--- Log opened Tue Aug 23 00:00:41 2016
20160823 00:01:34-!- travis-ci [~travis-ci@ec2-54-234-211-210.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 00:01:35< travis-ci> wesnoth/wesnoth#10512 (master - db167c0 : Charles Dang): The build is still failing.
20160823 00:01:35< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154308344
20160823 00:01:35-!- travis-ci [~travis-ci@ec2-54-234-211-210.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 00:03:52-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has quit [Ping timeout: 240 seconds]
20160823 00:17:59< celticminstrel> vultraz: Any objections to a line between the static and dynamic parts of the right panel?
20160823 00:18:07< celticminstrel> ie, just under the row of checkboxes.
20160823 00:19:50< vultraz> I guess not
20160823 00:20:21< irker616> wesnoth: Celtic Minstrel wesnoth:master b79a237f3508 / data/gui/window/ (mp_create_game.cfg mp_create_game/default_resolution.cfg): MP Create: Move WML around for easier definition of low-res versions https://github.com/wesnoth/wesnoth/commit/b79a237f3508effb3ffb73a25cd0222a8c2ed971
20160823 00:20:23< irker616> wesnoth: Celtic Minstrel wesnoth:master d604246a1a8f / data/gui/window/mp_create_game/default_resolution.cfg: MP Create: Add spacer line above tabbed section https://github.com/wesnoth/wesnoth/commit/d604246a1a8ffc1883beabd58a919ba9a6116995
20160823 00:20:45< celticminstrel> That commit might look big, but it's essentially zero-sum - ie, it has no effect on the dialog.
20160823 00:21:48< vultraz> that line is actually a good idea
20160823 00:24:38< vultraz> might remove the vertical lin
20160823 00:24:39< vultraz> e
20160823 00:24:52< celticminstrel> I think it's fine with both.
20160823 00:25:19-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 00:25:20< travis-ci> wesnoth/wesnoth#10513 (master - 396f5df : Celtic Minstrel): The build is still failing.
20160823 00:25:20< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154313141
20160823 00:25:20-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 00:26:10< vultraz> yeah, it does seem rather squashed without it
20160823 00:28:56< vultraz> ya know, I'm quite proud of this dialog
20160823 00:37:12< celticminstrel> Okay, so, I tried to rearrange it so that it has a top panel and bottom panel instead of left panel and right panel.
20160823 00:37:22< celticminstrel> However, the listbox is now taking up more than the height of the window.
20160823 00:38:00< vultraz> you have just stumbled upon GUI2 Listbox Bug #490272027 :P
20160823 00:38:11< vultraz> (that's not a real bug report, FYI)
20160823 00:39:05< vultraz> anyway, I know that in the case of two listboxes stacked on top of each other, gui2 just gives up and dedicates maximum space to them
20160823 00:39:20< celticminstrel> But there's just one listbox here?
20160823 00:39:28< celticminstrel> I guess there's also the horizontal one.
20160823 00:39:39< celticminstrel> ...oh right, there's a tree view too.
20160823 00:42:59< Aginor> sounds like someone should consider fixing that piece of code :)
20160823 00:43:05< Aginor> for the layouting
20160823 00:43:52< celticminstrel> It's pretty bad for it to be assigned about three times the height of the window, though.
20160823 00:51:54< vultraz> Aginor: one thing at a time :P
20160823 00:57:18-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 00:57:19< travis-ci> wesnoth/wesnoth#10514 (master - 14752e1 : Charles Dang): The build has errored.
20160823 00:57:19< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154314502
20160823 00:57:19-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 00:57:28< shadowm> vultraz: <shadowm> vultraz: You wouldn't happen to have an image comparison of before and after I fixed that warning, would you?
20160823 01:01:46< vultraz> I did a comparison, and they look surprisingly similar.
20160823 01:01:54< vultraz> Gimp's screwing with me so I can't make a side-by-side
20160823 01:02:40< vultraz> there are small, subtle differences, though
20160823 01:04:05< vultraz> before: https://www.dropbox.com/s/botbutolnj1ott9/wose%20before.PNG?dl=0
20160823 01:04:13< vultraz> after: https://www.dropbox.com/s/s7c6ao91h1vys42/wose%20after.PNG?dl=0
20160823 01:17:40< shadowm> Looks like it was better before. :\
20160823 01:23:29< shadowm> Or maybe not.
20160823 01:23:36< shadowm> I can't decide.
20160823 01:23:43< shadowm> Stupid fancy scaling algorithms.
20160823 01:26:16< celticminstrel> The wose probably isn't the greatest sample though.
20160823 01:28:48-!- travis-ci [~travis-ci@ec2-54-234-211-210.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 01:28:49< travis-ci> wesnoth/wesnoth#10515 (master - d604246 : Celtic Minstrel): The build has errored.
20160823 01:28:49< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154318957
20160823 01:28:49-!- travis-ci [~travis-ci@ec2-54-234-211-210.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 01:31:29< shadowm> Oh well who cares. It's not like the person who brought this in is going to return and decide that the non-deterministic version of the algorithm was better.
20160823 01:31:41< shadowm> *better/intended
20160823 01:37:36-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:b01d:7dbd:5d82:c3ee] has joined #wesnoth-dev
20160823 01:49:09< celticminstrel> Whoa! Finally got a layout that works.
20160823 01:49:29-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:b01d:7dbd:5d82:c3ee] has quit [Remote host closed the connection]
20160823 01:49:36< celticminstrel> Makes the game selection box pretty small, and long descriptions require scrolling, but...
20160823 01:50:02< vultraz> but does it look good
20160823 01:50:10< celticminstrel> Sorta?
20160823 01:50:17< celticminstrel> It doesn't look bad, at least.
20160823 01:50:23< vultraz> well, it's the small res version
20160823 01:50:26< vultraz> doesn't need to look good
20160823 01:50:37< celticminstrel> Oh, I should add a vertical line.
20160823 01:51:11< vultraz> gah, snprintf
20160823 01:51:13< vultraz> shoo shoo
20160823 01:51:19< celticminstrel> What? Where? Why?
20160823 01:51:30< celticminstrel> At least it's not sprintf.
20160823 01:51:41< shadowm> If it doesn't look good just add buzzwords like "responsive", "modern", "lean" and "material" to it and people will immediately act as if it does.
20160823 01:51:54< celticminstrel> Pfft.
20160823 01:52:29< vultraz> https://github.com/wesnoth/wesnoth/blob/master/src/units/drawer.cpp#L217
20160823 01:52:31< celticminstrel> Unfortunately, it still won't work for 800x600.
20160823 01:52:31< vultraz> :|
20160823 01:52:58< shadowm> Nothing inherently wrong about snprintf().
20160823 01:53:08< celticminstrel> Sorta works for 832x624.
20160823 01:53:08< shadowm> It's how you use it that makes the difference.
20160823 01:53:59< shadowm> Personally, regardless of the type safety argument, I find stream operations way more awkward than visibly conjointed placeholders.
20160823 01:54:34< vultraz> It would indeed be nice if C++ had a standard library equivalent of Lua's string.format.
20160823 01:54:47< shadowm> >Lua's
20160823 01:54:51< vultraz> or I guess in this case, snprintf :|
20160823 01:55:04< shadowm> string.format is inspired by the printf family of functions from C, yes.
20160823 01:55:26< shadowm> Boost.Format provides Boost's own... guess what.
20160823 01:55:56< vultraz> printf implementation?
20160823 01:56:04< shadowm> I used it ages ago, wasn't too bad but it's compile-time overhead was around Boost.Foreach's IIRC.
20160823 01:56:12< irker616> wesnoth: Celtic Minstrel wesnoth:master 23017d2f72dc / data/gui/window/ (mp_create_game.cfg mp_create_game/default_resolution.cfg): MP Create: More WML shifting to ease impementation of low-res layout https://github.com/wesnoth/wesnoth/commit/23017d2f72dcf253036c278217ef4a8fb2365ff1
20160823 01:56:14< irker616> wesnoth: Celtic Minstrel wesnoth:master 2c04ee6da695 / data/gui/window/ (mp_create_game.cfg mp_create_game/default_resolution.cfg): MP Create: Another batch of WML shifting to ease implementation of low-res layou https://github.com/wesnoth/wesnoth/commit/2c04ee6da6953dd098cf8551c1b0224d920ee710
20160823 01:56:16< irker616> wesnoth: Celtic Minstrel wesnoth:master f3962498005a / data/gui/window/ (mp_create_game.cfg mp_create_game/small_resolution.cfg): MP Create: New layout for lower resolutions https://github.com/wesnoth/wesnoth/commit/f3962498005ac8166908b830f48683af5d7612dc
20160823 01:56:18< celticminstrel> Feel free to tweak the layout a bit further.
20160823 01:56:20< shadowm> No, its own implementation of a printf-like formatter.
20160823 01:56:52< celticminstrel> For the tiny-res one I think I'll need some code changes.
20160823 01:56:54< shadowm> It'd be rather pointless to reimplement printf since it's already there and the C library version is often hand-optimized for performance.
20160823 01:57:02< shadowm> (Maybe you meant replace, though.)
20160823 01:57:54< vultraz> celticminstrel: this does not touch the default resolution, right?
20160823 01:58:10< celticminstrel> Well, it touches the default resolution, but only in mundane, code-shifting ways.
20160823 01:58:16< celticminstrel> No functional/layout changes.
20160823 01:58:19< vultraz> ok
20160823 01:58:25< vultraz> what resolution do I need to see it?
20160823 01:58:32< celticminstrel> 1024x768
20160823 01:58:58< celticminstrel> Oh wait, did I mess up the resolution specifier...
20160823 01:59:06< vultraz> I'm seeing the default resolution with the scroll label all on one line :|
20160823 01:59:30< celticminstrel> Sorry, yeah, I made a last-minute change and messed it up.
20160823 01:59:31< vultraz> ironically, the scroll label is proper in 1152x864
20160823 01:59:35< vultraz> damn gui2 >_>
20160823 01:59:51< vultraz> once again it's failing to place a label and resorting to "let's make it FULL SIZED"
20160823 02:00:13-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:b01d:7dbd:5d82:c3ee] has joined #wesnoth-dev
20160823 02:00:34< irker616> wesnoth: Celtic Minstrel wesnoth:master 617a12f3e6be / data/gui/window/mp_create_game/small_resolution.cfg: MP Create: Low-res fixup https://github.com/wesnoth/wesnoth/commit/617a12f3e6be2b3204e49a2be48cd47da4c55cea
20160823 02:00:35-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has joined #wesnoth-dev
20160823 02:00:42< celticminstrel> Now you should see it at 1024x768.
20160823 02:01:35< vultraz> oh god, scrollbars EVERYWHERE
20160823 02:01:57< celticminstrel> The only scrollbar I see is on the listbox.
20160823 02:02:08< celticminstrel> And also the description field if it's long enough.
20160823 02:02:41< celticminstrel> Which might only be Aethermaw.
20160823 02:03:49< celticminstrel> You start getting dialog-wide scrollbars around 1000x700.
20160823 02:04:19< vultraz> no I see them everywhere :|
20160823 02:04:25< vultraz> horizontal scrollbar in the mods list
20160823 02:04:31< celticminstrel> Weird.
20160823 02:04:32< vultraz> horizontal scrollbar in the tab bar
20160823 02:04:36< vultraz> horizontal scrollbar in the window
20160823 02:04:45< vultraz> tiny vertical scrollbar in the description area
20160823 02:05:04< vultraz> this is likely because i have mods
20160823 02:05:28< celticminstrel> Ah, could be...
20160823 02:05:43< celticminstrel> I do get the vertical scrollbar in the description, if Aethermaw is selected.
20160823 02:06:41< celticminstrel> If the mods are the problem, why not just put them in the same column as era/faction?
20160823 02:06:50< celticminstrel> That'd even work in both layouts.
20160823 02:07:14< celticminstrel> I was actually considering doing that earlier.
20160823 02:07:29< celticminstrel> The player count is in a weird place.
20160823 02:07:48< celticminstrel> Or the map size is. One of those.
20160823 02:07:56< vultraz> eh?
20160823 02:07:59< vultraz> how so :|
20160823 02:08:03< celticminstrel> Just under the Load Game button?
20160823 02:08:21< celticminstrel> Or were you asking how so about the mods?
20160823 02:09:07-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has quit [Quit: Konversation terminated!]
20160823 02:09:09< vultraz> ...what?
20160823 02:09:36< celticminstrel> Sorry, I can't answer further questions until you clarify what you're asking about.
20160823 02:09:50< vultraz> map size/player count
20160823 02:09:53< vultraz> where are they in an odd place
20160823 02:10:22< celticminstrel> For me, the 3 (player count) is directly under the Load Game button.
20160823 02:10:33< celticminstrel> While the map size is right next to the corresponding label.
20160823 02:11:03< vultraz> on low res?
20160823 02:13:41-!- celmin [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160823 02:13:45< celticminstrel> Yes
20160823 02:14:11< vultraz> interesting
20160823 02:20:39< celmin> http://celmin.pwcsite.com/wesnoth/mp_creat_mid_res.png (can't be bothered to fix typo)
20160823 02:21:18< vultraz> mid res?
20160823 02:21:38< celmin> Yeah, it doesn't work at 800x600.
20160823 02:21:43< vultraz> ah
20160823 02:29:16-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160823 02:30:32-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Client Quit]
20160823 02:34:12< vultraz> interesting that the tab bar asserts when its column has horizontal_grow = "true"
20160823 02:38:35-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 02:38:36< travis-ci> wesnoth/wesnoth#10516 (master - f396249 : Celtic Minstrel): The build failed.
20160823 02:38:36< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154332999
20160823 02:38:36-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 02:39:24< vultraz> and Campaigns asserts too..
20160823 02:40:39< irker616> wesnoth: Charles Dang wesnoth:master ff06bf8eb722 / data/gui/window/mp_create_game/small_resolution.cfg: MP Create: few small fixes to mid-res layout https://github.com/wesnoth/wesnoth/commit/ff06bf8eb72227ce6ef49d4c7af41c990768fa23
20160823 02:40:41< vultraz> celmin: ^ fixes the label thing 
20160823 02:44:59< celmin> Incidentally, my plan was to call the third size tiny_resolution.cfg
20160823 02:45:19< celmin> In case you were wondering why the mid-res one was small_resolution.cfg
20160823 02:45:58-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160823 02:46:28< vultraz> should i commit a removal of snprintf in unit/drawer.cpp?
20160823 02:46:41< celmin> I dunno.
20160823 02:46:45< celmin> Where was it again?
20160823 02:46:54< vultraz> https://github.com/wesnoth/wesnoth/blob/master/src/units/drawer.cpp#L217
20160823 02:48:37< celmin> I see.
20160823 02:48:57< celmin> I notice the only difference between the two is top vs botton.
20160823 02:48:59< celmin> ^bottom
20160823 02:49:18< vultraz> yes
20160823 02:49:21< celmin> I guess it could be replaced with formatter() or a stringstream.
20160823 02:49:38< vultraz> or just +
20160823 02:50:05< celmin> I don't think I'd use +
20160823 02:50:29< vultraz> why not?
20160823 02:50:31< celmin> The strings could also be refactored out of the ifdef...
20160823 02:50:43< vultraz> yes I did that
20160823 02:51:04< celmin> I think string concatenation is likely to be inefficient.
20160823 02:51:16< tad_> I'm trying to figure out [filter_recall] .. I think it would help with an idea I have but I can't seem to get it to actually do anything. Anyone have any insights?
20160823 02:51:21< celmin> So doing it 15 times in a row is probably not a great idea.
20160823 02:51:37< celmin> tad_: It takes a SUF, right? And you can only recall units matching that?
20160823 02:52:15< tad_> celmin: that's what the docs say, but I give is an SUF and it does nothing and when I look with inspect it's empty.
20160823 02:52:21< celmin> Huh.
20160823 02:52:40-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 02:52:41< travis-ci> wesnoth/wesnoth#10517 (master - 617a12f : Celtic Minstrel): The build failed.
20160823 02:52:41< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154333575
20160823 02:52:41-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 02:54:11< tad_> The only examples I could find are LoW and they're static .. every time the leader appears it uses the macro with a static filter. But I can't even get it to store the filter.
20160823 02:55:34< irker616> wesnoth: Charles Dang wesnoth:master 6100bac25875 / src/units/drawer.cpp: Refactored out some uses of snprintf in getting unit ellipses https://github.com/wesnoth/wesnoth/commit/6100bac25875aa96103dfd4d7c4661ee99c659a8
20160823 02:55:44< vultraz> dowwwwwnn with C-isms 
20160823 02:55:45< celmin> I dunno, could it only work in [side][unit]? Or only in ActionWML [unit]?
20160823 02:56:25< tad_> The comments in changelog say [side][unit] and it's clearly leaders only. 
20160823 02:56:36< celmin> Obviously leaders only.
20160823 02:56:58< celmin> Though it should work in non-leaders too and just do nothing (unless they later gain canrecruit).
20160823 02:58:16< tad_> I'm hoping to use it when a non-leader gains recruit, and I'm hoping to be able to change it but may need to drop to Lua for that because it looks like it's supposed to be static and not something one can ever change
20160823 02:58:45< tad_> But, I can't get it to filter because when I look it's always empty.
20160823 02:59:04 * celmin wonders if wedge009 is around.
20160823 02:59:05< tad_> Like it gets wiped out when S01 sets it and S02 does not.
20160823 03:00:14-!- louis94 [~~louis94@91.178.242.249] has quit [Ping timeout: 250 seconds]
20160823 03:00:23< tad_> What I'm doing is looking at all this storing and unstoring and thinking .. "Gee, if I could filter the units for THIS scenario all that would go away!"
20160823 03:01:24< celmin> …oh, I suppose I could just remove the vconfig argument and see where the compiler errors appear.
20160823 03:02:23< celmin> Sorry wedge, never mind after all.
20160823 03:02:30< tad_> celmin: Can have found where it's used, and where it's stored in the unit class / copied during copy contructor.
20160823 03:02:42< celmin> Hm?
20160823 03:03:04< celmin> I didn't think of the copy constructor...
20160823 03:03:58< tad_> As I understand it, when I start with S01 the [unit] for the leader is created. Thereafter, it's pulled from the recall list and the [side][unit(ish)] only applies if there is no leader in the recall list.
20160823 03:04:40< tad_> But when I set it in S01, go to S02, it's gone.
20160823 03:05:06-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160823 03:05:39< tad_> So, I was hoping for some insight. Otherwise I'll keep fiddling with it. Probably something I'm missing and I'm gonna have to put in the hours to learn it.
20160823 03:07:31< irker616> wesnoth: Charles Dang wesnoth:master 711dbdbc48ea / data/gui/window/game_load.cfg: Load Game: reorganized details panel https://github.com/wesnoth/wesnoth/commit/711dbdbc48ea784418ff463249bf5c9b16ea89f4
20160823 03:08:15< vultraz> eventually we'll want to display all human leaders there
20160823 03:08:32< celmin> I'll see if I can do anything about filter_recall.
20160823 03:08:48-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:b01d:7dbd:5d82:c3ee] has quit [Remote host closed the connection]
20160823 03:11:03< tad_> My goal is to be able to change it per scenario, and change it for alternate leaders when they take over. There's a mess of store/unstore going on simply to have the recall list read one way here and another there. And it would clean things up a LOT if the rule were 'all in the recall list, always' and simply filter as needed.
20160823 03:12:14-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:b01d:7dbd:5d82:c3ee] has joined #wesnoth-dev
20160823 03:13:18< tad_> As to 'doing anything' .. I don't know what's needed because I can't get the darned feature to even work once. So unless it's an obvious change, I'd say hold off until we undertand the use cases
20160823 03:14:00< celmin> So the vconfig constructor is only ever called from unit_creator...
20160823 03:14:06< tad_> At present the only use case I can find is LoW where it appears to be 'set once, never changes'
20160823 03:14:07-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:b01d:7dbd:5d82:c3ee] has quit [Remote host closed the connection]
20160823 03:14:14< celmin> Oh, wait, that's not true is it.
20160823 03:14:40< celmin> If I make it mandatory instead of deleting it, I should find other locations...
20160823 03:15:32< vultraz> celmin: so, I think we're committed to this MP Create layout. should we begin cleaning out the gui1 stuff?
20160823 03:15:41< tad_> In the [unit] creator it stores a config, copy copies it (by ref I think). And there's some uses of it in a couple functions via 'recall_filter' accessor
20160823 03:15:51< celmin> unit_creator is the only place a recall filter will work though...
20160823 03:16:04< celmin> vultraz: Not before you've done load tests for the MP lobby.
20160823 03:16:22< vultraz> How do I do that?
20160823 03:16:49< celmin> Well, the easiest way is probably using a Lua plugin, which means you need to make sure that the plugins context is set up.
20160823 03:17:03< vultraz> howww do I do that
20160823 03:17:20-!- travis-ci [~travis-ci@ec2-54-166-238-133.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 03:17:21< travis-ci> wesnoth/wesnoth#10518 (master - ff06bf8 : Charles Dang): The build is still failing.
20160823 03:17:21< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154338285
20160823 03:17:21-!- travis-ci [~travis-ci@ec2-54-166-238-133.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 03:17:39< tad_> OK. I'm off to play more. I'm sure I'll get it. Just gotta pay the dues.
20160823 03:17:43-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160823 03:17:46< celmin> Look at the GUI1 version. Search for plugins_context.
20160823 03:20:17< celmin> It's probably similar to binding callbacks to widgets.
20160823 03:25:30-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160823 03:26:45< tad_> So I did [filter_recall] exactly like LoW and got it working. It's a mess. The Wiki might best say "Only for LoW, do not use elsewhere." the feature needs a lot of work.
20160823 03:28:48 * celmin is curious what the actual restrictions are.
20160823 03:29:00< tad_> For example: it only filters when your leader is standing on a camp. It shows ALL units otherwise and simply says "Can't do that" in various ways depending on if it matches or does not match the filter.
20160823 03:30:24< tad_> I tried changing it from S01 to S02 and it didn't change. Maybe a [modify_unit] .. can you change a tag blocks there?
20160823 03:31:20< tad_> I can live with the Ctrl-R list showing too much if the leader isn't on a camp. They're there, you just can't do anything at all with them because you're not where you need to be on the map;.
20160823 03:31:38< tad_> Alt-R (Recall0
20160823 03:31:53< celmin> I get the impression that [filter_recall] works nowhere except in [side][unit].
20160823 03:32:10< celmin> But I haven't tracked down all uses of unit_creator, so maybe it works somewhere else too.
20160823 03:32:39< tad_> Well, it only 'works' during recall list pop-up.
20160823 03:33:57< tad_> I've not tested if a non-leader can have it at all .. that's easy enough to test.
20160823 03:34:25< tad_> It looks a bit fragile, though. and it certainly doesn't work as I expected from reading the wiki.
20160823 03:35:56< tad_> Anyway .. maybe I can get it to work, maybe not. The issues with what it shows, when, are probably non-issues if I'm playing normally and I'm seeing them because I'm jumping around trying to figure it out.
20160823 03:37:09< tad_> Can I [modify_unit] and set a sub-tag like [filter_recall]? It's only being set in the constructor, so I'm thinking not.
20160823 03:39:15-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 03:39:16< travis-ci> wesnoth/wesnoth#10519 (master - 6100bac : Charles Dang): The build is still failing.
20160823 03:39:16< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154339781
20160823 03:39:16-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 03:52:04< tad_> Hmm. Well, I don't know if [filter_recall] is really meant for anything OTHER than LoW. I can see why it's doing what it's doing for that campaign. Not sure I agree, but it makes a certain amount of sense, there, I guess. But on code-reading and testing, I don't think it's nearly the feature the wiki implies.
20160823 03:53:58< tad_> To make sense, you need several leaders. And at least one of them should be able to recall a given unit on the recall list. Otherwise, the recall list, error messages, and behavior don't really make sense.
20160823 03:56:12< tad_> which makes me wonder .. given that implementation as a starting point .. can I refine it to be something more generally useful.  *sigh* that means C++ .. well, I suppose I can get my head back into C++ for a couple days to see what's possible.
20160823 04:01:52-!- travis-ci [~travis-ci@ec2-54-166-238-133.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 04:01:53< travis-ci> wesnoth/wesnoth#10520 (master - 711dbdb : Charles Dang): The build is still failing.
20160823 04:01:54< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154340975
20160823 04:01:54-!- travis-ci [~travis-ci@ec2-54-166-238-133.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 04:05:34-!- JyrkiVesterinen [~JyrkiVest@87-100-217-59.bb.dnainternet.fi] has joined #wesnoth-dev
20160823 04:14:39< tad_> As I see it, the first disconnect for [filter_recall] is when you bring up the recall list. It shows too much. If none of the leaders can recall a unit that unit should not appear at all.
20160823 04:15:20< tad_> When I move a leader to a keep and bring up the list, it does show only the units that leader can recall. So that's good.
20160823 04:16:49< tad_> So, then I'd only need to change the filters on-the-fly for what I want to do. And it doesn't look like any of this would break LoW.
20160823 04:20:22-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160823 04:26:40< celmin> I think tad_ missed that I was looking at the filter_recall issue, though I dunno if my direction is the same as what he was planning...
20160823 04:35:44< vultraz> looks like a single plugin context is shared between all mp ui sections
20160823 04:36:18< celmin> Really?
20160823 04:36:21< vultraz> yes
20160823 04:36:22< celmin> I suppose that's helpful.
20160823 04:37:10< celmin> So (with GUI1) the context is set up in create and lasts until the end of connect?
20160823 04:38:05< vultraz> no, it's in the lobby too
20160823 04:38:13< celmin> Oh huh.
20160823 04:38:21< vultraz> it's reset at every stage, though
20160823 04:38:35< celmin> Ah, so it's probably not really the same context then...
20160823 04:39:11< vultraz> well, the same unique_prt at least
20160823 04:39:27< celmin> I wonder, should I force unit events to have an ID?
20160823 04:39:34< celmin> ie, ignore them if they don't.
20160823 04:41:34< celmin> (With an error, of course.9
20160823 04:41:34< celmin> ^)
20160823 04:41:34< celmin> Not going to do it right now, mind you.
20160823 04:41:34< vultraz> ok, I guess I just need to bind functions now
20160823 04:42:25-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160823 04:42:29-!- Kwandulin [~Miranda@p200300760F35BFCEDCB25943552F19B1.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160823 04:42:48< celmin> Even if it's the same plugins context object, are the functions the same for every stage?
20160823 04:42:56< celmin> Is the context name the same?
20160823 04:43:56< vultraz> no
20160823 04:44:01< vultraz> the name is not the same..
20160823 04:44:09< vultraz> and why do its functions all take a config
20160823 04:44:52< celmin> So that the functions can all choose their own interface.
20160823 04:44:57< celmin> Or something like that.
20160823 04:45:22< vultraz> blagh, I don't really want to inherit from mp::ui..
20160823 04:45:33< celmin> Then don"t?
20160823 04:47:29< celmin> Isn't it a GUI1 thing?
20160823 04:47:36< vultraz> yes
20160823 04:47:51< vultraz> I'll keep a single context pointer in create for now
20160823 04:48:52< vultraz> gah
20160823 04:49:00< vultraz> configs, configs, configs..
20160823 04:49:34< vultraz> I assume these functions are what specific plugin actions do?
20160823 04:49:37< celmin> Naturally each config has different meaningful keys.
20160823 04:49:44< celmin> Something like that?
20160823 04:49:53< vultraz> ok
20160823 04:50:01< vultraz> so why in hell would I need a config when quitting!
20160823 04:50:17< celmin> Because every function needs a config.
20160823 04:50:29< celmin> The quit one I think has a "code" key that sets the exit code.
20160823 04:50:40< celmin> So that the MP test for example can signal success or failure.
20160823 04:50:40< vultraz> jien2jnfiunuifnui34nfui3nfijnerjkfn 23ui4 f89p234hr89h3489rh23ui94nfui2ncin2dcuip23h9p8rh
20160823 04:51:01< celmin> If you want an example of how this is used from the Lua side, look in host.lua and join.lua.
20160823 04:51:33< celmin> I wonder which Lua kernal it runs in… app lua kernel?
20160823 04:51:36< celmin> ^kernel
20160823 04:51:53< vultraz> but I don't actually need a config to quit the dialog!!!
20160823 04:52:00< celmin> What?
20160823 04:52:14< celmin> You can always ignore the config parameter if it's not needed for anything...
20160823 04:52:46-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160823 04:53:22< tad_> celmin: I'm about there on [filter
20160823 04:53:43< tad_> celmin: I'm about there on [filter_recall] .. wondering your take on one change to the C++
20160823 04:53:48< celmin> Okay.
20160823 04:53:52 * vultraz curses
20160823 04:54:05 * celmin wishes vultraz would stop cursing.
20160823 04:54:28 * celmin will be celticminstrel in a few minutes, just noting.
20160823 04:55:05< tad_> I can modify it and have it take effect without chaning the C++. It's only slightly hacky since it uses [set_variables] but it's understandable enough.
20160823 04:55:27< celmin> ?
20160823 04:55:51< tad_> But the issue I have is if non of the leaders presently on the map can recall a unit, it displays the unit on the listbox. So .. how hard do you think it would be to not list such units?
20160823 04:56:28< celmin> I guess it would mean a change to the recall dialog or something.
20160823 04:56:42< vultraz> I can't pass any argument except a config :|
20160823 04:56:44< celmin> You'd need to find all leaders capable of recalling anything and run their recall filters.
20160823 04:56:46< tad_> Oh .. [store_unit] the leader [set_variables] to set the WML tags as desired then [unstore_unit] and [filter_recall] obeys my changes
20160823 04:56:57< celmin> vultraz: But a config can contain anything you need. :P
20160823 04:57:01< vultraz> NO!
20160823 04:57:07< celmin> Yes it can.
20160823 04:57:15< vultraz> if you want to quit you just call window.set_retval
20160823 04:57:21< celmin> Right.
20160823 04:58:18< tad_> At present when I pull up the recall list (Alt-R) it shows them all. If I click on one which no leader can recall, it says 'no leader can recall this unit' so it has the data, just after displaying rather than supressingthe display. 
20160823 04:59:09< celmin> Would you expect it to show all units that can be recalled by any leader, or only units that can be recalled by those leaders which are on a keep?
20160823 04:59:11< tad_> I note that the list is already correct if I pull it up on with a leader on a keep. Don't know about two leaders on two keeps ..
20160823 04:59:47< tad_> At the point it shows 'all units on recall list' I would prefer it show 'all units recallable by any leader presently on the map'
20160823 05:00:21< celmin> Okay.
20160823 05:00:35< tad_> istm that would not break LoW since it presents only data the player can use .. which LoW must be expecting and suppresses data the player can't use which LoW must not be presuming
20160823 05:00:54< celmin> So then you could still get the "can't recall" message if the relevant leader is not on a keep.
20160823 05:01:08< vultraz> celmin: so if you know how to get a window from a config, do tell me
20160823 05:01:09< tad_> That seems to be what LoW expects.
20160823 05:01:23< vultraz> because this makes no sense
20160823 05:01:26< celmin> vultraz: If you're trying to get a window from a config, you're probably approaching the problem wrong.
20160823 05:01:40< vultraz> I'm being sarcastic :
20160823 05:01:41< vultraz> :|
20160823 05:01:42< celmin> You bind the window into the function.
20160823 05:02:20< vultraz> don't tell me I need to create function wrappers just to deal with this atrocious plugin interface :|
20160823 05:02:43< tad_> You seem to have your hands full. I think I know enough to be able to run with it and simply note the desired change to the recall list pop-up for a (near?) future change.
20160823 05:03:06< celmin> It sounds relatively easy to do.
20160823 05:03:33< celmin> vultraz: You might have to, though not for functions that are expected to be called with an empty config.
20160823 05:03:35< tad_> I'll make notes in my issues and we can talk about it when (if?) vultraz ever sleeps ... :P
20160823 05:04:28-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Quit: Page closed]
20160823 05:04:37< celmin> Uh. Why the heck did I implement this command this way?
20160823 05:05:08< celmin> Oh right, I was probably copying the generic form just below that.
20160823 05:06:18-!- JyrkiVesterinen [~JyrkiVest@87-100-217-59.bb.dnainternet.fi] has quit [Quit: .]
20160823 05:07:08< celmin> It may be just a debug command, but why recreate the unit when it's easy to do this without that?
20160823 05:10:05-!- celmin [~celticmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The minstrel departs, to spread the music to the masses!]
20160823 05:14:06< vultraz> I just...cannot understand why a function would need a config when it doesn;t use it!
20160823 05:14:53< celticminstrel> Most of them probably do use it.
20160823 05:16:53< celticminstrel> Since most of them use it, all of them must accept one in order to have a simple unified interface.
20160823 05:16:57< celticminstrel> Or something like that.
20160823 05:17:18< celticminstrel> In practice, it's more of a string->attribute_value map. I don't think any of the functions use child tags.
20160823 05:17:29< celticminstrel> So it's like keyword arguments.
20160823 05:26:00< vultraz> but I don't need it for Quit :|
20160823 05:26:03< vultraz> or Create
20160823 05:26:29< celticminstrel> Then just ignore it?
20160823 05:31:10< vultraz> but I still need a window argument :|
20160823 05:31:58< celticminstrel> So bind it.
20160823 05:32:07< celticminstrel> Don't use a lambda, use bind.
20160823 05:32:23< celticminstrel> Alternatively, if you really prefer a lambda, then capture the window.
20160823 06:04:38-!- Kwandulin [~Miranda@p200300760F35BFCEDCB25943552F19B1.dip0.t-ipconnect.de] has quit [Ping timeout: 250 seconds]
20160823 06:07:43-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20160823 06:22:09-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160823 06:22:21< irker616> wesnoth: Wedge009 wesnoth:master 872e2e45ad40 / projectfiles/VC12/ (wesnoth.vcxproj wesnoth.vcxproj.filters): Tidying VC project files. https://github.com/wesnoth/wesnoth/commit/872e2e45ad4088af80bc220c01aa786b8e4580dd
20160823 06:49:27-!- boucman_work [~boucman@193.56.60.161] has joined #wesnoth-dev
20160823 06:50:40-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20160823 06:53:10< shadowm> Is the conversation above related to the fact that I'm getting a warning on src/gui/dialogs/multiplayer/mp_create_game.cpp:455 right now?
20160823 06:55:16< vultraz> is it an "unused argument" warning?
20160823 06:55:19< vultraz> if so then no
20160823 06:55:38< shadowm> Okay, but you're aware of it at least.
20160823 06:55:42< vultraz> Yes
20160823 06:56:04< vultraz> The call that uses it is commented out since it causes a crash for the same reason you get a crash in the gui2 file browser dialog
20160823 06:56:10< vultraz> (at least, I think)
20160823 06:56:14< vultraz> it's being looked into
20160823 06:58:41< vultraz> with any luck, it shall be fixed soon
20160823 06:59:05< shadowm> Well, if you manage to wire into this byzantine mess a mechanism to make event processing aware of widget lifecycles then certainly I'd appreciate a heads up when you're done.
20160823 07:01:09-!- travis-ci [~travis-ci@ec2-54-159-50-178.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 07:01:10< travis-ci> wesnoth/wesnoth#10521 (master - 872e2e4 : Wedge009): The build is still failing.
20160823 07:01:10< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154364085
20160823 07:01:10-!- travis-ci [~travis-ci@ec2-54-159-50-178.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 07:16:25< irker616> wesnoth: Charles Dang wesnoth:master 9348ef21885e / src/gui/dialogs/multiplayer/ (mp_create_game.cpp mp_create_game.hpp): MP Create: set up Lua plugin context https://github.com/wesnoth/wesnoth/commit/9348ef21885efad7ad88f7abf6e6694efeac8cfa
20160823 07:16:27< vultraz> celticminstrel: ^
20160823 07:17:44< vultraz> is this acceptable?
20160823 07:18:32 * celticminstrel pulls.
20160823 07:19:03< celticminstrel> The way to be sure is to run the MP test with the new lobby enabled.
20160823 07:19:09-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20160823 07:19:14< vultraz> can you do that?
20160823 07:19:17< celticminstrel> No.
20160823 07:19:28< vultraz> will trvais do that?
20160823 07:19:35< celticminstrel> I don't know why, but the server I build crashes on my computer.
20160823 07:19:38< celticminstrel> I suggest you do it.
20160823 07:19:45< vultraz> I don't have the tests
20160823 07:19:51< celticminstrel> Yes you do.
20160823 07:19:57< vultraz> eh?
20160823 07:20:14< celticminstrel> You can't directly run the shell-script, but the key commands in it will work the same on Windows as on Mac/Linux.
20160823 07:20:30< celticminstrel> The shell-script is at utils/travis/mp_test_executor.sh
20160823 07:20:40< vultraz> I do have bash
20160823 07:20:44< celticminstrel> Just open three Command Prompt windows.
20160823 07:20:46< vultraz> if that helps
20160823 07:21:02< celticminstrel> Well, you could use bash, but it's not really necessary.
20160823 07:21:20< celticminstrel> If you use bash you can just call the shell-script.
20160823 07:21:52< celticminstrel> If you're in an ordinary command prompt, you'd do "bash utils\travis\mp_test_executor.sh" I assume.
20160823 07:22:07< celticminstrel> If in a bash prompt, "./utils/travis/mp_test_executor.sh"
20160823 07:22:37< celticminstrel> Or you could open three command prompts, run the server in one (copy-paste from the shell-script) and the clients in the other two (likewise).
20160823 07:23:32< vultraz> ok, what should I be looking for
20160823 07:24:07< vultraz> it's saying a lot lot of "Playing multiplayer"
20160823 07:24:14< vultraz> now it
20160823 07:24:17< vultraz> s not doing anything
20160823 07:24:29< celticminstrel> That might be a bad sign. Not quite sure though.
20160823 07:24:38-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has joined #wesnoth-dev
20160823 07:24:41< celticminstrel> You could compare it to the Travis output from the MP test.
20160823 07:25:32< celticminstrel> Your effort at setting up the plugin contexts looks good at a glance, but I do see at least one possible issue.
20160823 07:25:50< vultraz> oh?
20160823 07:26:17< celticminstrel> The select_level callback looks like it might not be doing what it's supposed to.
20160823 07:26:33< celticminstrel> Unless programmatically selecting a row triggers its callback?
20160823 07:27:11< celticminstrel> It's not really necessary to update the GUI to reflect the new selection (though it's probably a good idea to do so) - more important is that the create_engine is updated.
20160823 07:27:36< celticminstrel> I can't tell at a glance whether select_type shares the same problem. It might not.
20160823 07:27:44< vultraz> hm, I just realized the gui1 version of load allowed passing some extra parameters
20160823 07:27:59< celticminstrel> Well, obviously you need to handle those too then. :)
20160823 07:28:13< vultraz> anyway, the tests decided to just sit there
20160823 07:28:25< celticminstrel> I don't think the select_level problem alone would be enough to make the tests fail...
20160823 07:28:49< vultraz> let's try again
20160823 07:29:40< celticminstrel> I think you actually need to run the plugins context or something at some point.
20160823 07:30:00< celticminstrel> You can think of the plugins context as a parallel thread that Wesnoth defers to at specific times.
20160823 07:30:18< celticminstrel> So control passes between the Lua code and the Wesnoth code.
20160823 07:30:45< celticminstrel> But unlike threading it only happens when one side explicitly hands off control.
20160823 07:30:55< vultraz> how do I do that
20160823 07:31:12< celticminstrel> On the Lua side that's done by calling coroutine.yield(), but I dunno how it's done on the C++ side.
20160823 07:31:19< celticminstrel> I don't think it should happen in pre_show though.
20160823 07:31:50< celticminstrel> It should somehow be tied into the event loop or something, I guess.
20160823 07:31:59< celticminstrel> I dunno.
20160823 07:36:26< irker616> wesnoth: Charles Dang wesnoth:master 51e53c83838a / src/gui/dialogs/multiplayer/mp_create_game.cpp: MP Create: comment out unused function argument for now https://github.com/wesnoth/wesnoth/commit/51e53c83838a545ed603b32fb5047b497727d2b7
20160823 07:36:40< celticminstrel> So what you need to do is call ->play_slice(), the next question though is where.
20160823 07:37:03< celticminstrel> You could install a pump monitor, I suppose.
20160823 07:37:11< celticminstrel> Or some sort of timer mechanism.
20160823 07:37:12< vultraz> :|
20160823 07:37:21< vultraz> gui2 has a timer mechanism...
20160823 07:37:39< celticminstrel> For the record, the GUI1 call is multiplayer.cpp:92.
20160823 07:38:43< shadowm> Finished updating the wiki, hopefully I didn't break anything this time (Rhonda, celticminstrel).
20160823 07:38:44< celticminstrel> Does the GUI2 MP lobby set up its plugin context already?
20160823 07:39:05< vultraz> no
20160823 07:39:13< celticminstrel> Ah, well, you'll need to do that then.
20160823 07:39:42< shadowm> I also somehow found out just now that the last time I dropped the necessary files and permissions for MW uploads through the web interface.
20160823 07:39:47< vultraz> I guess I need a class all the dialogs inherit from like the gui1 version
20160823 07:39:52< celticminstrel> It's nice that you did it for MP Create, but the MP test at least will never get there without the MP Lobby's being set up! I think there might be a different test though if you just want to test MP Create...
20160823 07:40:21< shadowm> Fortunately I found the files from like two years ago and I don't think the MW upload has been actually used ever since I enabled it so it's probably fine.
20160823 07:40:29< celticminstrel> The play test doesn't use a plugin, okay.
20160823 07:40:30< shadowm> If you don't know what I'm talking about that's fine too.
20160823 07:40:50< celticminstrel> I guess you could write your own plugin to test it.
20160823 07:40:54< vultraz> D:
20160823 07:40:59< vultraz> extra work!
20160823 07:41:24< celticminstrel> Someday we should probably document the plugin mechanism.
20160823 07:43:58< vultraz> I guess i need to dd the configure plugin settings here too
20160823 07:44:00< vultraz> :/
20160823 07:44:01< vultraz> add*
20160823 07:44:22< celticminstrel> Sort of?
20160823 07:45:31< celticminstrel> I think the configure callbacks need to be registered by the create/load hooks.
20160823 07:46:00< celticminstrel> Maybe directly, or maybe just after calling play_slice.
20160823 07:48:26< celticminstrel> (The latter implies they set a flag or something, which is checked after calling play_slice.)
20160823 07:50:30< celticminstrel> (If you do want to use your own plugin, it might be simple to tweak host.lua or join.lua... not really sure though.)
20160823 07:50:48-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has quit [Ping timeout: 265 seconds]
20160823 07:52:12-!- travis-ci [~travis-ci@ec2-54-159-50-178.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 07:52:13< travis-ci> wesnoth/wesnoth#10522 (master - 9348ef2 : Charles Dang): The build is still failing.
20160823 07:52:13< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154372062
20160823 07:52:13-!- travis-ci [~travis-ci@ec2-54-159-50-178.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 07:54:23-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: Rebooting]
20160823 07:54:59-!- Kwandulin [~Miranda@p200300760F35BF8DA540F76ED7844112.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160823 07:57:15< celticminstrel> I guess I'll sleep.
20160823 07:57:27-!- celticminstrel is now known as celmin|sleep
20160823 07:58:36-!- boucman_work [~boucman@193.56.60.161] has quit [Ping timeout: 250 seconds]
20160823 08:04:11-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160823 08:06:14-!- boucman_work [~boucman@193.56.60.161] has joined #wesnoth-dev
20160823 08:16:20-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 08:16:21< travis-ci> wesnoth/wesnoth#10523 (master - 51e53c8 : Charles Dang): The build is still failing.
20160823 08:16:21< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154375339
20160823 08:16:21-!- travis-ci [~travis-ci@ec2-54-227-23-16.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 08:18:40< vultraz> le sigh
20160823 08:19:02< vultraz> now we're back to that memory access error
20160823 08:20:25< vultraz> something with the inspector, it seems..
20160823 08:29:13< Aginor> celmin|sleep: you had an interesting discussion/monolouge about lua execution
20160823 08:29:32< Aginor> I may have some thoughts on the matter ;)
20160823 08:40:44-!- prkc [~prkc@192.40.89.72] has quit [Ping timeout: 244 seconds]
20160823 08:49:00-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20160823 08:55:18-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160823 08:56:21< wedge009> vultraz: There's another read error?
20160823 08:56:29< vultraz> hm?
20160823 08:56:32< vultraz> see travis
20160823 08:56:36-!- prkc [~prkc@catv-89-133-21-131.catv.broadband.hu] has joined #wesnoth-dev
20160823 08:56:43< wedge009> Oh okay, I don't really know how to read those.
20160823 09:14:10-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20160823 09:29:41-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has joined #wesnoth-dev
20160823 09:29:51-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has quit [Client Quit]
20160823 09:49:43-!- prkc [~prkc@catv-89-133-21-131.catv.broadband.hu] has quit [Ping timeout: 250 seconds]
20160823 09:57:29-!- ideuler [~textual@0.213.62.94.rev.vodafone.pt] has joined #wesnoth-dev
20160823 10:02:57-!- prkc [~prkc@192.40.89.13] has joined #wesnoth-dev
20160823 10:16:57-!- prkc [~prkc@192.40.89.13] has quit [Quit: Leaving]
20160823 10:17:11-!- prkc [~prkc@192.40.89.13] has joined #wesnoth-dev
20160823 10:19:15-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160823 10:34:21-!- ideuler [~textual@0.213.62.94.rev.vodafone.pt] has quit [Quit: Chakalaka.]
20160823 10:36:35-!- irker616 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160823 10:39:09-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has joined #wesnoth-dev
20160823 10:52:07-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Ping timeout: 244 seconds]
20160823 10:57:14-!- prkc [~prkc@192.40.89.13] has quit [Remote host closed the connection]
20160823 11:21:29< RatArmy> hi
20160823 11:21:43< RatArmy> Have female_message and male_message been already implemented in 1.13.6+dev?
20160823 11:21:56< RatArmy> I'm reviewing tutorial translation, but Li'sar's female_message is not working.
20160823 11:27:05-!- hk238 [~kvirc@unaffiliated/hk238] has joined #wesnoth-dev
20160823 11:35:01-!- kidneb [~kidneb@not.allthetime.xyz] has quit [Ping timeout: 250 seconds]
20160823 11:36:48-!- kidneb [~kidneb@178.79.173.107] has joined #wesnoth-dev
20160823 11:40:08-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160823 11:43:39-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 264 seconds]
20160823 11:43:40-!- wedge010 is now known as wedge009
20160823 11:46:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160823 11:51:12< vultraz> female_message?
20160823 11:52:33< RatArmy> yes
20160823 11:52:59< vultraz> I'm not sure where that is
20160823 11:54:45< RatArmy> in tutorial scenario1, [message] tag
20160823 11:55:23-!- fabi_ [~fabi@176.7.88.214] has joined #wesnoth-dev
20160823 11:55:34-!- fabi [~fabi@wesnoth/developer/fendrin] has quit [Read error: Connection reset by peer]
20160823 11:55:59< vultraz> when was that implemented..
20160823 11:56:42< fabi_> vultraz: Hello. I haven't seen Elvish_Hunter for a while, do you know more? Where is wesbot?
20160823 11:57:03< shadowm> wesbot has been dead for like two years now?
20160823 11:57:10< vultraz> Elvish_Hunter hasn't been around recently
20160823 11:58:40< vultraz> RatArmy: hm.. I'll ask celmin|sleep about this
20160823 11:58:51< RatArmy> vultaz: in 1.13.2
20160823 11:59:13< shadowm> https://raw.githubusercontent.com/wesnoth/wesnoth/1.13.2/changelog yep
20160823 11:59:14< RatArmy> from changelog
20160823 12:00:32< vultraz> looks like it might be a bug in the implementation
20160823 12:01:05< vultraz> do the messages not show up?
20160823 12:01:20< shadowm> wesbot has been away for so long its NickServ registration is gone apparently.
20160823 12:01:37< shadowm> I thought it was cloaked and that staff didn't expire cloaked accounts. :\
20160823 12:01:52< RatArmy> vultraz: message= is showed
20160823 12:01:57< shadowm> *project-cloaked
20160823 12:05:40-!- tamanegi [~chatzilla@180.4.243.84] has joined #wesnoth-dev
20160823 12:06:48-!- kidneb [~kidneb@178.79.173.107] has quit [Ping timeout: 276 seconds]
20160823 12:07:40< fabi_> Several wml tag openings in a single line is valid WML code? But it never happens in mainline because wmlindent explodes it?
20160823 12:08:15-!- kidneb [~kidneb@178.79.173.107] has joined #wesnoth-dev
20160823 12:14:59-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:b01d:7dbd:5d82:c3ee] has joined #wesnoth-dev
20160823 12:19:13-!- Greg-Boggs [~greg_bogg@2601:1c2:901:e170:b01d:7dbd:5d82:c3ee] has quit [Ping timeout: 250 seconds]
20160823 12:26:34-!- Duthlet [~Duthlet@dslb-188-104-254-170.188.104.pools.vodafone-ip.de] has joined #wesnoth-dev
20160823 12:35:33-!- Kwandulin [~Miranda@p200300760F35BF8DA540F76ED7844112.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160823 13:00:23-!- RatArmy [~RatArmy@om126212092082.11.openmobile.ne.jp] has quit [Quit: Konversation terminated!]
20160823 13:03:21-!- Kwandulin [~Miranda@p200300760F35BF8D5027BCD4C4B306E2.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160823 13:10:26-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160823 13:32:10-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160823 13:42:39< vultraz> zookeeper: I thought you were going away?
20160823 13:43:36-!- louis94 [~~louis94@91.178.241.1] has joined #wesnoth-dev
20160823 13:48:56< zookeeper> vultraz, well if you look the clock i'd expect to have been gone exactly as long as i predicted, no? :P
20160823 13:51:04-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160823 13:53:09-!- louis94 [~~louis94@91.178.241.1] has quit [Ping timeout: 260 seconds]
20160823 14:02:14-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20160823 14:15:29-!- Shiki [~Shiki@141.39.226.227] has joined #wesnoth-dev
20160823 14:42:08-!- Kwandulin [~Miranda@p200300760F35BF8D5027BCD4C4B306E2.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160823 14:43:44-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20160823 14:57:01-!- tamanegi [~chatzilla@180.4.243.84] has quit [Quit: ChatZilla 0.9.92 [Firefox 45.1.1/20160507231935]]
20160823 15:07:04-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160823 15:10:16< zookeeper> mattsc, maybe for you: https://github.com/GregoryLundberg/wesnoth/issues/28
20160823 15:10:27< zookeeper> was discussed on https://www.wesnoth.org/irclogs/2016/08/%23wesnoth-dev.2016-08-17.log
20160823 15:11:46-!- Kwandulin [~Miranda@p200300760F35BF8DCD7950226E90D948.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160823 15:12:16-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160823 15:17:11< mattsc> zookeeper: thanks for the pointer, I’ll have a look; although just looking at the error message it might be something for celmin|sleep rather than me (I’m saying that not having read the irc discussion yet)
20160823 15:22:15-!- louis94 [~~louis94@91.178.241.1] has joined #wesnoth-dev
20160823 15:32:14< celmin|sleep> Aginor: What about Lua execution?
20160823 15:38:17-!- celmin|sleep is now known as celticminstrel
20160823 15:49:08-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 15:51:24-!- ideuler [~textual@0.213.62.94.rev.vodafone.pt] has joined #wesnoth-dev
20160823 15:53:20-!- boucman_work [~boucman@193.56.60.161] has quit [Ping timeout: 244 seconds]
20160823 15:55:55-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has joined #wesnoth-dev
20160823 15:56:10-!- tad_ [add94167@gateway/web/freenode/ip.173.217.65.103] has quit [Client Quit]
20160823 15:56:58< vultraz> didn
20160823 15:57:08< vultraz> 't someone add an option to download dependencies 
20160823 15:57:11< vultraz> for a game
20160823 15:57:12< vultraz> from mp
20160823 15:57:14< vultraz> somewhere
20160823 16:00:40< celticminstrel> What's but #18488...
20160823 16:00:43< celticminstrel> ^bug
20160823 16:01:47< vultraz> something fixed 5 years ago?
20160823 16:01:59< celticminstrel> It was fixed?
20160823 16:02:19< vultraz> are you looking at a comment?
20160823 16:02:22< vultraz> in the source?
20160823 16:02:31< celticminstrel> I'm asking because I see a "testcase for bug #18488" which asserts when you do ;unit fail
20160823 16:02:39< vultraz> "bug #18488: unit map iterators are invalidated after '.erase(loc) .add(loc)' sequence"
20160823 16:02:44< celticminstrel> I see.
20160823 16:02:52< vultraz> https://gna.org/bugs/index.php?18488
20160823 16:03:06< celticminstrel> I wonder if we still need that then.
20160823 16:03:14< vultraz> not likely
20160823 16:04:33-!- louis94 [~~louis94@91.178.241.1] has quit [Ping timeout: 244 seconds]
20160823 16:05:30< vultraz> i'm working on the plugin context for the lobby
20160823 16:05:38< vultraz> but we still need a play_slice call
20160823 16:05:44< vultraz> or whatever it is
20160823 16:05:48< celticminstrel> Right.
20160823 16:06:48< vultraz> also really need to go through and do a more thorough cleanup of the lobby code..
20160823 16:06:56< vultraz> I did some cleanup recently but it's not near enough
20160823 16:07:51< celticminstrel> I assume .erase(end(), end()) is a no-op...
20160823 16:08:44< vultraz> I'd be really nice if we could start clearing out the GUI1 stuff and replacing it with gui2 proper
20160823 16:14:39< vultraz> i'll work on this tomorrow
20160823 16:15:56-!- JyrkiVesterinen [~JyrkiVest@87-92-18-178.bb.dnainternet.fi] has joined #wesnoth-dev
20160823 16:17:44< celticminstrel> Not when the GUI2 side isn't complete, fully-functional, and working. :P
20160823 16:22:26< vultraz> oh come on, it's very close :|
20160823 16:22:57< celticminstrel> Well, that's great, but it needs to be more than close before cleaning old stuff out.
20160823 16:23:23< celticminstrel> Also we probably shouldn't clear anything out until after it's been the default in at least one release.
20160823 16:23:26< JyrkiVesterinen> celticminstrel: Yes, .erase(end(), end()) is a no-op.
20160823 16:23:40< celticminstrel> I ended up doing that a different way though.
20160823 16:23:49< celticminstrel> (About the erase thing.)
20160823 16:24:23< celticminstrel> Not because I wasn't sure, but because I realized that the list was a copy in the first place, so filtering during the copy is better.
20160823 16:24:54< celticminstrel> I want to merge PR 744...
20160823 16:25:04< celticminstrel> I dunno if wedge009 tested it again though.
20160823 16:25:38< celticminstrel> (Or anyone else.)
20160823 16:28:11< celticminstrel> So about [filter_recall], if the only issue is that it's not applied to the recall dialog, I think I've fixed that.
20160823 16:28:21< celticminstrel> But if there were other issues, then I dunno.
20160823 16:37:54-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160823 16:42:26-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 16:43:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 16:43:31< celticminstrel> Ah, looks like he did.
20160823 16:44:15< celticminstrel> vultraz: What should we do with PR 744?
20160823 16:46:30-!- ideuler [~textual@0.213.62.94.rev.vodafone.pt] has quit [Quit: Chakalaka.]
20160823 16:48:09-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20160823 16:50:51-!- horrowind [~Icedove@2a02:810a:83c0:404:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160823 17:06:22-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 17:13:09-!- Kwandulin [~Miranda@p200300760F35BF8DCD7950226E90D948.dip0.t-ipconnect.de] has quit [Quit: Kwandulin]
20160823 17:16:08-!- louis94 [~~louis94@91.178.241.1] has joined #wesnoth-dev
20160823 17:34:16-!- fabi [~fabi@176.0.34.124] has joined #wesnoth-dev
20160823 17:35:58-!- irker130 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160823 17:35:58< irker130> wesnoth: Jyrki Vesterinen wesnoth:master af2303df7782 / src/gui/dialogs/multiplayer/ (mp_create_game.cpp mp_create_game.hpp): Experimental MP Create: update options list immediately on reset to default https://github.com/wesnoth/wesnoth/commit/af2303df778225cb837461cc9943470835d01e1f
20160823 17:36:15< JyrkiVesterinen> Heh. Fixing the crash was easier than I expected. 8-)
20160823 17:37:05< celticminstrel> Ooh, previously unused features of the event handling system.
20160823 17:37:05-!- fabi_ [~fabi@176.7.88.214] has quit [Ping timeout: 265 seconds]
20160823 17:37:32< celticminstrel> Those unused parameters that everyone has just been ignoring all this time.
20160823 17:39:45-!- mjs-de [~mjs-de@x4db6a4d9.dyn.telefonica.de] has joined #wesnoth-dev
20160823 17:45:07-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160823 17:48:58-!- boucman_work [~boucman@2a02-8428-034f-f800-9e32-0c7c-b391-6223.rev.sfr.net] has joined #wesnoth-dev
20160823 18:00:09-!- hk238 [~kvirc@unaffiliated/hk238] has quit [Quit: http://www.kvirc.net/ 4.9.1 Aria]
20160823 18:04:37< mattsc> celticminstrel: have you ever figured out why the debug_ai table does not work when you access it the second time?
20160823 18:04:53< celticminstrel> I completely forgot about that.
20160823 18:05:21< mattsc> I had too, I’m just now going through my notes after being back from traveling
20160823 18:05:32< celticminstrel> What was the problem, again? Was it that the table was missing, or that it was set to read-only, or what?
20160823 18:06:11< mattsc> You could only access it once after starting Wesnoth, the second time you get an error
20160823 18:06:47-!- boucman_work [~boucman@2a02-8428-034f-f800-9e32-0c7c-b391-6223.rev.sfr.net] has quit [Remote host closed the connection]
20160823 18:07:10< mattsc> as in, the ai = wesnoth.debug_ai(side).ai only worked once
20160823 18:07:23< mattsc> *works
20160823 18:07:35< celticminstrel> So, if you did it a second time...
20160823 18:07:39< celticminstrel> 1) You get nil?
20160823 18:07:47< celticminstrel> 2) a.read_only is true?
20160823 18:08:04< celticminstrel> 3) Something else?
20160823 18:08:07< celticminstrel> ^ai.read_only
20160823 18:08:28< celticminstrel> If I recall correctly, wasn't the error something about "move does not exist"?
20160823 18:12:12-!- boucman_work [~boucman@2a02-8428-034f-f800-9e32-0c7c-b391-6223.rev.sfr.net] has joined #wesnoth-dev
20160823 18:12:22-!- boucman_work [~boucman@2a02-8428-034f-f800-9e32-0c7c-b391-6223.rev.sfr.net] has left #wesnoth-dev []
20160823 18:12:32-!- louis94 [~~louis94@91.178.241.1] has quit [Ping timeout: 240 seconds]
20160823 18:12:56-!- travis-ci [~travis-ci@ec2-54-92-146-1.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 18:12:57< travis-ci> wesnoth/wesnoth#10524 (master - af2303d : Jyrki Vesterinen): The build is still failing.
20160823 18:12:58< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154523790
20160823 18:12:58-!- travis-ci [~travis-ci@ec2-54-92-146-1.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 18:14:06< mattsc> celmin: http://pastebin.com/wkbRrHs9
20160823 18:14:17< mattsc> that’s the code I am using and the output
20160823 18:14:25-!- louis94 [~~louis94@91.178.241.1] has joined #wesnoth-dev
20160823 18:14:27< mattsc> celticminstrel: ^
20160823 18:14:43< mattsc> that’s the first time after starting Wesnoth
20160823 18:14:57< mattsc> the second time executing this code you get that error message on the first call
20160823 18:16:28< celticminstrel> So the debug_ai function itself throws an error...
20160823 18:16:38< mattsc> it seems so
20160823 18:28:07< mattsc> celticminstrel: do you know if the Xcode project is currently up to date?  I just tried to compile for the first time after being away for 3 weeks and get a boost_filesystem linker error in wesnothd
20160823 18:28:29< celticminstrel> The XCode project should be up-to-date.
20160823 18:28:31-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 18:28:35< mattsc> hmm
20160823 18:28:38< celticminstrel> You might need to update your libs though.
20160823 18:28:47< celticminstrel> I don't know if ancestral updated that yet though.
20160823 18:28:52< mattsc> oh …
20160823 18:29:09< mattsc> it’s a library not found error though, not a problem inside the libraries
20160823 18:29:21< celticminstrel> Oh, dyld?
20160823 18:29:38< celticminstrel> ie, library not found at runtime?
20160823 18:29:42-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 18:29:42-!- oldlaptop [~quassel@50-107-81-99.adr02.mskg.mi.frontiernet.net] has quit [Ping timeout: 250 seconds]
20160823 18:29:44< celticminstrel> Or is it a linker error?
20160823 18:30:06< mattsc> a linker error when compiling/linking wesnothd
20160823 18:30:17< mattsc> strange
20160823 18:30:57< celticminstrel> Oh, I remember now.
20160823 18:31:13< celticminstrel> Ancestral changed the Boost library names that the project expects.
20160823 18:31:22< celticminstrel> So they should end in -mt instead of w
20160823 18:31:57< mattsc> I see ...
20160823 18:32:03< mattsc> Any reason why this was done?
20160823 18:32:09< celticminstrel> I don't remember.
20160823 18:32:11< mattsc> (and yes, I had just noticed that myself)
20160823 18:33:10< mattsc> actually, I think the -mt is the default name, and the ending in w is something we did to rename them
20160823 18:33:34< mattsc> I might have a very vague memory from a long time ago that ancestral asked me whether it was fine to use the default names
20160823 18:33:39< mattsc> I probably said yes
20160823 18:34:47< mattsc> “we” up there meaning: whoever set up the Xcode project before I was the packager
20160823 18:36:32< mattsc> I got past the linker error of wesnothd by renaming the boost libraries.  Will see what happens when we get to linking Wesnoth, but I expect no problems.
20160823 18:36:36< mattsc> celticminstrel: thanks
20160823 18:43:04-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Ping timeout: 260 seconds]
20160823 18:44:00-!- louis94 [~~louis94@91.178.241.1] has quit [Ping timeout: 250 seconds]
20160823 18:45:47-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20160823 18:55:37-!- Kwandulin [~Miranda@p200300760F35BF8DF06CEFF115C578D5.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160823 18:56:18< mattsc> libpcre ?
20160823 19:01:09< celticminstrel> For some reason it's a dependency of glib, so...
20160823 19:01:32< mattsc> so why did I not need it before?
20160823 19:01:38< celticminstrel> I don't know.
20160823 19:01:53< celticminstrel> To be more precise, it's a dependency of my glib installed via MacPorts.
20160823 19:01:57< mattsc> did glib get updated?
20160823 19:02:05< celticminstrel> It seemed like ancestral's glib didn't need it.
20160823 19:02:14< celticminstrel> I don't know which is the older glib.
20160823 19:03:08< mattsc> mine doesn’t either, apparently
20160823 19:03:19< mattsc> sigh, I guess I have to compile some more libraries …
20160823 19:03:23< celticminstrel> We used my MacPorts libs because they're 10.7-compatible, while ancestral's brew(?)-built ones were not.
20160823 19:03:34< celticminstrel> You don't actually need pcre then.
20160823 19:03:46< celticminstrel> You could disable it in your copy of the project, for now.
20160823 19:03:58< mattsc> right, but then I always need to do that
20160823 19:04:05< celticminstrel> (Though installing it is still a good idea, because it's not good to have local changes to your project.)
20160823 19:04:11< mattsc> Btw, I am using MacPorts libs also
20160823 19:04:19< celticminstrel> Right, I meant as a temporary measure type thing.
20160823 19:04:29< mattsc> sure
20160823 19:04:33< celticminstrel> If you just want it to build now, you don't need pcre.
20160823 19:04:56< mattsc> it’s not a difficult problem to solve, I’m just too lazy …
20160823 19:04:59< mattsc> right
20160823 19:06:34< celticminstrel> Hmm, I had a sudden idea for transient message...
20160823 19:09:05-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160823 19:09:29< mattsc> which is?
20160823 19:09:46< celticminstrel> A "don't ask again" checkbox.
20160823 19:10:27< celticminstrel> Where did dbms come from again...
20160823 19:10:41-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160823 19:11:05< celticminstrel> ...oh, I just realized that I messed up. I assigned debug_ai() to a local variable, but I'm pretty sure that local variables do not work in the inspector.
20160823 19:11:38< celticminstrel> Well whatever, the goal wasn't actually to view the contents anyway.
20160823 19:13:22< celticminstrel> So debug_ai() works no matter how many times I call it...
20160823 19:14:01-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 19:14:15< mattsc> in other words, you cannot reproduce the error?
20160823 19:15:41< mattsc> celticminstrel: dbms come from Wesnoth Lua Tools, but I have a local version in AI-demos that I use
20160823 19:17:09< mattsc> Oh, great.  I just compiled myself a version of libpcre and copied it into my project directory.  Since I don’t need it, I don’t care whether it’s compatible with other libs I am using.
20160823 19:17:21< mattsc> But no I get white-box-syndrome for all text  :(
20160823 19:17:34< mattsc> s/no/now
20160823 19:17:56< mattsc> celticminstrel: are the new libraries in 1.13.5 already?
20160823 19:18:52< mattsc> the wesnoth_compile_mac zips don’t seem to have been updated yet
20160823 19:20:49< celticminstrel> I cannot reproduce, yeah.
20160823 19:21:01< celticminstrel> Pretty sure I do have dbms, but I guess it's not in global scope.
20160823 19:21:18< celticminstrel> The new libs are in 1.13.5, yes, including pcre.
20160823 19:21:34< mattsc> celticminstrel: okay, I’ll steal the libs from there then
20160823 19:21:49< mattsc> I thought I had compiled after the release already …
20160823 19:22:15< mattsc> celticminstrel: but you say it (debug_ai) may not be assigned to a local variable?
20160823 19:22:22< mattsc> is that the problem I am having?
20160823 19:22:33< celticminstrel> Uh, no, what I said is just in the inspector.
20160823 19:22:37< mattsc> (I cannot test right now since I don’t have a working version of master right now)
20160823 19:22:48< celticminstrel> Basically each command is in its own scope, so if you use local variables, they immediately expire.
20160823 19:23:06< celticminstrel> Each command executed from the Lua console, that is.
20160823 19:23:27< mattsc> and you did try the code I pasted as is?
20160823 19:23:42< celticminstrel> I just did wesnoth.debug_ai(1), three times.
20160823 19:24:07< celticminstrel> Just now I tried getting aggression, and got back 3.
20160823 19:24:11< mattsc> could you try accessing ai/get_aggression() after the first time?
20160823 19:24:25< celticminstrel> ...which actually seems a bit odd, but maybe I did something weird in the test scenario.
20160823 19:24:32< mattsc> 3?  that’s an unlikely value?
20160823 19:24:58< mattsc> did you try getting aggression and _then_ assigning debug_ai again?
20160823 19:24:59< celticminstrel> Yup, but it looks like it's correct.
20160823 19:25:24< celticminstrel> I think it's from when I was fixing [modify_side], I chose an unlikely value because it would be easy to see that it was working.
20160823 19:25:36< mattsc> ok
20160823 19:27:10< mattsc> anyway, IIRC I only get the error only if I tried to access aggression in between the two assignments the two calls to debug_ai
20160823 19:27:37< mattsc> and this is from test code, not from the console
20160823 19:28:24< celticminstrel> debug_ai() still works after I called get_aggression, too.
20160823 19:29:46< mattsc> hmm, so there must be something different from how I am using it
20160823 19:29:56< mattsc> yours is from the console?
20160823 19:30:00< celticminstrel> Yes.
20160823 19:30:25< mattsc> could that be the difference?
20160823 19:30:36< celticminstrel> I can't think of a reason why it could be, but who knows.
20160823 19:30:43< celticminstrel> How are you using it?
20160823 19:31:03< mattsc> in the ai-demos test scenario
20160823 19:31:19< celticminstrel> I mean, how is the problematic code triggered.
20160823 19:31:51< celticminstrel> Looking at this code, I'm wondering how the heck it works at all.
20160823 19:31:56< mattsc> right-click option
20160823 19:32:00-!- mjs-de [~mjs-de@x4db6a4d9.dyn.telefonica.de] has quit [Remote host closed the connection]
20160823 19:32:23< celticminstrel> Okay.
20160823 19:32:33< celticminstrel> I'll try doing that then.
20160823 19:35:04< mattsc> sorry, was afk for a couple minutes
20160823 19:35:15< mattsc> celticminstrel: which part surprises you that it works at all?
20160823 19:35:45< celticminstrel> I mean the C++ code that loads the AI table into scope for the purpose of debug_ai.
20160823 19:36:03< mattsc> if you have the AI-demos addon, you can just put the code into the test scenario, that has the right click option (and keyboard shortcut for it) installed already
20160823 19:36:15< celticminstrel> Oh, wait, I see - it leaves it on the stack no matter what.
20160823 19:36:16< mattsc> oh, not my code then  (phew)  ;)
20160823 19:36:19< celticminstrel> That's why it works.
20160823 19:36:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 19:37:15< mattsc> I’m installing 1.13.5 at the moment.  I’d been away for so long that I had not even done that yet.
20160823 19:37:23< mattsc> Then I’ll steal those libraries.
20160823 19:37:31< mattsc> And then I’ll hopefully be back in business.
20160823 19:39:29< celticminstrel> Works from a right-click menu too.
20160823 19:39:54-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 250 seconds]
20160823 19:40:19< celticminstrel> Hmm, actually, I see something suspicious here...
20160823 19:41:12< celticminstrel> The ai table from the second call is a different table.
20160823 19:41:52< celticminstrel> It doesn't contain the get_dst_src family of functions that are installed by the default Lua engine.
20160823 19:44:42< mattsc> celticminstrel: if you have the AI-demos add-on, I can easily have you replicate exactly what I am doing, if you want to
20160823 19:44:58< mattsc> maybe I am doing something “dirty” that by chance worked in 1.12, but not any more in master
20160823 19:45:13< celticminstrel> I don't have the add-on.
20160823 19:45:18< mattsc> I just got 1.13.5 up and am definitely still get the error there
20160823 19:45:32< mattsc> well, it’s a very small add-on, it should be easy to get  :)
20160823 19:45:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 19:45:40-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160823 19:45:56< celticminstrel> Though I'm not getting the error, I'm pretty sure something is wrong.
20160823 19:45:59-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 19:46:04< mattsc> okay
20160823 19:46:22< celticminstrel> Not sure if it's the root cause as your problem though.
20160823 19:46:26< celticminstrel> ^the same
20160823 19:46:51< mattsc> I am also getting something else that is strange, in that the code gets executed twice every time I press the hotkey (it did not in 1.12)
20160823 19:47:14< mattsc> but that should not have anything to do with this, the errors happen on the first time through (and then right away on the second time)
20160823 19:47:23-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 19:47:50-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 19:49:37< mattsc> Ah, rats, I don’t get the headers from the packaged version …
20160823 19:49:48< celticminstrel> Yeah, I mentioned that earlier.
20160823 19:50:00< celticminstrel> That you'll need to also get the correct version of the headers.
20160823 19:50:20< mattsc> right, you did ...
20160823 19:50:33< mattsc> you don’t happen to have whatever package you sent to ancestral, do you?
20160823 19:51:01< mattsc> Compiling worked just fine, but I am getting the white boxes for text in all the buttons etc. now.
20160823 19:51:22< mattsc> So that usually means some version problem with pango or fontconfig or something
20160823 19:51:34-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 19:51:53< celticminstrel> The package I sent to Ancestral didn't include Boost, but I think it's still up in http://celmin.pwcsite.com/wesnoth/
20160823 19:52:00-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 19:52:45-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160823 19:52:54< celticminstrel> Though if you had 1.13.5 it should be the same as what's there.
20160823 19:52:56< celticminstrel> ^have
20160823 19:53:15< celticminstrel> I guess from 1.13.5 it may not be obvious which libs actually need to be copied.
20160823 19:53:37< mattsc> So ancestral still got the headers separately then too?
20160823 19:54:00< celticminstrel> I dunno what he did about that. 1.13.5 is using his version for Boost though.
20160823 19:54:08-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160823 19:54:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 19:54:38-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 19:55:13-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 19:55:40-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 19:55:48-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 19:56:14-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 19:56:38-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 19:57:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 19:57:56-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 19:58:47-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 19:58:54< mattsc> celticminstrel: I simply copied in your libs into the frameworks directory and now I get text in the buttons again
20160823 19:59:03-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 19:59:09< celticminstrel> Yay!
20160823 19:59:14< mattsc> I’ll figure out later how to make it all compile
20160823 19:59:25< mattsc> … and link and copy files ...
20160823 19:59:29-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 19:59:48< mattsc> So is there anything I can help with figuring out the debug_ai bug right now?
20160823 19:59:58< celticminstrel> This code looks like it was doing things very wrong before, and yet, somehow it seems like things worked anyway...?
20160823 20:00:55< celticminstrel> ...what the heck is update_self...
20160823 20:00:58< mattsc> okay …  I cannot comment on that …
20160823 20:01:14 * celticminstrel means debug_ai(n).update_self there
20160823 20:01:33< mattsc> Looks like I better get myself into position to compile again
20160823 20:02:28< celticminstrel> The lua_ai_load class was supposed to load the current AI context onto the stack, but if already in an AI context, it was instead leaving the AI table on the stack rather than the context.
20160823 20:02:37< celticminstrel> (The context being the return value of debug_ai())
20160823 20:03:24< mattsc> okay; that sounds like it could have someting to do with what I am seeing
20160823 20:03:55< celticminstrel> I dunno. It seems like it shouldn't affect debug_ai since that's normally only called when the AI context is not active, but...
20160823 20:04:53< celticminstrel> There is clearly something weird happening with the second call to debug_ai.
20160823 20:06:59-!- JyrkiVesterinen [~JyrkiVest@87-92-18-178.bb.dnainternet.fi] has quit [Quit: .]
20160823 20:19:08< celticminstrel> I guess update_self is the initialization function.
20160823 20:19:30< celticminstrel> Though calling it doesn't actually update the self variable.
20160823 20:19:42< celticminstrel> You'd need x.self = x.update_self(x.params, x.data)
20160823 20:20:06< celticminstrel> (Where x = wesnoth.debug_ai(n))
20160823 20:21:11< celticminstrel> I guess the params that are supposdely immutable can actually be changed if you use debug_ai.
20160823 20:21:13< mattsc> celticminstrel: it looks like the Xcode project file is missing copying libpcre to the frameworks folder
20160823 20:21:22< celticminstrel> Okay...
20160823 20:23:27-!- oldlaptop [~quassel@162.247.150.37] has joined #wesnoth-dev
20160823 20:23:48< mattsc> once I add that, and using your libraries, things compile, link and run just fine
20160823 20:23:49< celticminstrel> Hmm, that didn't seem to help.
20160823 20:24:07< celticminstrel> Still missing get_src_dst with the second call.
20160823 20:26:51< celticminstrel> Why is update_state called twice...
20160823 20:28:39< mattsc> Why doesn’t Xcode let me reorder the list of files to copy?
20160823 20:28:49< celticminstrel> I don't know.
20160823 20:28:51< celticminstrel> It's really dumb.
20160823 20:29:01< mattsc> I’m getting that impression.
20160823 20:29:22< celticminstrel> I think I sorted them at some point by editing the project file in a text editor.
20160823 20:29:29< mattsc> Do you want me to commit that change to the project file (addind copying pcre, I mean) ?
20160823 20:29:40< mattsc> That’s what I figred I’ll have to do.
20160823 20:29:53< celticminstrel> Well, I don't mind, though it's annoying if I have to pull --rebase when the XCode project was touched.
20160823 20:30:08< celticminstrel> Then again, that's probably true for you too, so either way one of use will have to do it.
20160823 20:30:40< mattsc> I’m not having any active development going on (yet), so if it’s easier for you to do so, please go ahead
20160823 20:34:21-!- Kwandulin [~Miranda@p200300760F35BF8DF06CEFF115C578D5.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160823 20:34:36< celticminstrel> Trying to fix the issue I noticed, which is that the AI table seems to be replaced with a fresh copy at some point.
20160823 20:36:20-!- irker130 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160823 20:59:19< celticminstrel> vultraz: Did you remove a horizontal line from the mid-res MP Create?
20160823 20:59:26< celticminstrel> Or am I just confused?
20160823 21:03:19-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 21:03:31-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 21:06:26< celticminstrel> I must've been confused. There's still a horizontal line.
20160823 21:07:31< celticminstrel> Wait! debug_ai is constructing a dummy engine?
20160823 21:09:11-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 21:09:51< celticminstrel> And in so doing, the engine is initialized twice. Weird.
20160823 21:12:51-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 21:13:21< celticminstrel> So the first debug_ai call is immediately after initializing the engine...
20160823 21:14:00< celticminstrel> While the second uses that same engine... no idea if that's relevant though...
20160823 21:16:01< celticminstrel> What? If I try it on side 2 instead, it works...
20160823 21:17:55< celticminstrel> mattsc: Does the side you're using debug_ai on have a Lua engine defined? Either explicitly, or by having some Lua component?
20160823 21:17:57-!- Shiki [~Shiki@141.39.226.227] has quit [Quit: Verlassend]
20160823 21:18:22< mattsc> celticminstrel: let me check …
20160823 21:18:41< celticminstrel> Inspector might tell you, not quite sure...
20160823 21:19:00< mattsc> celticminstrel: I just looked at the scenario cfg.  And yes, it does.  Explicitely.
20160823 21:19:11< mattsc> Let me remove that and try again.
20160823 21:19:25< celticminstrel> Nah, my theory was the opposite.
20160823 21:19:31< celticminstrel> That it only works if you do have it.
20160823 21:19:54-!- oldlaptop [~quassel@162.247.150.37] has quit [Ping timeout: 265 seconds]
20160823 21:20:57< mattsc> celticminstrel: there might be something to it though
20160823 21:21:03< celticminstrel> I suppose.
20160823 21:21:07< mattsc> if I remove that, it already does not work the first time
20160823 21:21:16< celticminstrel> Huh, interesting.
20160823 21:21:35< mattsc> Well, no, it’s different
20160823 21:21:45< mattsc> I get error messages up front, but then it works both time
20160823 21:21:56< mattsc> *times
20160823 21:22:25-!- oldlaptop [~quassel@162.247.150.37] has joined #wesnoth-dev
20160823 21:22:41< celticminstrel> Interesting...
20160823 21:23:01< celticminstrel> So the error is weird in itself, "bad argument #2 to set_variable"...
20160823 21:23:21< celticminstrel> How would debug_ai internally be calling that...
20160823 21:23:38< mattsc> yes, wondering too
20160823 21:23:59< celticminstrel> In particular the presence of helper.lua in the stack trace...
20160823 21:24:13< celticminstrel> helper.lua is totally unknown to the C++ code.
20160823 21:24:38< celticminstrel> So if it's in the stack trace, I think that must imply that, somewhere along the line, debug_ai is causing some user code to be called...
20160823 21:24:43< mattsc> I get a similar but somewhat different error message now
20160823 21:24:46< mattsc> let me pastebin it
20160823 21:25:07< mattsc> http://pastebin.com/c3jy7ixm
20160823 21:27:25< mattsc> Huh, so these new errors are gone once I remove some of my includes
20160823 21:27:38< mattsc> and then the debug_ai call works
20160823 21:27:51< celticminstrel> Okay, so the problem is caused by the engine somehow?
20160823 21:27:56< celticminstrel> The custom engine.
20160823 21:28:37< mattsc> no, hold on ...
20160823 21:28:54< mattsc> sigh
20160823 21:29:02< mattsc> now it also works when I use the custom engine
20160823 21:29:09< celticminstrel> ...
20160823 21:29:16< mattsc> yeah, exactly
20160823 21:29:20< celticminstrel> So... maybe it was somehow caused by one of those includes?
20160823 21:29:26< mattsc> apparently some of my other code interfered with it
20160823 21:29:58< mattsc> the thing was though that the error message appeared _after_ the print statement I had just before the code I pastebinned to you earlier
20160823 21:30:20< mattsc> and not the first time around, only the second
20160823 21:38:13< celticminstrel> Is it not possible to make radio-buttons from Lua?
20160823 21:38:20< celticminstrel> I mean the appearance, not the behaviour.
20160823 21:39:36 * celticminstrel was just noticing that the "quick debug" thing uses checkboxes for the direction.
20160823 21:41:46< celticminstrel> I'm stumped on this, so I'm going to take a break for a bit and work on the low-res MP Create.
20160823 21:44:09-!- Appleman1234 [~Appleman1@KD036012028006.au-net.ne.jp] has quit [Ping timeout: 258 seconds]
20160823 21:45:03-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160823 21:45:50-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 250 seconds]
20160823 21:46:09-!- VultCave is now known as vultraz
20160823 21:48:52-!- irker566 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160823 21:48:52< irker566> wesnoth: Celtic Minstrel wesnoth:master 09854e99ad1b / src/actions/create.cpp: Only show recall units that a leader on the map can recall https://github.com/wesnoth/wesnoth/commit/09854e99ad1b6bbc5ca7710e637d324a48b9c2c2
20160823 21:48:52< irker566> wesnoth: Celtic Minstrel wesnoth:master 1e575480a409 / src/scripting/ (game_lua_kernel.cpp lua_common.cpp lua_common.hpp): When creating units in Lua from a vconfig, don't parse any contained [filter_rec https://github.com/wesnoth/wesnoth/commit/1e575480a409a3b377bd85052a6aaef741e861fd
20160823 21:48:53< irker566> wesnoth: Celtic Minstrel wesnoth:master 19e173d1a7f9 / src/synced_commands.cpp: Don't recreate unit with ;unit status command https://github.com/wesnoth/wesnoth/commit/19e173d1a7f9fd92aef727b6e9fe3b584b5bb04a
20160823 21:48:54< irker566> wesnoth: Celtic Minstrel wesnoth:master 888a3e26cb7a / src/game_events/action_wml.cpp: Don't parse [filter_recall] or [event] with [unit]to_variable= https://github.com/wesnoth/wesnoth/commit/888a3e26cb7a6d834ba5224b051323ca64c50032
20160823 21:48:55< irker566> wesnoth: Celtic Minstrel wesnoth:master b7a6f16f11d9 / src/scripting/game_lua_kernel.cpp: Add recall_filter to Lua unit proxy https://github.com/wesnoth/wesnoth/commit/b7a6f16f11d9e916270535d0656a4cbbdfa5fa0e
20160823 21:48:57< irker566> wesnoth: Celtic Minstrel wesnoth:master f71031d8fd51 / changelog: Update changelog https://github.com/wesnoth/wesnoth/commit/f71031d8fd51e8d0ff74136fd9d2bbed9ea154f8
20160823 21:48:59< irker566> wesnoth: Celtic Minstrel wesnoth:master 6b623147c8b3 / changelog: Neaten changelog https://github.com/wesnoth/wesnoth/commit/6b623147c8b3b742bf8116248cefa03d9d85f6fd
20160823 21:50:21-!- louis94 [~~louis94@91.178.241.1] has joined #wesnoth-dev
20160823 22:09:03-!- fabi_ [~fabi@176.4.125.129] has joined #wesnoth-dev
20160823 22:09:06-!- fabi [~fabi@176.0.34.124] has quit [Read error: Connection reset by peer]
20160823 22:14:51-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160823 22:17:47-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20160823 22:17:51-!- Duthlet [~Duthlet@dslb-188-104-254-170.188.104.pools.vodafone-ip.de] has quit [Quit: leaving]
20160823 22:22:09< celticminstrel> I have a layout that works for 800x600 but is a bit ugly... but it makes me think I can make this work without a fourth tab.
20160823 22:24:20-!- Appleman1234 [~Appleman1@KD036012023035.au-net.ne.jp] has joined #wesnoth-dev
20160823 22:27:11-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has joined #wesnoth-dev
20160823 22:27:12< travis-ci> wesnoth/wesnoth#10525 (master - 6b62314 : Celtic Minstrel): The build has errored.
20160823 22:27:12< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/154587117
20160823 22:27:12-!- travis-ci [~travis-ci@ec2-54-204-199-165.compute-1.amazonaws.com] has left #wesnoth-dev []
20160823 22:30:18-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20160823 22:32:09-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 22:36:20-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 22:39:01-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160823 22:43:26-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160823 22:49:32-!- Shiki [~Shiki@141.39.226.227] has joined #wesnoth-dev
20160823 22:57:21-!- Jetrel_bot is now known as Trollio
20160823 22:58:24-!- Trollio is now known as Jetrel_bot
20160823 23:14:37-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 258 seconds]
20160823 23:21:54-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 260 seconds]
20160823 23:23:39< Aginor> celticminstrel: I have, mainly for my own amusement, pondered how to rearchitecture wesnoth to fix some of the issues
20160823 23:23:50< Aginor> it'll probably introduce new ones instead though ;)
20160823 23:27:54< irker566> wesnoth: Celtic Minstrel wesnoth:master 4e6787639ae2 / data/gui/schema.cfg: GUI2 schema: fixup indentation https://github.com/wesnoth/wesnoth/commit/4e6787639ae22c9dcc48eba583adc0895cc2ab9a
20160823 23:27:56< irker566> wesnoth: Celtic Minstrel wesnoth:master aefe945218ca / data/gui/window/ (3 files in 2 dirs): GUI2 dialogs: Fix missing textdomains https://github.com/wesnoth/wesnoth/commit/aefe945218ca58585225a836be1de5511385abce
20160823 23:27:58< irker566> wesnoth: Celtic Minstrel wesnoth:master 44219de141a0 / data/gui/window/ (mp_create_game.cfg mp_create_game/tiny_resolution.cfg): MP Create: Layout for tiny resolutions (eg 800x600) https://github.com/wesnoth/wesnoth/commit/44219de141a0bb69043144e97785a6e066e7029a
20160823 23:28:14< celticminstrel> Still some problems with that layout - there's a slight vertical scrollbar, the alignment of the header is wrong, and the checkboxes at the top of the third tab are centred.
20160823 23:28:17< celticminstrel> vultraz: ^
20160823 23:28:40< celticminstrel> Aginor: Oh?
20160823 23:34:45< Aginor> pondering how to differentiate gamestate and graphics and event processing
20160823 23:35:14< Aginor> because I think those three things are too interwoven at the moment
20160823 23:37:00< celticminstrel> Graphics and gamestate, definitely. I'm not sure on event processing.
20160823 23:37:37< Aginor> events drive state changes ;)
20160823 23:38:05< Aginor> mind you, we have two types of events to consider. I/O events and game state events
20160823 23:38:17< celticminstrel> Actually, isn't splitting those three up ain to MVC?
20160823 23:38:24< celticminstrel> ^akin
20160823 23:38:57< celticminstrel> gamestate = model, graphics = view, event processing = controller?
20160823 23:39:16-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160823 23:39:40< Aginor> kind of, but there's a few other paradigms in there as well
20160823 23:39:50< Aginor> it allows us to have a process, update, draw cycle
20160823 23:40:06< Aginor> which is the most fundumental pattern in game design
20160823 23:40:08-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20160823 23:41:47< Aginor> I'm loath to call it a direct mapping to MVC as it is a bit too broad at that point
20160823 23:42:08< celticminstrel> Yeah, maybe it's not MVC exactly, but I was seeing parallels at least.
20160823 23:42:29< Aginor> yeah, there's certainly a parallel
20160823 23:43:26-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 250 seconds]
20160823 23:43:27-!- wedge010 is now known as wedge009
20160823 23:46:18-!- celmin [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160823 23:46:41< mattsc> celticminstrel: so I figured it’s one of my own AI debug hacks in my testing suite that was responsible for the problems we talked about earlier
20160823 23:46:45< mattsc> it isn’t
20160823 23:47:01< celmin> Okay.
20160823 23:47:12< mattsc> I have a helper.set_wml_var_metatable(_G) in there.  When I remove that, the problems disappear.
20160823 23:47:22< celmin> I see.
20160823 23:47:25< mattsc> I don’t even remember why I have that in the file.
20160823 23:47:53< mattsc> It’s probably left over in there from some early tests I did years ago when it was either needed or I had no idea what I was doing (or both)
20160823 23:48:03< Aginor> celticminstrel: is there anything in the lua/wml that does anything besides affecting gamestate?
20160823 23:48:24< celmin> If the C API function lua_getglobal() respects the metatable of _G, then that could certainly explain how set_variable was being called internally from debug_ai().
20160823 23:48:51< celmin> Aginor: Um, probably not? I guess they can install event handlers with [event] and [(set_)menu_item].
20160823 23:49:05< celmin> (I actually wonder if putting [menu_item] at scenario level would work.)
20160823 23:49:15< celmin> (Instead of [set_menu_item] in an event.)
20160823 23:49:52< celmin> Oh, some Lua/WML things do affect how the screen is drawn, like the fade-in-out stuff...
20160823 23:52:34-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
--- Log closed Wed Aug 24 00:00:45 2016