160824 17:57:51< vultraz> verbatim
20160824 17:58:53< celticminstrel> Ugh, I wanted to avoid having the needless extra grid layer, but I'll try it, I guess.
20160824 17:59:26< vultraz> true
20160824 17:59:37< vultraz> but it's better to keep the top row together than splitting it
20160824 18:00:03< celticminstrel> It wasn't really visibly split though.
20160824 18:00:29< vultraz> right
20160824 18:00:32 * celticminstrel pipes curl through pbcopy and gets the paste directly to the clipboard. >_>
20160824 18:00:32< vultraz> but internally
20160824 18:00:40< vultraz> o_O
20160824 18:00:45< celticminstrel> Internally I don't think it matters.
20160824 18:00:47< vultraz> t'is advanced magic
20160824 18:01:25< celticminstrel> The code you gave me looks exactly the same as adding horizontal grow like you said. I'll update the screenshot.
20160824 18:01:40< celticminstrel> Just add a 2 to the filename.
20160824 18:03:03< vultraz> can't get it
20160824 18:03:16< celticminstrel> ?
20160824 18:03:59< vultraz> shot wont load
20160824 18:04:19< celticminstrel> http://celmin.pwcsite.com/wesnoth/mp_create_unbalanced2.png ?
20160824 18:04:41< tad_> worked for me
20160824 18:05:25< vultraz> still no
20160824 18:06:54< tad_> celticminstrel: can you give me a pasteup for the changes you need tested for the unstore problem. I'm at a stand-still because it's defeating the scene.
20160824 18:07:28< celticminstrel> Yeah, one sec.
20160824 18:07:56< vultraz> oh come on
20160824 18:08:05< vultraz> listboxes are designed to ONLY accept toggle buttons/panels :|
20160824 18:08:13< vultraz> ART THOU KIDDING
20160824 18:08:27< celticminstrel> Why is that a problem?
20160824 18:09:13< celticminstrel> http://celmin.pwcsite.com/wesnoth/recall_fix.diff
20160824 18:09:35< nore> Can somebody look at #757 please?
20160824 18:09:47< celticminstrel> I'll push that once I get this MP Create thing working.
20160824 18:10:08< vultraz> nore: did you test that it works?
20160824 18:10:11< tad_> celticminstrel: thanks.
20160824 18:10:11< nore> it fiwes the network commands only appearing in non-networked games
20160824 18:10:17< nore> vultraz: yes, just tested it
20160824 18:10:26< nore> s/fiwes/fixes/
20160824 18:10:27< irker813> wesnoth: Ekdohibs wesnoth:master 38698137aa3a / src/menu_events.cpp: Fix networked commands only appearing in non-networked games. https://github.com/wesnoth/wesnoth/commit/38698137aa3abfe02f73b91adb18452cecac9fcf
20160824 18:10:29< irker813> wesnoth: Charles Dang wesnoth:master d6d62f04b82a / src/menu_events.cpp: Merge pull request #757 from Ekdohibs/network-fix https://github.com/wesnoth/wesnoth/commit/d6d62f04b82a5dd6306fbf8c723cb57c1d3fd126
20160824 18:10:39< nore> that was fast
20160824 18:10:54< celticminstrel> If you want to do another server thing, you could re-enable the room manager.
20160824 18:11:49< vultraz> I thought we were gonna drop that
20160824 18:11:52-!- mjs-de [~mjs-de@x4db6df9c.dyn.telefonica.de] has joined #wesnoth-dev
20160824 18:11:59< celticminstrel> Why would we drop it?
20160824 18:12:18< vultraz> i dunno
20160824 18:12:21< vultraz> is it worth keeping?
20160824 18:12:49< celticminstrel> If the lobby every gets really active, I think it's definitely worth having.
20160824 18:13:18< vultraz> I see
20160824 18:13:47< celticminstrel> Whee, time to step through generator stuff again...
20160824 18:14:03< tad_> celticminstrel: your patch does solve the error, tyvm
20160824 18:14:32< celticminstrel> Guh? I said step in, not out! :(
20160824 18:15:34< celticminstrel> Why are some of the names in the stack trace still mangled...
20160824 18:15:45< celticminstrel> Most of them are unmangled, so what's wrong with those ones?
20160824 18:15:51< celticminstrel> Buffer overflow?
20160824 18:15:54< vultraz> celticminstrel: worth fixing: has_minimum implicitly means has_maximum = false too
20160824 18:16:09< celticminstrel> That doesn't appear to be the case?
20160824 18:16:36< celticminstrel> The generator I'm specifying has minimum policy tnone and maximum policy tone.
20160824 18:16:48< vultraz> look at the generator codepath
20160824 18:16:51< celticminstrel> Oh wait, you mean the opposite maybe?
20160824 18:16:55< vultraz> in build()
20160824 18:16:59< vultraz> it's a split path
20160824 18:17:02< celticminstrel> has_minimum=true implies has_maximum=false?
20160824 18:17:10< vultraz> no
20160824 18:17:26< vultraz> if you set has_minimum = false, one expects you to be able to select either 0 or 1 items
20160824 18:17:34< vultraz> but you can select 0 to n items
20160824 18:18:12< celticminstrel> Again, that doesn't appear to have happened here.
20160824 18:18:41< vultraz> try setting has_minimum = "false" in a listbox 
20160824 18:18:42< celticminstrel> I set has_minumum=false and didn't touch has_maximum, and it's clearly using the tone maximum policy.
20160824 18:18:45< vultraz> and go clicking items
20160824 18:18:55< celticminstrel> This is in the drop down list though.
20160824 18:19:23< vultraz> which uses a listbox
20160824 18:20:01< celticminstrel> Seems like it doesn't let you select items if there isn't one initially selected or something?
20160824 18:20:27< vultraz> it does
20160824 18:20:32< tad_> vultraz: yeah .. with cm's patch I can play through and not see those RC>1 errors. now .. on to cleaning up the recall-vs-store proof-of-concept so zookeeper can review the idea and whack me upside the head for making such a huge change to a working mainline campaign ...
20160824 18:20:34< vultraz> just that the box dismisses on a click
20160824 18:20:47< celticminstrel> No, I'm going by the code here.
20160824 18:20:51< vultraz> tad_: cm's patch.. ? I fixed the rc stuff :/
20160824 18:20:53< celticminstrel> Not by the visual effect.
20160824 18:21:03< vultraz> celticminstrel: oh
20160824 18:21:06< vultraz> right, that is true
20160824 18:21:08< vultraz> I did notice that
20160824 18:21:12< vultraz> seems like ab ug
20160824 18:21:56< tad_> vultraz: yes but the unstore problem prevented me retesting for your changes. All is good now.
20160824 18:22:03< vultraz> ah
20160824 18:26:59< tad_> celticminstrel: PR 758 has your patch. Reject and close if you have more. It's there so we don't forget is all.
20160824 18:27:28< vultraz> celticminstrel: not being able to select if none is initial selected is a bug. but it's also a bug that a visual effect appears
20160824 18:27:34< vultraz> if one cannot select
20160824 18:27:48< vultraz> celticminstrel: I think it'd be useful to have a 'cannot select any' mode
20160824 18:27:52< vultraz> (tnone?)
20160824 18:29:19-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160824 18:30:37< vultraz> tad_, celticminstrel, zookeeper: https://drive.google.com/file/d/0B-mR9s8FduLLV2toY01oeGFBMGs/view?usp=sharing
20160824 18:31:10< zookeeper> ehh why would you name them?
20160824 18:31:19< celticminstrel> That's the leader dialog?
20160824 18:31:31< celticminstrel> ^loadgame
20160824 18:31:40< celticminstrel> I don't think I can view Google drive without a browser though...
20160824 18:31:41< zookeeper> how is it at all relevant what your leader's name is?
20160824 18:32:10< vultraz> How is it at all relevant what your leader's sprite is :P
20160824 18:32:14< zookeeper> basically the current leader image mainly serves as a visual reminder about which campaign you're playing.
20160824 18:32:14< celticminstrel> vultraz: Can't select any should be easy in principle, though I suspect there's currently no max tnone policy.
20160824 18:32:43< vultraz> zookeeper: well it's a rather bad one, then
20160824 18:32:59< zookeeper> maybe
20160824 18:32:59< celticminstrel> vultraz: What buggy visual effect are you talking about though?
20160824 18:33:57< zookeeper> vultraz, it doesn't sound unreasonable to me to replace the unit sprite with the campaign icon, for instance.
20160824 18:34:03< vultraz> celticminstrel: despite the bug where you cannot make a new selection if one wasn't already selected, you can click rows and the toggle panel's 'selected' indicator shows up
20160824 18:34:05-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 18:34:18< vultraz> ie, the row looks selected
20160824 18:34:36< celticminstrel> Ah, that's probably the toggle panel's fault then.
20160824 18:34:44< vultraz> celticminstrel: the fix for that is probably to avoid setting child widget states
20160824 18:34:47< vultraz> if the click is not valid
20160824 18:35:03< celticminstrel> Something like that, I suppose.
20160824 18:35:16< vultraz> zookeeper: ehhh...
20160824 18:35:47< zookeeper> if you think about what information the player needs to be presented with when they're browsing their saves, the names and other specifics of the leader units seem rather irrelevant.
20160824 18:37:03< vultraz> i shall consider
20160824 18:37:34< zookeeper> the leader is almost always the main character and protagonist, sure, but they're far from being an embodiment of each campaign
20160824 18:38:01< tad_> I like the layout. I'm with zookeeper on not being sure it's needed but it used to show just an icon and if you're going to do it, I do like having the name with the icon like that
20160824 18:38:36-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 18:38:37< travis-ci> wesnoth/wesnoth#10545 (master - d6d62f0 : Charles Dang): The build is still failing.
20160824 18:38:37< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154833815
20160824 18:38:37-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 18:39:45< tad_> When I'm playing (and not playtesting) though, I usually just look at the top line of the file list and, maybe, a few lines down from it. But I don't jump from campaign to campaign when I'm playing. Others may and the added info might help them.
20160824 18:40:08-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 18:40:19< vultraz> gonna get some sleep, will probably commit tomorrow
20160824 18:40:24< vultraz> i hope celticminstrel fixes bugs
20160824 18:40:30< celticminstrel> Which bugs?
20160824 18:40:31< vultraz> that he's fixing
20160824 18:40:44-!- mjs-de [~mjs-de@x4db6df9c.dyn.telefonica.de] has quit [Remote host closed the connection]
20160824 18:40:55< celticminstrel> Well, I've fixed the Lua one and the dropdown one and some MP Create ones, just have the layout now.
20160824 18:41:30< tad_> So you're about half a million from being done. Think you'll finish tonight?  :)
20160824 18:41:39< celticminstrel> Me?
20160824 18:41:45 * tad_ chuckles
20160824 18:41:46< celticminstrel> Probably.
20160824 18:43:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160824 18:45:20-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160824 19:11:33< irker813> wesnoth: Celtic Minstrel wesnoth:master 2cad49162d1e / / (5 files in 3 dirs): MP Create: Fix a few issues on 800x600 layout https://github.com/wesnoth/wesnoth/commit/2cad49162d1efd1bb27bdafce26df1c4b99a52ef
20160824 19:11:35< irker813> wesnoth: Celtic Minstrel wesnoth:master 824655c67f2a / / (4 files in 4 dirs): Fix first menuitem being outlined in all menus https://github.com/wesnoth/wesnoth/commit/824655c67f2a07bc973481f553a59a078d916f05
20160824 19:11:37< irker813> wesnoth: Celtic Minstrel wesnoth:master 91ed4c31080e / src/scripting/game_lua_kernel.cpp: Lua: Fix issue with wesnoth.put_recall_unit https://github.com/wesnoth/wesnoth/commit/91ed4c31080eab6062d77d165a7cbdffcb38c7e0
20160824 19:21:55-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160824 19:23:36-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160824 19:29:29-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 19:30:38-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160824 19:32:15-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 19:33:20-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160824 19:36:49-!- travis-ci [~travis-ci@ec2-54-224-50-50.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 19:36:50< travis-ci> wesnoth/wesnoth#10547 (master - 91ed4c3 : Celtic Minstrel): The build is still failing.
20160824 19:36:50< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154849759
20160824 19:36:50-!- travis-ci [~travis-ci@ec2-54-224-50-50.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 19:57:44< vultraz> cannot sleep
20160824 19:57:55< vultraz> celticminstrel: crash when selecting Campaigns in create tiny
20160824 19:58:08< celticminstrel> Yaaay... :|
20160824 20:00:18< vultraz> can also confirm that you've fixed has_minimum = "false" implying has_maximum = "false"
20160824 20:00:29< celticminstrel> I have? Huh, okay.
20160824 20:00:36< celticminstrel> That wasn't the intent there.
20160824 20:01:23< vultraz> must've been the change to generator_private
20160824 20:01:55< celticminstrel> Obviously, but that wasn't the intent of that change, is all.
20160824 20:02:12< celticminstrel> The intent was to allow selecting an item in a tone-max list that has no initial selection.
20160824 20:02:44< celticminstrel> This naming convention is so ridiculous. I want to change it but it's just such a huge task that ugh.
20160824 20:03:17< vultraz> really?
20160824 20:03:21< vultraz> cannot just find-replace?
20160824 20:03:48< celticminstrel> Probably could, but that's still a huge task. Do you understand just how many classes would need to be renamed?
20160824 20:04:04< vultraz> yes
20160824 20:04:59-!- oldlaptop [~quassel@162.247.150.37] has quit [Ping timeout: 250 seconds]
20160824 20:05:12-!- Kwandulin [~Miranda@p200300760F35BFF8748F2530F77E8BD2.dip0.t-ipconnect.de] has quit [Quit: Kwandulin]
20160824 20:05:32< celticminstrel> game_events/manager.hpp also uses a slightly saner variation on the naming convention, I think.
20160824 20:05:45< celticminstrel> I guess game_events/ in general.
20160824 20:08:47< celticminstrel> Okay, outside of the gui dir we have t_pump, t_event_handlers, twesnothd_connection, tredirect_output_setter, tlua_jailbreak_exception, ttext, titor (related to previous), t_ref_counter; and in sdl/, timage and twindow.
20160824 20:15:41< nore> hm, I think I've just found a crash related to planning mode
20160824 20:15:54< nore> how to reproduce:
20160824 20:16:26< nore> 1. start a local game (anything will do), and create an enemy unit not far away from leader, but not next to it
20160824 20:16:40< nore> 2. enable planning mode, and plan to attack that unit with leader
20160824 20:16:55< nore> 3. disable planning mode and attack that unit with leader
20160824 20:16:59< nore> 4. crash.
20160824 20:17:02< celticminstrel> What...
20160824 20:17:22< nore> error message: wesnoth/src/whiteboard/utility.cpp:66: unit_const_ptr wb::find_backup_leader(const unit&): Assertion `resources::gameboard->map().is_keep(leader.get_location())' failed.
20160824 20:18:11< nore> I've not yet tried to investigate further
20160824 20:18:15< celticminstrel> Ugh, assertions.
20160824 20:18:41< celticminstrel> I feel that Wesnoth's use of assertions is frequently just a sign of lazy coding.
20160824 20:19:10< celticminstrel> They really should only be used on conditions that you absolutely can't imagine ever being false...
20160824 20:20:21< vultraz> I think I've found out why leader images often don't appear
20160824 20:20:23< vultraz> :D
20160824 20:20:29< vultraz> (in Load Game)
20160824 20:21:48< vultraz> bug in filesystem::get_independent_image_path
20160824 20:22:17< vultraz> if provided a campaign-specific image is derps out and fails
20160824 20:25:58-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160824 20:27:01-!- mjs-de [~mjs-de@x4db6df9c.dyn.telefonica.de] has joined #wesnoth-dev
20160824 20:28:19 * vultraz adds some debugging
20160824 20:29:07-!- mjs-de [~mjs-de@x4db6df9c.dyn.telefonica.de] has quit [Remote host closed the connection]
20160824 20:30:05-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160824 20:33:26< vultraz> this is rather odd
20160824 20:33:50< celticminstrel> Yay oddness?
20160824 20:34:51< vultraz> I can't really understand the code :/
20160824 20:35:02< vultraz> maybe you can take a peek?
20160824 20:35:08< celticminstrel> Maybe!
20160824 20:35:12< vultraz> filesystem_boost.cpp:1192
20160824 20:35:20-!- oldlaptop [~quassel@162.247.150.37] has joined #wesnoth-dev
20160824 20:42:32-!- Greg-Boggs [~greg_bogg@74.92.168.145] has joined #wesnoth-dev
20160824 20:42:49< vultraz> hmm
20160824 20:42:55< vultraz> no, looks like the problem is in get_binary_file_location
20160824 20:42:59< vultraz> line 1074
20160824 20:43:24-!- JyrkiVesterinen [~JyrkiVest@87-100-204-68.bb.dnainternet.fi] has quit [Quit: .]
20160824 20:43:34< vultraz> wat
20160824 20:43:36< vultraz> "using 'for' instead 'if' to allow putting delcaration and check into the brackets"
20160824 20:44:20< vultraz> ah, I think I see
20160824 20:44:25< celticminstrel> I will look in a minute, just waiting for XCode to finish linking.
20160824 20:45:21< vultraz> keep in mind if you actually test anything you keep having to delete your save_index file
20160824 20:45:51< celticminstrel> I thought I was looking with the intent of helping you understand the code.
20160824 20:46:11< vultraz> yes
20160824 20:47:44-!- Greg-Boggs [~greg_bogg@74.92.168.145] has quit [Remote host closed the connection]
20160824 20:47:44< vultraz> I think the problem is it doesn't actually check in data/campaigns/...
20160824 20:48:01< celticminstrel> [binary_path]?
20160824 20:48:18< celticminstrel> I bet that's the problem - binary paths for the campaign are not set up at this point.
20160824 20:49:11-!- Greg-Boggs [~greg_bogg@74.92.168.145] has joined #wesnoth-dev
20160824 20:50:17< vultraz> hmmmm
20160824 20:50:37< vultraz> you're right
20160824 20:50:42< vultraz> they're inside ifdefs
20160824 20:50:45< celticminstrel> And there's no way they could be without reparsing the whole config every time you select a new saved game, pretty much.
20160824 20:50:58< celticminstrel> So maybe the save index needs to store the path in a binary_path-independent way.
20160824 20:51:30< celticminstrel> ie, store the resolved image path instead of the short image path.
20160824 20:51:31< vultraz> moving the binary path outside the ifdef fixes the problem
20160824 20:51:44< celticminstrel> But don't commit that.
20160824 20:52:31< vultraz> ok, if we don't want the image path run through get_independent_image_path
20160824 20:52:35< vultraz> what should it go through?
20160824 20:52:51< vultraz> the docs indicate get_independent_image_path was supposed to do as you say :/
20160824 20:53:05< vultraz> " Returns an image path to @a filename for binary path-independent use in saved games."
20160824 20:53:13< vultraz> so maybe it's instead the implementation that's broken
20160824 20:53:28-!- Greg-Boggs [~greg_bogg@74.92.168.145] has quit [Ping timeout: 244 seconds]
20160824 20:53:43< celticminstrel> From the code it looks like that is, in fact, exactly what it's doing.
20160824 20:54:08< vultraz> but it only looks in binary_paths
20160824 20:54:26< vultraz> no wonder leader images sometimes show up once you exit a campaign
20160824 20:54:29< celticminstrel> Is get_independent_image_path when creating the save index or parsing it?
20160824 20:54:39< celticminstrel> ^called when
20160824 20:54:43< vultraz> creating
20160824 20:55:01< vultraz> save_index is added to whenever accessing a new save for the first time
20160824 20:55:03< celticminstrel> So maybe the problem is actually when parsing it.
20160824 20:55:32< vultraz> not..really?
20160824 20:55:52< celticminstrel> Well, what's called to resolve the image path taken from the save index?
20160824 20:56:05< vultraz> get_independent_image_path
20160824 20:56:12< vultraz> which looks at all available binary_paths
20160824 20:56:21< celticminstrel> Huh? I thought you said that was called when adding to the save index.
20160824 20:56:31< vultraz> oh right
20160824 20:56:34< vultraz> misread
20160824 20:56:36< vultraz> nothing is
20160824 20:56:48< vultraz> it's accessed directly
20160824 20:56:57< celticminstrel> So it just goes through whatever standard resolution process there is?
20160824 20:57:07< vultraz> yes
20160824 20:57:14< vultraz> at least in the case of the load game dialog
20160824 20:57:17< celticminstrel> Wouldn't that be the problem, then?
20160824 20:57:29< vultraz> but what should it be parsed with?
20160824 20:58:13< celticminstrel> I suppose at the moment it's being passed through get_binary_file_location("image", name)?
20160824 20:58:31< vultraz> when the save_index entry is created, yes
20160824 20:58:46< celticminstrel> I mean when the save index is loaded.
20160824 20:59:03< vultraz> when it's loaded nothing happens
20160824 20:59:06< celticminstrel> Set a breakpoint there and see if it hits with a leader image?
20160824 20:59:16< vultraz> what?
20160824 20:59:42< celticminstrel> The save index is loaded when you view the load game dialog, I guess, right?
20160824 20:59:49< vultraz> accessed
20160824 20:59:58< vultraz> but that's not.. really an issue.
20160824 21:00:06< celticminstrel> Huh?
20160824 21:00:35< vultraz> the issue is that running the image path taken from the savefile when creating the save_index through get_independent_image_path will give an empty string if the necessary binary path isn't available 
20160824 21:00:58< celticminstrel> Oh, so the save index may not be created while the correct campaign is in scope?
20160824 21:01:06< vultraz> exactly
20160824 21:02:00< vultraz> and i assume it's not good to move all binary_path tags outside their ifdefs?
20160824 21:02:05< celticminstrel> So, if you ensure the save index is created while the campaign is in scope, then the image does show up correctly?
20160824 21:02:10< celticminstrel> It's not good, no.
20160824 21:02:12-!- Greg-Boggs [~greg_bogg@74.92.168.145] has joined #wesnoth-dev
20160824 21:03:11< vultraz> the answer is yes
20160824 21:03:34< vultraz> so we have an odd problem :|
20160824 21:04:01< vultraz> what do you advise?
20160824 21:06:26< celticminstrel> Well, I can think of two ideas, but... wouldn't the save index entry normally be created when the game is saved, and thus the scope is correct?
20160824 21:06:37< zookeeper> from looking at the backlog, it sure sounds like get_independent_image_path isn't working correctly
20160824 21:07:20< zookeeper> that it "only looks in binary_paths" should be a non-issue because when the save is created, the corresponding binary_path will have been set
20160824 21:07:21< vultraz> one would think
20160824 21:07:37< vultraz> let us test that theory
20160824 21:09:25< vultraz> ot, but im noticing massive lag when opening the campaign menu..
20160824 21:09:27< vultraz> weird
20160824 21:10:09< vultraz> well now the image isn't loading at all :/
20160824 21:11:54< vultraz> odd
20160824 21:12:04< vultraz> the path is correct and it was working a minute ago :/
20160824 21:12:47< vultraz> but the output in the save_index is correct..
20160824 21:12:53< vultraz> full path
20160824 21:12:57< vultraz> perhaps zookeeper is right
20160824 21:13:09< vultraz> but...
20160824 21:27:26-!- Greg-Bog_ [~greg_bogg@74.92.168.145] has joined #wesnoth-dev
20160824 21:29:51-!- Greg-Bog_ [~greg_bogg@74.92.168.145] has quit [Remote host closed the connection]
20160824 21:30:21-!- Greg-Boggs [~greg_bogg@74.92.168.145] has quit [Ping timeout: 250 seconds]
20160824 21:39:27-!- Samual [~Samual@xonotic/core-team/Samual] has quit [Ping timeout: 250 seconds]
20160824 21:44:30-!- Appleman1234 [~Appleman1@KD036012023035.au-net.ne.jp] has quit [Ping timeout: 250 seconds]
20160824 21:45:37-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20160824 21:47:56-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160824 21:48:18-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 276 seconds]
20160824 21:49:05-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has quit [Client Quit]
20160824 22:00:46-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 22:01:03-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160824 22:01:10-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 22:04:22-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160824 22:06:05< celticminstrel> Okay, so the player slider in MP Create causes a crash.
20160824 22:06:31< tad_> Is there a way, in Lua or WML, to ask the map what the starting position for a side? For example, with a no_leader=yes side can I recall/recruit a leader into the keep without having to hard-code the (x,y) position?
20160824 22:06:55< celticminstrel> Good question!
20160824 22:07:33< celticminstrel> I think there is a way to get starting positions defined in the map.
20160824 22:07:55< celticminstrel> But no way to get the leader's actual initial position if it was specified as x,y in [side].
20160824 22:08:51< tad_> I'm in a prestart event and am creating a leader, if not already present. If present, [side] auto-recalls him to the keep. But if not, I need to figure out where he should have gone.
20160824 22:10:40< tad_> I think this issue is the reason S05 changes leaders for S06 .. so it's correct when [side] runs in S06. But it's a pain if I'm coming at it from choose_level other than from S05. There are other conditions where it would be nice, but that's the first of the top of my head.
20160824 22:11:37< celticminstrel> Are the starting locations defined in the map?
20160824 22:12:06< tad_> Yes. Map editor supports it and most mainline campaigns use it.
20160824 22:12:15-!- irker813 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160824 22:12:26< celticminstrel> Just give me a moment to look it up or something.
20160824 22:12:40< celticminstrel> It might not be documented yet on the wiki.
20160824 22:12:48< zookeeper> [store_starting_locations]?
20160824 22:12:54< zookeeper> or whatever it was
20160824 22:12:56< celticminstrel> Wait, that exists?
20160824 22:13:07< zookeeper> [store_starting_location]
20160824 22:13:14< zookeeper> documented exactly where you'd expect
20160824 22:13:27< celticminstrel> That's not what I was thinking of though.
20160824 22:13:34< celticminstrel> Obviously it would work.
20160824 22:13:47< tad_> Well, if you're looking at unstore_unit or recall it's not where you'd expect.
20160824 22:14:17-!- irker098 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160824 22:14:17< irker098> wesnoth: Celtic Minstrel wesnoth:master dfb6e865ce08 / src/gui/dialogs/lobby/lobby.cpp: MP Lobby: Allow plugins to select games by ID https://github.com/wesnoth/wesnoth/commit/dfb6e865ce08bbb54700c58de85b1583239dd4cb
20160824 22:14:17< irker098> wesnoth: Celtic Minstrel wesnoth:master 7084eee8a76b / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Fixup XCode project for PR 744 https://github.com/wesnoth/wesnoth/commit/7084eee8a76b1b477a4e18b858c940edbc4439de
20160824 22:14:17< irker098> wesnoth: Celtic Minstrel wesnoth:master a0350198c348 / / (7 files in 4 dirs): MP Lobby / Create: Yield to plugins https://github.com/wesnoth/wesnoth/commit/a0350198c3487236adff9325b1f5ec91080f0131
20160824 22:14:18< irker098> wesnoth: Celtic Minstrel wesnoth:master 970c6ee914d1 / src/gui/dialogs/multiplayer/mp_create_game.cpp: Fix assertion failure in MP Create https://github.com/wesnoth/wesnoth/commit/970c6ee914d131d8712bd6db63ccb1289e28fda6
20160824 22:14:19< irker098> wesnoth: Celtic Minstrel wesnoth:master db9678997cd2 / src/gui/dialogs/multiplayer/mp_create_game.cpp: MP Create: Set up configure plugins context https://github.com/wesnoth/wesnoth/commit/db9678997cd251e6e27398a3eb278f86ea42c2d9
20160824 22:14:21< irker098> wesnoth: Celtic Minstrel wesnoth:master 086c7f5b4c51 / src/gui/dialogs/multiplayer/ (mp_create_game.cpp mp_create_game.hpp): MP Create: Add more methods to plugins contexts https://github.com/wesnoth/wesnoth/commit/086c7f5b4c513c3f8251baa3ff6fc0eac723e9b1
20160824 22:14:29< zookeeper> well, i meant more like exactly where you'd expect _after_ knowing what the tag is called :p
20160824 22:14:38< tad_> But zookeeper is right .. it's just what I needed.
20160824 22:15:18< celticminstrel> Apparently the Lua is wesnoth.get_starting_location.
20160824 22:15:40< tad_> So much of the Wiki is written for an expert who needs a reminder and not for a noob who has no idea if it's even there.
20160824 22:15:45-!- fabi [~fabi@176.0.91.28] has joined #wesnoth-dev
20160824 22:15:52-!- fabi_ [~fabi@176.4.125.129] has quit [Read error: Connection reset by peer]
20160824 22:16:02< celticminstrel> Well, there's that massive sidebar on all the WML pages,
20160824 22:16:13< celticminstrel> Though it's kinda dense, I suppose.
20160824 22:16:19< celticminstrel> Not good for scanning.
20160824 22:16:28< zookeeper> tad_, well that tends to apply to any language's reference documentation...
20160824 22:16:33< tad_> Which is itty-bitty, too. I try to keep the page zoomed so it's even readable.
20160824 22:18:06< Aginor> mattc: I think it sounds like an issue with when hotkeys are processed in lua, it sounds like both key down and key up...
20160824 22:18:07< celticminstrel> wesnoth.special_locations[side_num] also works. Might deprecate get_starting_location in favour of that.
20160824 22:19:54-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160824 22:26:07-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160824 22:27:26-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 22:29:39-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 22:33:29-!- celmin [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160824 22:34:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160824 22:37:12-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20160824 22:38:40-!- travis-ci [~travis-ci@ec2-54-158-184-207.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 22:38:41< travis-ci> wesnoth/wesnoth#10548 (master - 086c7f5 : Celtic Minstrel): The build is still failing.
20160824 22:38:41< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154890878
20160824 22:38:41-!- travis-ci [~travis-ci@ec2-54-158-184-207.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 22:40:18-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160824 22:40:49-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Client Quit]
20160824 22:41:09-!- Appleman1234 [~Appleman1@KD036012026220.au-net.ne.jp] has joined #wesnoth-dev
20160824 22:49:27-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160824 22:53:05-!- Shiki [~Shiki@141.39.226.227] has quit [Quit: Verlassend]
20160824 23:00:52< irker098> wesnoth: Celtic Minstrel wesnoth:master 9dbae3944839 / src/ (5 files in 3 dirs): Minor code cleanup regarding special map locations https://github.com/wesnoth/wesnoth/commit/9dbae3944839f2e55086f0e7c221ef7083c6be06
20160824 23:00:54< irker098> wesnoth: Celtic Minstrel wesnoth:master be456aa1c35e / src/pathfind/pathfind.cpp: Attempt to fix Travis by removing custom lower_bound implementation https://github.com/wesnoth/wesnoth/commit/be456aa1c35efadf7fbf11a68b6293eafee2ab9d
20160824 23:00:58< celmin> vultraz: I think the plugins context might be ready for the MP test now, would you like to try it again?
20160824 23:01:19< vultraz> did you commit something?
20160824 23:01:30< celmin> Yes.
20160824 23:07:03-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 240 seconds]
20160824 23:07:20< vultraz> damn codeblocks won't start blah
20160824 23:07:35< celmin> Speaking of codeblocks, there's a new header to add.
20160824 23:07:59< vultraz> noted
20160824 23:08:05< vultraz> thanks for moving that to its proper place
20160824 23:08:23-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160824 23:08:27< celmin> Well, not sure it's really the proper place for the file, but whatever, I guess.
20160824 23:09:03< vultraz> better than what I had :P
20160824 23:09:14< celmin> It also Does Something now.
20160824 23:09:20< celmin> Rather than just being a container.
20160824 23:12:35< vultraz> it'd be nice if c++ allowed had local functions like lua does
20160824 23:12:42< vultraz> so one didn't have to use macros
20160824 23:13:26< celmin> Uhh, lambdas.
20160824 23:13:42< celmin> (But they won't work for this, because of all those #)
20160824 23:13:59< vultraz> yeah, I meant in this case
20160824 23:14:24< vultraz> wait
20160824 23:14:28< vultraz> ... 
20160824 23:14:33< celmin> I guess this could be done with a function if it was Lua, given differences in how Lua looks up names.
20160824 23:14:44< celmin> Something wrong?
20160824 23:14:50< vultraz> nah
20160824 23:14:56< vultraz> just a thought
20160824 23:14:59< celmin> What was it?
20160824 23:15:04< celmin> Something interesting? :P
20160824 23:15:29< vultraz> can you define a local function using a lambda outside of the context of a function argument or something?
20160824 23:15:33< vultraz> like a variable?
20160824 23:15:35< vultraz> and call it?
20160824 23:15:53< celmin> You can do something like "auto name = [](){…};"
20160824 23:15:57< celmin> In any scope you want.
20160824 23:16:05< vultraz> ... I did not know this :|
20160824 23:16:08< celmin> Though doing it at global scope would be very silly.
20160824 23:16:29< vultraz> I mean it seems obvious, but I didn't think of it
20160824 23:16:50< celmin> I think Wesnoth even does it already in a few places.
20160824 23:16:54< vultraz> (I assume it doesn't necessarily need to be auto?)
20160824 23:17:13< celmin> It kinda does.
20160824 23:17:18< celmin> You could use std::function.
20160824 23:17:22< vultraz> oh
20160824 23:17:27< celmin> Some people say that's less efficient.
20160824 23:17:36< celmin> If the square brackets are empty, you can also use a function pointer type.
20160824 23:17:37< vultraz> but you can still specify the return type with ->, right?
20160824 23:17:44< celmin> Yeah, of course.
20160824 23:17:57< vultraz> hm
20160824 23:18:00< vultraz> so..
20160824 23:18:17< celmin> Each lambda with a capture is implicitly a unique type, so there's no way for you to specify the exact type.
20160824 23:18:33< celmin> ^with a closure
20160824 23:18:42< vultraz> is one able to do this: auto foo = [](auto& bar) { ... }; as opposed to template<typename T> foo(T& bar) { ... };?
20160824 23:18:52< celmin> Not until C++14.
20160824 23:19:06< celmin> Generic lambdas are not available in C++11.
20160824 23:19:17< vultraz> right
20160824 23:19:20< vultraz> but in theory
20160824 23:19:26< vultraz> can generic lambdas take the place of templates
20160824 23:19:50< celmin> That might even need C++17 if you want to assign it to a variable, but should work in C++14 as an argument to a template function, I think...
20160824 23:20:00< celmin> I think generic lambdas are basically templates, yes.
20160824 23:20:06< celmin> As far as I know.
20160824 23:21:39< vultraz> c++17? D:
20160824 23:21:44< vultraz> we won't be using that any time soon
20160824 23:21:48< vultraz> it's not even out
20160824 23:22:17< celmin> Well, to assign that lambda to a variable, I think you need template variables, and if I recall correctly (which I may not), that's C++17.
20160824 23:22:53-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160824 23:22:55< celmin> Anyway, we're still on C++11, so no template lambdas here. :P
20160824 23:23:07< celmin> tad_: I just implemented [filter]type_tree=
20160824 23:23:41< vultraz> variable templates or template variables?
20160824 23:23:42< celmin> (Wouldn't recommend using for [role] though since it wouldn't have the same special handling as type does, unless I add that separately.)
20160824 23:23:54< celmin> vultraz: Template variables. What does "variable template" even mean?
20160824 23:24:02< vultraz> http://en.cppreference.com/w/cpp/language/variable_template
20160824 23:24:10< celmin> Hmm.
20160824 23:24:26< celmin> Okay, that is indeed what I was talking about.
20160824 23:24:34< vultraz> it's c++14
20160824 23:24:39< celmin> Okay, so I was wrong.
20160824 23:24:54< vultraz> which we're not too far away from being able to use
20160824 23:25:15< celmin> We can't use C++14 until we can drop MSVC 2013, so probably not as close as you might think.
20160824 23:25:24< tad_> celmin: Philosophy question: in [modify_unit] as specified, if I give [filter_recall] it adds but does not remove. I have a quick-fix to special-case [filter_recall] to remove all before adding. Or should I figure out a more general approach which works for any top-level subtag? 
20160824 23:25:52< celmin> If push comes to shove I could get XCode to compile with the latest clang, though it might be a pain and not quite work as well as with the older version.
20160824 23:26:19< celmin> tad_: Basically it's impossible to remove anything with [modify_unit].
20160824 23:27:05< tad_> celmin: correct. So I added a special-case in the Lua to remove the entire [filter_recall] if [filter_recall] appears as a top-level subtag.
20160824 23:27:07< celmin> I actually kinda want to rewrite [modify_unit] to not rely on storing and unstoring the unit.
20160824 23:27:11< celmin> I see.
20160824 23:27:44< tad_> celmin: makes it much cleaner than having to store, replace, unstore
20160824 23:28:17< celmin> Yes and no. It also makes it less generic and higher-maintenance, unless I'm missing something.
20160824 23:28:33< tad_> celmin: BUT it will mean it depends upon the internal store/unstore because [filter_recall] is only set on unit instance create
20160824 23:28:50< celmin> As far as I know, it would mean manually handling every possible key or subtag,
20160824 23:28:52< celmin> What>
20160824 23:28:54< celmin> ^?
20160824 23:29:48< celmin> I guess it would have to still store in order to get $this_unit, but it wouldn't need to unstore.
20160824 23:30:07< tad_> celmin: in the c++, as I read it, the [filter_recall] variable is read-only and only set when the constructing the instance. 
20160824 23:30:20< celmin> Was it read-only? Well, if so, that's easily changed.
20160824 23:31:03< celmin> Oh, BTW vultraz: I got a crash in MP Create while fiddling with the players slider.
20160824 23:33:14< tad_> [modify_unit] is much cleaner for my use than [store_unit]/[set_variable]/[unstore_unit] but, at present, there is no way to remove parts of a sub-tag .. only merge in new ones. And, as I understood the C++, to have [filter_recall] take effect it MUST be there for instance constructor and there is no method to change it. But I might be misunderstanding how the C++ works for sub-tabs like filter_recall.
20160824 23:33:51< celmin> It seems like Travis isn't even trying to build my latest push… :|
20160824 23:33:56< celmin> Can anyone do anything about that?
20160824 23:34:36< celmin> Wait, maybe it is?
20160824 23:34:50< celmin> There, found it.
20160824 23:34:57< celmin> One job already failed, sigh.
20160824 23:36:14< celmin> tad_: I can't remember if there are Lua helper functions to remove things from configs?
20160824 23:36:19< tad_> celmin: Oh, I will have to look at [filter]type_tree ... 
20160824 23:36:32< celmin> If it's a parsed config, you can of course use the normal table.whatever function.
20160824 23:36:54< celmin> Once I push it. :)
20160824 23:37:35-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has joined #wesnoth-dev
20160824 23:37:36< travis-ci> wesnoth/wesnoth#10549 (master - be456aa : Celtic Minstrel): The build is still failing.
20160824 23:37:36< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154899527
20160824 23:37:36-!- travis-ci [~travis-ci@ec2-23-22-203-102.compute-1.amazonaws.com] has left #wesnoth-dev []
20160824 23:37:39< tad_> celmin: I didn't see one so I used the wesnoth.set_variable and set filter_recall to {} and that solved it. But it's a special-case for my use case and seems like a more general case may be desirable but not sure if it's really required.
20160824 23:38:42< celmin> Ah, set_variable would certainly work in the case of [modify_unit], yeah.
20160824 23:38:54< celmin> I was thinking of generic config manipulation functions, not variables.
20160824 23:39:22< celmin> Hmm...
20160824 23:39:31< celmin> I think a general-use case is desirable...
20160824 23:39:45< celmin> I can think of a few possibilities.
20160824 23:39:46< tad_> Looks to me like the Lua for modify_unit is working on the [unit] instance totally with set_variable and never as a config/table 
20160824 23:40:06-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160824 23:40:08< celmin> Yeah, it stores it and works with it as a WML variable, then unstores it afterwards.
20160824 23:41:31< celmin> 1. There could be special tags/attributes for removing stuff, which are checked at all levels - [remove_tag]name=filter_recall, for example, or remove_attributes=list,of,keys,to,be,removed,at,this,level.
20160824 23:41:49< tad_> celmin: OK. Proposal: add a top-level attribute to [modify_unit] 'replace_subtags=yes|no' default 'no' when 'yes' and top-level sub-tag which appears in [modify_unit] causes that tag to go to {} before merging the value from [modify_unit].
20160824 23:42:04< celmin> 2. [modify_unit] could check for certain "magic attributes" in the tag, such as "_remove=yes".
20160824 23:42:12< celmin> Hmm.
20160824 23:42:34-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160824 23:42:37 * celmin looks up [set_variables]
20160824 23:43:21-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 276 seconds]
20160824 23:43:21-!- wedge010 is now known as wedge009
20160824 23:43:35< celmin> If you think of it from the perspective of [set_variables]mode, I guess the current behaviour is mode=merge, right?
20160824 23:43:41< celmin> And you want mode=replace.
20160824 23:44:22< tad_> Yes. That sounds good. If you want to do both, no problem doing two [modify_unit] calls, right?
20160824 23:44:29< celmin> Hmm.
20160824 23:44:37< celmin> Well, that's certainly a possibility.
20160824 23:44:56< celmin> I'm thinking a major use-case of this would be manipulating unit attacks, though.
20160824 23:46:17< celmin> I guess mode=insert or mode=replace make sense there. Or mode=append to add a new one.
20160824 23:46:38< celmin> The only problem would be if the attack has specials.
20160824 23:46:55-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20160824 23:46:57< celmin> Maybe that wouldn't be a problem actually.
20160824 23:47:34< celmin> Well, anyway. If you'd like to implement something like "mode=replace|merge" (or maybe edit_mode or something?), that seems fine, I guess.
20160824 23:47:48< celmin> If you want to implement append and insert as well, go ahead.
20160824 23:47:59< celmin> Not sure how you'd do insert without knowing the index though.
20160824 23:49:55< tad_> I think mode=replace|merge work well. default would be merge to prevent breakage. And if you want both, do it twice, no problem. 'delete' for a subtag seems like 'replace with {}' anyway.
20160824 23:50:04< celmin> (But since it's replace that you specifically want, I'd be fine with you implementing that only. The others could be added later if someone else felt like it.)
20160824 23:50:16-!- Shiki [~Shiki@141.39.226.227] has joined #wesnoth-dev
20160824 23:52:14< tad_> ok. I'll take a look at adding a top-level attribute 'mode' (not a [unit] key) and if it's 'replace' any sub-tags are cleared before merging. Not sure if 'replace' on an attribute makes any difference. 
20160824 23:52:28< celmin> It doesn't.
20160824 23:52:48< tad_> Are there any [unit] takes which it makes a different if it appears and is empty (or "") vs not appears at all?
20160824 23:53:05< celmin> I have no idea.
20160824 23:53:31< celmin> You're asking about unit subtags?
20160824 23:54:11< tad_> I'm asking about top-level attributes.  "experience" for (a bad) example/
20160824 23:54:20< celmin> Ah.
20160824 23:54:52< celmin> I'm still not sure, but it gives warnings about unknown attributes, which would happen even if they're empty.
20160824 23:54:59< tad_> That's a "delete" operation. Not even sure if it's possible. Not sure it's ever neede.d
20160824 23:55:09< celmin> Oh, right.
20160824 23:55:31< celmin> Yeah, I don't know of any top-level unit tags for which delete would be different from setting to empty.
20160824 23:55:47< celmin> There could be one or two I haven't thought of, I suppose.
20160824 23:55:54-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160824 23:56:22< tad_> Well, if it breaks, we'll have to fix it. Same if we missed something. Just like always. :P
20160824 23:57:06< tad_> OK. I'm gonna go work on a change to [modify
20160824 23:57:21< tad_> _unit] and will PR it when I think it's ready.
20160824 23:57:26< celmin> Okay.
20160824 23:57:30-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
--- Log closed Thu Aug 25 00:00:52 2016