--- Log opened Sun Feb 28 00:00:06 2016
20160228 00:08:56-!- louis94 [~~louis94@91.178.240.5] has quit [Ping timeout: 244 seconds]
20160228 00:09:19< Aginor> I take that back
20160228 00:09:25< Aginor> vultraz: you broke SDL1 on master
20160228 00:12:18 * celticminstrel isn't all that surprised.
20160228 00:13:15< Aginor> I'll fix it on guifixes
20160228 00:14:21 * celticminstrel is altering Lua goals so that you can use type="explicit" or type="battle aid" or whatever instead of integer constants. Also, it'll default to explicit if there's no type key in your return table.
20160228 00:16:13< celticminstrel> (If you prefer the integer constants though, you can still use them.)
20160228 00:17:05< Aginor> celticminstrel: I've stayed out of all of the wml and lua stuff so far
20160228 00:17:08< Aginor> but that sounds like a good thing
20160228 00:17:25< celticminstrel> That comment was technically aimed at mattsc if he's around.
20160228 00:17:40< Aginor> anything that increases readability and maintainability has my approval :)
20160228 00:18:09< Aginor> I figured I wan't the intended person, but I decided to weigh in anyway
20160228 00:18:18< celticminstrel> That's fine.
20160228 00:18:38< Aginor> :)
20160228 00:20:04< celticminstrel> I need to decide what to do if an invalid type is used.
20160228 00:20:30< celticminstrel> The choices are to silently convert to the default (explicit) or panic and print an error message.
20160228 00:22:22-!- irker012 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160228 00:22:22< irker012> wesnoth: Andreas Löf wesnoth:guifixes bdac437efc1e / src/ (video.cpp video.hpp): Fix compilation with SDL1 https://github.com/wesnoth/wesnoth/commit/bdac437efc1e33cad80381b50fc2dde363356c6d
20160228 00:22:39< Aginor> I honestly think panic and error is the right thing
20160228 00:22:47< Aginor> fail early and fail explicitly
20160228 00:23:05< Aginor> instead of silently eating the error and ending up with a weird state
20160228 00:23:58< celticminstrel> Hmm... I probably can't use lua_error or whatever here. I don't think this is a protected context. I'll catch the bad_enum_cast and print an error...
20160228 00:24:07 * mattsc felt a cold shiver go down his spine, as if an Ancient Lich were materializing in a dark corner of the room
20160228 00:24:43< celticminstrel> Since the return value is actually a list of goals, I suppose it's probably easiest to just drop the faulty goal from the list... or I could make it an empty list...
20160228 00:24:45< Aginor> mattsc: nothing as friendly as that :D
20160228 00:24:57< mattsc> I was afraid of that ...
20160228 00:25:04< mattsc> celticminstrel: I don’t prefer using the integer constant.  I don’t really prefer using lua_goals at all, but this is definitely a step in the right direction.
20160228 00:25:37< celticminstrel> Why is this code using a back_insert_iterator...
20160228 00:25:54< celticminstrel> Maybe there's a possibility that elements are left over from an earlier call or something?
20160228 00:26:02< celticminstrel> Though even with that it's kinda strange.
20160228 00:26:49< celticminstrel> Wait, where is the iterator incremented?
20160228 00:27:11< celticminstrel> Does assigning repeatedly to the same iterator continuously insert?
20160228 00:27:29< celticminstrel> I've never used back_insert_iterator except as arguments to <algorithm> stuff...
20160228 00:29:59-!- louis94 [~~louis94@91.178.240.5] has joined #wesnoth-dev
20160228 00:42:04< celticminstrel> Hmm, it's a bit hard to drop only the faulty targets, so I'll go with dropping the whole list.
20160228 00:43:05-!- louis94 [~~louis94@91.178.240.5] has quit [Quit: Konversation terminated!]
20160228 00:44:56-!- horrowind [~Icedove@2a02:810a:8380:834:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160228 00:49:23< celticminstrel> Okay, I went and edited make_enum.hpp. >_>
20160228 00:50:25< celticminstrel> I feel like this is going to be another big recompile. :/
20160228 01:05:48-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 276 seconds]
20160228 01:12:59< vultraz> Aginor: thanks
20160228 01:35:53< vultraz> Aginor: I'm trying to reduce the number of places redraw_everything() is called. do you agree with this?
20160228 01:48:37< irker012> wesnoth: Charles Dang wesnoth:master b655ec8f9b03 / data/campaigns/Two_Brothers/maps/01_Rooting_Out_a_Mage.map: AToTB S1: spruced up map https://github.com/wesnoth/wesnoth/commit/b655ec8f9b032acca12745e929adfa430dee7906
20160228 02:01:21-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160228 02:02:22< irker012> wesnoth: Charles Dang wesnoth:master 1e7abb4615cf / src/menu_events.cpp: Don't trigger a full redraw when toggling label visibility https://github.com/wesnoth/wesnoth/commit/1e7abb4615cf919288aa2b17f214205c6934316b
20160228 02:14:13-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20160228 02:16:06< Aginor> vultraz: wait until after guifixes has gone in. It should eliminate the need to call it externally from _most_ places
20160228 02:16:18< Aginor> the editor may be the one exception
20160228 02:17:53< vultraz> I already eliminated the main use from the editor yesterday
20160228 02:17:59< vultraz> but I'll wait for the rest
20160228 02:18:29-!- travis-ci [~travis-ci@ec2-54-157-222-14.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 02:18:30< travis-ci> wesnoth/wesnoth#8643 (master - b655ec8 : Charles Dang): The build passed.
20160228 02:18:30< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112338035
20160228 02:18:30-!- travis-ci [~travis-ci@ec2-54-157-222-14.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 02:19:39< Aginor> vultraz: I reintroduced it :D
20160228 02:20:38< Aginor> because that's one of the places where it's actually needed. The draw_everything call is recreating the buttons on master, which the editor relies on. I've been removing that behaviour, but I think it's still needed. I may be wrong though
20160228 02:25:43< vultraz> Aginor: but why would you need to recreate the buttons when switching map contexts 
20160228 02:25:53< vultraz> (assuming we're talking about the same block of code here)
20160228 02:26:17< Aginor> because someone made silly design decisions I think
20160228 02:26:37< Aginor> there's too much recreating of buttons and the like going on all over the place
20160228 02:26:46< vultraz> context_manager::refresh_all() is where I removed it
20160228 02:27:16< Aginor> that's where I reintroduced it when I merged master into guifixes because of the conflict
20160228 02:27:38< vultraz> ah.. :(
20160228 02:27:47< Aginor> it may have been a mistake
20160228 02:27:56< Aginor> but I'd rather not jump the shark on it
20160228 02:28:04< vultraz> I'll check after guifixes gets merged
20160228 02:28:16< Aginor> speaking of which, have you tested it?
20160228 02:28:23< vultraz> guifixes?
20160228 02:28:31< Aginor> yes
20160228 02:28:32< Aginor> lately
20160228 02:28:35< vultraz> not since the other day since I was finishing up b655ec8f9b03 
20160228 02:28:45< vultraz> (the atotb map)
20160228 02:28:59< Aginor> I have no idea about any of that work :)
20160228 02:29:03< vultraz> I'll give it a shot now
20160228 02:29:13< Aginor> but I've fixed the undrawing of the dialogs you were complaining about
20160228 02:29:26< Aginor> there may be others, but we can deal with them as they appear
20160228 02:30:33< Aginor> hmm
20160228 02:30:42< Aginor> the new hotkey preferences are looking weird
20160228 02:30:47< Aginor> there's an empty column in there
20160228 02:31:24< vultraz> takes about 15 minutes to build guifixes
20160228 02:31:47< vultraz> so I don't test often
20160228 02:33:47< Aginor> fair enough
20160228 02:33:56< Aginor> I'd appreciate input though
20160228 02:35:37< irker012> wesnoth: Wedge009 wesnoth:master 798a99b27a86 / src/hotkey/hotkey_item.cpp: Avoid polluting preferences file with null commands (Bug #21969) https://github.com/wesnoth/wesnoth/commit/798a99b27a86e5c15c6cd21d35e5b3f44332bf7a
20160228 02:35:39< irker012> wesnoth: Wedge009 wesnoth:master 389285e5803a / src/hotkey/ (command_executor.cpp hotkey_item.cpp hotkey_item.hpp): Add disabled flag to allow clearing of default hot-keys (Bugs #21983/#22218/#239 https://github.com/wesnoth/wesnoth/commit/389285e5803a21e62c8a345578a1219201490b79
20160228 02:35:41< irker012> wesnoth: Wedge009 wesnoth:master 807d78a0b18d / changelog: Update change log. https://github.com/wesnoth/wesnoth/commit/807d78a0b18d75ec5cdb3c3fd8b37288ad4a69be
20160228 02:35:43< irker012> wesnoth: Andreas wesnoth:master a3f6abfc85a7 / / (4 files in 2 dirs): Merge pull request #564 from Wedge009/bug_21969_fix https://github.com/wesnoth/wesnoth/commit/a3f6abfc85a750927b8e484f4bd1d64364584af6
20160228 02:35:57-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20160228 02:36:12-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 260 seconds]
20160228 02:36:20-!- VultCave is now known as vultraz
20160228 02:37:25-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20160228 02:37:25-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160228 02:38:28< vultraz> Aginor: anyway, the reason I wanted to reduce redraw_everything() calls is specifically because of the flicker when everything is redrawn
20160228 02:38:57< vultraz> it looks bad
20160228 02:39:27< Aginor> vultraz: I agree, but there's better ways to solve it
20160228 02:39:36< vultraz> oh?
20160228 02:39:53< Aginor> (not that removing the coupling between that class and other parts of the game is a bad thing)
20160228 02:39:53-!- prkc [~prkc@563BD80A.dsl.pool.telekom.hu] has quit [Read error: Connection reset by peer]
20160228 02:40:10< vultraz> well, I do have to replace them with other calls
20160228 02:40:15< vultraz> usually invalidate_all
20160228 02:40:23< Aginor> the real problem is when the flip method gets called, and that it gets called multiple times in a render cycle
20160228 02:40:41< Aginor> ideally, it should only ever be called at the end of the cycle
20160228 02:40:50< vultraz> ah, yes
20160228 02:40:53< vultraz> I tried to do that
20160228 02:40:55< vultraz> but
20160228 02:41:02< vultraz> I couldn't figure out where to put it
20160228 02:41:02< Aginor> that way you won't get your flickering, which is caused by different layers being updated and then flipped
20160228 02:41:12< vultraz> and do you mean CVideo::flip() or display::flip()
20160228 02:41:20< Aginor> CVideo::flip()
20160228 02:41:30< Aginor> display::flip() should be aquished
20160228 02:41:34< vultraz> yes
20160228 02:41:42< vultraz> in sdl2 i think it basically does nothing
20160228 02:41:53< Aginor> not quite true
20160228 02:41:57< Aginor> but close enough
20160228 02:42:09< Aginor> above all, _it shouldn't be called willy nilly_
20160228 02:42:31< Aginor> wedge009: thank you for your effort with the hotkeys, I appreciate it
20160228 02:43:14< vultraz> Do you think we can clean up this 'sunset' crap sitting around the code?
20160228 02:43:42< Aginor> everything is possible
20160228 02:43:49< Aginor> but I don't think we want to do it right now
20160228 02:44:10< vultraz> ok, guifixes built in 14 minutes
20160228 02:44:18< Aginor> :D
20160228 02:44:33< Aginor> I just recompiled everything and it started my computer beeping about the heat
20160228 02:44:43< vultraz> confirmed fixes with the dialogs
20160228 02:45:16< vultraz> gui1 dialogs still cause flicker, but we'll deal with that
20160228 02:45:49< vultraz> in the editor, that is
20160228 02:46:36< vultraz> in-game they only flicker the buttons, so that's most important
20160228 02:47:06< Aginor> yeah
20160228 02:47:13< vultraz> major flicker when opening Terrain Descriptions...
20160228 02:47:36< Aginor> I can't fix the flickering the way the rendering pipeline is set up
20160228 02:48:17< Aginor> the flickering will happen any time a window/dialog is opened or closed, or when the window is resized or exposed
20160228 02:48:32< Aginor> the fix to the flickering is what I spoke about above
20160228 02:48:36< vultraz> ah
20160228 02:48:40< vultraz> well besides that
20160228 02:48:42< vultraz> stuff looks good
20160228 02:48:45< Aginor> but doing that will take some mucking about with deep internals
20160228 02:48:46< vultraz> I recommend merging 
20160228 02:48:58< Aginor> thank you
20160228 02:49:15< Aginor> celticminstrel: what do you think?
20160228 02:49:31< vultraz> we'll have all week in case something more serious comes up
20160228 02:50:56< Aginor> yes
20160228 02:51:08< Aginor> please keep in mind that my free time is pretty limited though
20160228 02:51:18< Aginor> and this cuts into my free time :)
20160228 02:51:25< Aginor> or counts as it
20160228 02:51:28< Aginor> or whatever
20160228 02:52:26< vultraz> I just tested the reapplication of that commit you reverted, and it seems to reduce flickering in the editor
20160228 02:52:58< vultraz> probably because you introduced redraw routines elsewhere which means redrawing was done twice
20160228 02:53:01< vultraz> (just a randomg uess)
20160228 02:53:09< vultraz> random guess
20160228 02:53:11< Aginor> excellent, and you still have the the buttons?
20160228 02:53:20< vultraz> yes
20160228 02:53:48< Aginor> I actually think all/almost all calls to display::redraw_everything are redundant at this point
20160228 02:54:46< Aginor> but it needs to be removed and tested ever so closely
20160228 02:54:53< vultraz> yeah
20160228 02:54:59< vultraz> each case individually 
20160228 02:55:15< Aginor> celticminstrel: have you had an opportunity to test so far? - what's your opinion? more work or merge?
20160228 02:59:33< vultraz> once it's merged I'll reapply that commit
20160228 03:02:28< vultraz> and I know of at least 1 flip() call that can be removed 
20160228 03:02:49< vultraz> as for the flickering in help, it's because help calls flip() when it opens :|
20160228 03:06:55< vultraz> *twice*
20160228 03:07:31< celticminstrel> Aginor: Sorry - between distractions and slow computer, I still haven't gotten to guifixes. vultraz: Since you're doing campaign stuff, do you want to transform the Inspire ability in THoT to be more like the new implementation of leadership?
20160228 03:08:49< vultraz> perhaps
20160228 03:08:54< vultraz> might not be done before 1.13.3
20160228 03:08:56< vultraz> I'll see
20160228 03:09:28< celticminstrel> If you don't get to it first, I'll probably do it eventually.
20160228 03:09:48-!- travis-ci [~travis-ci@ec2-54-157-222-14.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 03:09:49< travis-ci> wesnoth/wesnoth#8644 (master - 1e7abb4 : Charles Dang): The build passed.
20160228 03:09:49< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112339472
20160228 03:09:49-!- travis-ci [~travis-ci@ec2-54-157-222-14.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 03:26:57-!- legoktm [~quassel@wikipedia/Legoktm] has quit [Read error: Network is unreachable]
20160228 03:27:31-!- legoktm [~quassel@wikipedia/Legoktm] has joined #wesnoth-dev
20160228 03:43:45< celticminstrel> The deprecation message says that [goal]name=protect_my_unit will be removed in version 1.13.0.
20160228 03:44:24< celticminstrel> Should I go ahead and remove it?
20160228 03:51:33-!- travis-ci [~travis-ci@ec2-54-157-222-14.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 03:51:34< travis-ci> wesnoth/wesnoth#8645 (master - a3f6abf : Andreas): The build has errored.
20160228 03:51:34< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112342590
20160228 03:51:34-!- travis-ci [~travis-ci@ec2-54-157-222-14.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 03:54:52< vultraz> celticminstrel: I didn't?
20160228 03:55:13< vultraz> I thought I removed all pre 1.13 deprecated code
20160228 03:55:36< celticminstrel> This one's in the AI engine.
20160228 03:55:44< celticminstrel> Maybe you just missed it?
20160228 03:57:34< vultraz> oh
20160228 03:57:36< vultraz> wait
20160228 03:57:38< vultraz> no, I remember now
20160228 03:57:46-!- legoktm [~quassel@wikipedia/Legoktm] has quit [Quit: No Ping reply in 180 seconds.]
20160228 03:57:53< vultraz> there *was* something in the ai engine but I left it alone because I didn't know how to deal with it
20160228 03:57:55< vultraz> go ahead
20160228 04:10:32-!- legoktm [~quassel@wikipedia/Legoktm] has joined #wesnoth-dev
20160228 04:12:14-!- kidneb [kidneb@not.allthetime.xyz] has quit [Ping timeout: 260 seconds]
20160228 04:13:51< celticminstrel> There's also the testing_ai_default syntax which should be removed at some point... but maybe that can wait a bit longer.
20160228 04:19:39-!- kidneb [kidneb@not.allthetime.xyz] has joined #wesnoth-dev
20160228 04:25:58-!- Elsi_ [~Elsi@luwin.ulrar.net] has quit [Ping timeout: 268 seconds]
20160228 04:29:40-!- kidneb [kidneb@not.allthetime.xyz] has quit [Ping timeout: 268 seconds]
20160228 04:30:40-!- Samual [~Samual@2601:547:1000:e3e2:bc46:4696:c0d9:d7db] has joined #wesnoth-dev
20160228 04:30:59-!- Samual [~Samual@2601:547:1000:e3e2:bc46:4696:c0d9:d7db] has quit [Changing host]
20160228 04:30:59-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20160228 04:33:07-!- Elsi [~Elsi@luwin.ulrar.net] has joined #wesnoth-dev
20160228 04:42:40-!- kidneb [kidneb@not.allthetime.xyz] has joined #wesnoth-dev
20160228 04:44:28-!- Samual [~Samual@xonotic/core-team/Samual] has quit [Ping timeout: 268 seconds]
20160228 04:45:42-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20160228 04:45:49-!- legoktm [~quassel@wikipedia/Legoktm] has quit [Read error: Connection reset by peer]
20160228 04:47:30-!- legoktm [~quassel@wikipedia/Legoktm] has joined #wesnoth-dev
20160228 04:50:01-!- kidneb [kidneb@not.allthetime.xyz] has quit [Ping timeout: 268 seconds]
20160228 04:55:17-!- kidneb [kidneb@not.allthetime.xyz] has joined #wesnoth-dev
20160228 04:57:00< celticminstrel> mattsc: The change to string target types requires me to change the get_targets() function. Should I include an undocumented boolean parameter to tell the function to return integer target types instead of string target types?
20160228 04:57:25< celticminstrel> Or should I just go ahead and break compatibility?
20160228 04:59:38-!- ancestral [~ancestral@97.116.184.84] has joined #wesnoth-dev
20160228 05:02:43< celticminstrel> ancestral: Did you try out the guifixes branch?
20160228 05:04:36< mattsc> celticminstrel: not quite sure I understand; I thought the get_targets() function returns an array of locations.  Are you going to change that too?
20160228 05:05:26-!- kidneb [kidneb@not.allthetime.xyz] has quit [Ping timeout: 268 seconds]
20160228 05:05:41< celticminstrel> mattsc: It looks like it returns an array of targets - the same format as you're expected to return from a Lua goal.
20160228 05:06:15-!- kidneb [kidneb@2a01:7e00::f03c:91ff:fe33:363c] has joined #wesnoth-dev
20160228 05:10:32< mattsc> right
20160228 05:10:53< celticminstrel> At the moment, the type key is an integer. I want to change it to a string.
20160228 05:10:56< mattsc> so I guess I don’t understand what you’re suggesting
20160228 05:10:59-!- Elsi [~Elsi@luwin.ulrar.net] has quit [Ping timeout: 268 seconds]
20160228 05:11:16< celticminstrel> But, doing that would break any code that checked the types of targets from get_targets().
20160228 05:11:56< celticminstrel> Like if you had some code like "get_targets()[1].type == 3".
20160228 05:12:07< mattsc> oh ...
20160228 05:12:24< mattsc> right, type is specified inside that array …
20160228 05:12:27< mattsc> I’d forgotten
20160228 05:12:54< mattsc> Just break compatibility
20160228 05:13:02< celticminstrel> 'kay
20160228 05:13:19< celticminstrel> It can probably get a mention in the release notes, too.
20160228 05:13:42< mattsc> sure
20160228 05:13:54< mattsc> I would be very surprised if anybody had used that though
20160228 05:14:09< mattsc> read the target type and done anything with it, I mean
20160228 05:15:27< mattsc> I just checked, there’s not a single occurrence of get_targets in all of mainline
20160228 05:16:32-!- kidneb [kidneb@2a01:7e00::f03c:91ff:fe33:363c] has quit [Ping timeout: 268 seconds]
20160228 05:16:42< ancestral> celticminstrel: How recent?
20160228 05:16:55< ancestral> Like, last 24 hours?
20160228 05:17:14-!- Elsi [~Elsi@luwin.ulrar.net] has joined #wesnoth-dev
20160228 05:17:17< celticminstrel> Not necessarily that recent.
20160228 05:17:22< celticminstrel> Past few days, I guess?
20160228 05:18:29-!- kidneb [~kidneb@not.allthetime.xyz] has joined #wesnoth-dev
20160228 05:28:21< ancestral> Okay so
20160228 05:28:21< ancestral> celticminstrel: I rebuilt since last time, which I swear was maybe 26 hours ago
20160228 05:29:36< ancestral> Anyway, the stop-reset button works now
20160228 05:29:36< ancestral> But the all-AI crash still happens
20160228 05:29:43-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160228 05:29:56< celticminstrel> So um, is that a yes or a no...
20160228 05:30:04< ancestral> Application Specific Information:
20160228 05:30:04< ancestral> Assertion failed: (teamindex < dc_->teams().size()), function set_team, file /Users/martinproud/Documents/Pilot/wesnoth 1.13/src/display.cpp, line 395.
20160228 05:41:37-!- ancestral [~ancestral@97.116.184.84] has quit [Quit: i go nstuf kthxbai]
20160228 06:02:47-!- molgrum [~molgrum@unaffiliated/molgrum] has quit [Ping timeout: 268 seconds]
20160228 06:04:42-!- molgrum [~molgrum@h-94-103.a230.priv.bahnhof.se] has joined #wesnoth-dev
20160228 06:04:42-!- molgrum [~molgrum@h-94-103.a230.priv.bahnhof.se] has quit [Changing host]
20160228 06:04:42-!- molgrum [~molgrum@unaffiliated/molgrum] has joined #wesnoth-dev
20160228 06:06:43< vultraz> Aginor: ok, I've reduced help screen flicker by removing one of the flip() calls
20160228 06:11:27-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 260 seconds]
20160228 06:21:03-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160228 07:03:37-!- irker012 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160228 07:12:39-!- Kwandulin [~Miranda@p5DDD036A.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160228 07:16:22-!- irker431 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160228 07:16:22< irker431> wesnoth: Charles Dang wesnoth:guifixes d39bbc23f331 / src/editor/map/context_manager.cpp: Reapply 86746aaaf094ece1540d143c8d58c8f9e551e243 https://github.com/wesnoth/wesnoth/commit/d39bbc23f331bb1d60adb01a0d36989dbccbe116
20160228 07:16:23< irker431> wesnoth: Charles Dang wesnoth:guifixes 716ffa524b20 / src/ (game_launcher.cpp help/help.cpp hotkey/hotkey_preferences_display.cpp): Remove unnecessary calls to CVideo::flip() (and one call to CVideo::delay()) https://github.com/wesnoth/wesnoth/commit/716ffa524b2071525c19cddd0a2e0a36ff1a1d50
20160228 07:16:30< vultraz> Aginor: ^ ok, this significantly reduces flicker in general
20160228 07:21:37-!- travis-ci [~travis-ci@ec2-54-157-222-14.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 07:21:38< travis-ci> wesnoth/wesnoth#8646 (guifixes - 716ffa5 : Charles Dang): The build has errored.
20160228 07:21:38< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112363895
20160228 07:21:38-!- travis-ci [~travis-ci@ec2-54-157-222-14.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 07:22:03< celticminstrel> I have one AI-related thing I want to do in the MP player options screen, and then I'll probably check out guifixes.
20160228 07:22:15< celticminstrel> Let it build while I sleep, then try it out tomorrow.
20160228 07:22:54< celticminstrel> Do you happen to know off the top of your head whether scenarios can disable the AI selection dropdown?
20160228 07:23:05 * celticminstrel will take a look in a few minutes if you don't.
20160228 07:23:33< vultraz> afaik you can't
20160228 07:24:06< Aginor> vultraz: have you checked so that animations and water still looks alright?
20160228 07:24:50< vultraz> water?
20160228 07:25:18< vultraz> yes, the water is still fine
20160228 07:25:39< celticminstrel> By animations, do you mean terrain or units?
20160228 07:27:23< Aginor> both
20160228 07:27:49 * Aginor ponders
20160228 07:27:57< vultraz> both look fine
20160228 07:29:50< vultraz> Aginor: I did notice gui1 dialogs don't seem to be affected by window resize events
20160228 07:29:59< vultraz> gui2 does
20160228 07:30:10< Aginor> vultraz: what exactly do you mean?
20160228 07:31:49< vultraz> Aginor: https://www.dropbox.com/s/wsyo9jrgppa13bn/gui1dialogsdontlikeresize.PNG?dl=0
20160228 07:32:28< Aginor> vultraz: that's expected behaviour and I will not fix that because it requires me to rewrite gui1
20160228 07:32:33< vultraz> ok
20160228 07:32:41< vultraz> and there's one other bug I observed
20160228 07:33:46< vultraz> Aginor: https://www.dropbox.com/s/7xcupzqrvbf9wkz/hotkeyprefsissue.PNG?dl=0 in master, that popup doesn't show up. in guifixes, it shows up, but puts the gui1 dialog (hotkeys) under the gui2 one (prefs) and won't dismiss on click
20160228 07:44:32< vultraz> (it's not a merge blocker, but should optimally be fixed prior to release)
20160228 07:46:04< celticminstrel> While I can't see the image, from the description I feel like it probably should be a merge blocker.
20160228 07:48:38< vultraz> celticminstrel: https://drive.google.com/file/d/0B-mR9s8FduLLdWpteW5KamcyYmc/view?usp=sharing
20160228 07:49:03< celticminstrel> I'll look once I've tested this thing I'm testing.
20160228 07:51:41-!- danni [~quassel@203-206-161-81.perm.iinet.net.au] has joined #wesnoth-dev
20160228 08:09:38-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160228 08:14:08< irker431> wesnoth: Celtic Minstrel wesnoth:master e9fc60ff0cb2 / src/make_enum.hpp: Add additional accessors to bad_enum_cast exception https://github.com/wesnoth/wesnoth/commit/e9fc60ff0cb23509327ff82a86a563c314efcdbe
20160228 08:26:48-!- celticminstrel is now known as celmin|sleep
20160228 08:27:27-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20160228 08:28:12< celmin|sleep> Now building guifixes. Will test after sleep.
20160228 08:45:24-!- Kwandulin [~Miranda@p5DDD036A.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160228 08:46:49-!- Samual [~Samual@xonotic/core-team/Samual] has quit [Ping timeout: 268 seconds]
20160228 08:47:29-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20160228 08:48:31-!- travis-ci [~travis-ci@ec2-107-22-51-111.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 08:48:32< travis-ci> wesnoth/wesnoth#8648 (master - e9fc60f : Celtic Minstrel): The build passed.
20160228 08:48:32< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112367562
20160228 08:48:32-!- travis-ci [~travis-ci@ec2-107-22-51-111.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 08:52:22-!- Samual [~Samual@xonotic/core-team/Samual] has quit [Ping timeout: 268 seconds]
20160228 08:57:31-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20160228 09:01:12-!- Appleman1234 [~Appleman1@KD106161153065.au-net.ne.jp] has quit [Remote host closed the connection]
20160228 09:09:52-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160228 09:16:32< Aginor> celmin|sleep, vultraz: that behaviour is completely broken on master, the game just locks up there
20160228 09:16:45< vultraz> yeah
20160228 09:16:49< Aginor> so it'd be a release blocker
20160228 09:17:07< vultraz> guifixes fixes it a little bit but the game still hangs
20160228 09:17:17< vultraz> I'll bump the bug
20160228 09:17:26< Aginor> I'm compiling guifixes now to look at it
20160228 09:17:35< Aginor> or at least start looking at it
20160228 09:17:52< Aginor> I shouldn't complain, but I so want to go and buy a new computer
20160228 09:18:07< Aginor> something that isn't an big honkin' AMD so it runs cooler
20160228 09:18:57< vultraz> new computers are good
20160228 09:19:15< Aginor> money though
20160228 09:19:21< Aginor> it's $2k I don't want to spend
20160228 09:19:27< vultraz> ah
20160228 09:19:38< Aginor> I'd rather put that towards a mortage for a house
20160228 09:19:38< vultraz> yeah, that's a lot
20160228 09:19:45< Aginor> hmm
20160228 09:19:50< vultraz> I spent about $700 of my saving for this shitty laptop I have now :P
20160228 09:19:54< vultraz> savings*
20160228 09:20:24< vultraz> the screen died within a year and the hdd had to be replaced
20160228 09:20:29< Aginor> I'd not get something shitty, I'd probably look at an i7 with 16 or 32gb of RAM
20160228 09:21:03< Aginor> and a nice SSD for system/home
20160228 09:21:20 * Aginor dreams
20160228 09:21:31< Aginor> so yeah, problem reproduced
20160228 09:21:41-!- Kwandulin [~Miranda@p5DDD036A.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160228 09:21:46< Aginor> not sure how to deal with GUI2 ending up on top of GUI1
20160228 09:22:01< Aginor> actually, isn't the hokey dialog a gui2 dialog?
20160228 09:22:31< vultraz> Aginor: I just noticed, in events.cpp pump(), the DRAW_ALL_EVENT case is encapsulated in the bracket block for SDL_MOUSEBUTTONDOWN. Is that deliberate?
20160228 09:22:38< vultraz> Aginor: and no, only prefs is gui2
20160228 09:23:14< Aginor> vultraz: no. that's a bug
20160228 09:23:17< vultraz> Aginor: gfgtf wanted to convert it to gui2 but gui2 has too much indrection and apparently can't get the key input for setting new hotkeys
20160228 09:24:00< Aginor> I'll commit a fix for that bracket
20160228 09:24:04< Aginor> thanks for picking up on it
20160228 09:25:09 * Aginor humms
20160228 09:25:19< Aginor> it seems to be gui2 showmessage that's a problem
20160228 09:25:37-!- iceiceice_ [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20160228 09:27:12< Aginor> hmm
20160228 09:27:20< Aginor> stuck in showing that transient message
20160228 09:27:31< Aginor> I wonder if it's not getting the events
20160228 09:33:10-!- Jetrel_ [~Jetrel@c-73-228-139-39.hsd1.mn.comcast.net] has quit [Ping timeout: 244 seconds]
20160228 09:33:15< Aginor> ok, problem diagnosed...
20160228 09:33:22< Aginor> fix... is harder
20160228 09:33:29< vultraz> what's the problem?
20160228 09:33:56< Aginor> the gui2 event handler is never passed any events
20160228 09:34:01 * Aginor ponders
20160228 09:34:18-!- Samual [~Samual@xonotic/core-team/Samual] has quit [Ping timeout: 268 seconds]
20160228 09:34:18< Aginor> if might be the Right Thing to have GUI2 join the global event context
20160228 09:34:30< Aginor> or that may confuse GUI2 greatly
20160228 09:34:42-!- Samual [~Samual@2601:547:1000:e3e2:bc46:4696:c0d9:d7db] has joined #wesnoth-dev
20160228 09:34:42-!- Samual [~Samual@2601:547:1000:e3e2:bc46:4696:c0d9:d7db] has quit [Changing host]
20160228 09:34:42-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20160228 09:35:04< vultraz> if it does would that allow the hotkey dialog to receive key input for setting new hotkeys?
20160228 09:36:49< Aginor> yes
20160228 09:37:49< vultraz> ahhh :D
20160228 09:37:53< vultraz> that would be beneficial 
20160228 09:38:11-!- Jetrel [~Jetrel@c-73-228-139-39.hsd1.mn.comcast.net] has joined #wesnoth-dev
20160228 09:39:26 * Aginor compiles to try it out
20160228 09:41:52< Aginor> hmm
20160228 09:41:56< Aginor> not ideal
20160228 09:42:31< Aginor> I think that GUI1 dialogs over GUI2 dialogs will always make GUI2 break
20160228 09:42:58< Aginor> which is unfortunate when the GUI1 dialog spawns yet another GUI2 dialog over itself with "show_transient_message"
20160228 09:43:20< vultraz> the joys of dual gui systems :|
20160228 09:43:21< Aginor> the breakage we're seeing is brought on by the GUI2 preferences implementation not having hotkeys
20160228 09:43:28< Aginor> as GUI2 elements
20160228 09:43:58< Aginor> but the GUI1 hotkeys dialog calls GUI2 to show the message
20160228 09:44:07< Aginor> all in all, a big mess
20160228 09:44:16< vultraz> well again, the hotkeys dialog could be made gui2 but it needs to be able to receive the input :/
20160228 09:44:20< Aginor> global event context is not the right solution
20160228 09:44:37< Aginor> make it GUI2 and it'll be on top of the event stack
20160228 09:44:44< Aginor> so it should get the events
20160228 09:45:05< Aginor> hell, add a custom event scope to the class just for getting those events if you have to
20160228 09:45:21< Aginor> (and have that join the global context)
20160228 09:45:36< Aginor> I can help you solve that bit
20160228 09:45:57< vultraz> you should talk to gfgtdf, since he was working on makig it gui2
20160228 09:45:59< vultraz> making
20160228 09:46:17< vultraz> I don't want to touch hotkeys with a ten foot pole :P
20160228 09:46:23< vultraz> I never understood the system at all
20160228 09:46:44< Aginor> I'm tempted to say "neither" but that's not true, I rewrote it
20160228 09:46:56< vultraz> especially how events related to the them ui somehow seem to be the same thing as hotkey commands
20160228 09:46:59< vultraz> theme
20160228 09:47:03< vultraz> I really can't type tonight
20160228 09:48:37< Aginor> it's all implemented the same way
20160228 09:49:30-!- Appleman1234 [~Appleman1@KD106161153065.au-net.ne.jp] has joined #wesnoth-dev
20160228 09:49:39< zookeeper> Jetrel_bot, so, my main idea currently for nym is to not have her as a scout, but as the same "desert youth" unit type as kaleh, with the same advancement options (but with unique sprites just like now, obviously). i'm not sure whether that would feel weird in-game or not.
20160228 09:50:07< vultraz> zookeeper: why?
20160228 09:50:14< Jetrel_bot> Frankly, no, it wouldn't feel weird at all.
20160228 09:50:32< zookeeper> vultraz, because i'm suspicious about making her mounted
20160228 09:50:32< Jetrel_bot> It would be understood by the player that that's the special class for "hero characters".
20160228 09:50:38< vultraz> zookeeper: oh
20160228 09:50:47< vultraz> didn't realize that was a stipulation 
20160228 09:51:04< vultraz> in that case, sure
20160228 09:53:04-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20160228 09:53:34< vultraz> Aginor: so the bottom line is that that dialog has to be gui2
20160228 09:54:41-!- Ivanovic__ [~ivanovic@p579FBD31.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160228 09:54:50-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 244 seconds]
20160228 09:55:14-!- Ivanovic__ is now known as Ivanovic
20160228 09:55:45< Aginor> vultraz: I'm pondering possibilities, but that's the best I can come up with
20160228 09:55:54< vultraz> alright
20160228 09:56:01< vultraz> I'll ask gfgtdf if he still has his code
20160228 09:56:05< Aginor> I'm trying to figure out how it might have worked in the past
20160228 09:56:17< vultraz> in the meantime, I'd like to deal with this flip() stuff
20160228 09:56:23< Aginor> because titlescreen is GUI2, right
20160228 09:56:33< vultraz> yes
20160228 09:56:40< Aginor> vultraz: after guifixes. On a separate branch.
20160228 09:56:47< vultraz> right
20160228 09:56:49< Aginor> it can break everything :)
20160228 09:56:59< Aginor> thoroughly and subtly
20160228 09:57:03-!- Ivanovic [~ivanovic@p579FBD31.dip0.t-ipconnect.de] has quit [Changing host]
20160228 09:57:03-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20160228 09:57:04< vultraz> indeed
20160228 09:57:15< vultraz> at least I've removed the unnecessary ones 
20160228 09:58:10< Aginor> mmm
20160228 09:58:28-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 244 seconds]
20160228 09:58:32< Aginor> ideally nothing should call flip on its own, it should happen at the end of a render cycle
20160228 10:03:49< vultraz> Aginor: so is there anything I can do right now?
20160228 10:04:18< Aginor> redisign the event-system? :D
20160228 10:04:26< Aginor> re-design even
20160228 10:04:39< Aginor> I don't think so
20160228 10:05:17< Aginor> at the moment, this is the context stack: default context, gui2 context, preferences_context
20160228 10:05:34< Aginor> the topmost context only receives events by default
20160228 10:05:46< Aginor> which means the dialog here
20160228 10:06:32< Aginor> It might be possible to twiddle the hotkey preferences to live in the default context instead, but then it'd be broken in the titlescreen
20160228 10:06:48< Aginor> or it may be possible to have it join the gui2 context...
20160228 10:06:50 * Aginor ponders
20160228 10:07:35 * Aginor ponders harder
20160228 10:15:55-!- iceiceice_ [~chris@wesnoth/developer/iceiceice] has quit [Quit: Leaving]
20160228 10:16:32< Aginor> nope, not easily, not without redesigning gui2 dialogs
20160228 10:17:47< irker431> wesnoth: Andreas Löf wesnoth:guifixes 6978296137d0 / src/events.cpp: Move stray curly bracket to right place https://github.com/wesnoth/wesnoth/commit/6978296137d0e046b6a03c67a9431a5a3405af97
20160228 10:17:49< irker431> wesnoth: Andreas Löf wesnoth:guifixes 2ad8939de7fb / src/ (4 files in 4 dirs): Merge branch 'guifixes' of github.com:wesnoth/wesnoth into guifixes https://github.com/wesnoth/wesnoth/commit/2ad8939de7fb714555b205fc359d90b6a32c4050
20160228 10:20:19< Aginor> I don't know how to fix this nicely at this point.
20160228 10:20:29< Aginor> I don't even have an idea for an ugly hack
20160228 10:21:42< vultraz> then maybe you should forget the hack and just wait for the dialog tobe gui2
20160228 10:22:59< Aginor> it doesn't solve the fundamental problem though
20160228 10:23:07< Aginor> it's going to crop up more and more often
20160228 10:23:14< vultraz> why?
20160228 10:23:26< vultraz> the number of gui2 dialogs is going down
20160228 10:23:29< vultraz> er
20160228 10:23:31< vultraz> gui1
20160228 10:23:48< vultraz> Help is the only other one with a gui2 dialog over a gui1 one
20160228 10:23:57< Aginor> as you start to mix in more gui2 elements with gui1, we're more likely to see this behaviour
20160228 10:24:21< vultraz> (but that does work correctly)
20160228 10:24:56< Aginor> I expect there to be more of the same behaviour
20160228 10:25:05< Aginor> in the long run
20160228 10:25:15< vultraz> well I'm trying to get rid of the gui1 cases so this does not come up
20160228 10:25:27-!- Crendgrim [~crend@wesnoth/forum-moderator/crendgrim] has joined #wesnoth-dev
20160228 10:26:21-!- Crendgrim_ [~crend@wesnoth/forum-moderator/crendgrim] has quit [Ping timeout: 244 seconds]
20160228 10:26:31< Aginor> mmm
20160228 10:26:35< Aginor> anyway
20160228 10:26:40< Aginor> I'm real tired
20160228 10:26:44< Aginor> I'm going to go to bed
20160228 10:26:54< vultraz> cya
20160228 10:26:56< Aginor> good night
20160228 10:28:28 * vultraz switches back to master
20160228 10:34:07-!- Elsi [~Elsi@luwin.ulrar.net] has quit [Ping timeout: 268 seconds]
20160228 10:36:18-!- Elsi [~Elsi@2001:41d0:2:d43b::1] has joined #wesnoth-dev
20160228 10:39:54-!- mjs-de [~mjs-de@x4db5c93f.dyn.telefonica.de] has joined #wesnoth-dev
20160228 10:52:36-!- travis-ci [~travis-ci@54.157.222.14] has joined #wesnoth-dev
20160228 10:52:36< travis-ci> wesnoth/wesnoth#8649 (guifixes - 2ad8939 : Andreas Löf): The build has errored.
20160228 10:52:36< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112377854
20160228 10:52:36-!- travis-ci [~travis-ci@54.157.222.14] has left #wesnoth-dev []
20160228 11:45:48-!- prkc [~prkc@563BD80A.dsl.pool.telekom.hu] has joined #wesnoth-dev
20160228 11:50:41 * zookeeper thought that it was possible to make all units of a given type automatically have an [object] by putting [modifications] in the [unit_type], but apparently not
20160228 11:51:45< wedge009> Aginor, vultraz: Was your last discussion about the hot-key reset? It's all to do with gui2 migration, then?
20160228 11:52:00< vultraz> yes
20160228 11:52:14< wedge009> Huh, bugger.
20160228 12:04:40< vultraz> I've narrowed down the cause of that replay crash
20160228 12:10:19-!- Samual [~Samual@xonotic/core-team/Samual] has quit [Ping timeout: 268 seconds]
20160228 12:10:51-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20160228 12:17:45< vultraz> got it
20160228 12:29:12< irker431> wesnoth: Charles Dang wesnoth:master ddb873d492db / src/ (game_state.hpp playsingle_controller.cpp): Fixed assertion in replays with all AI sides (bug #24439) https://github.com/wesnoth/wesnoth/commit/ddb873d492dbebe65e7ac23b2b60420bde1236a9
20160228 12:29:26< vultraz> ancestral ^
20160228 12:29:40< vultraz> also gfgtdf ^
20160228 12:34:02-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160228 12:34:10< iceiceice> vultraz, did you ever play an old 90s PC game called "zone66" ?
20160228 12:34:18< vultraz> nope
20160228 12:34:25< iceiceice> i was randomly reminded of it
20160228 12:34:33< iceiceice> i think, if we still don't have a trailer on April 6,
20160228 12:34:37< vultraz> i wasn't alive for most of the 90's
20160228 12:34:51< iceiceice> should just take this, take the first two minutes, and somewhere in it, put like a big red MsPaint arrow to the guy saying "Konrad"
20160228 12:34:52< iceiceice> https://www.youtube.com/watch?v=9zuwIQ3Y6ck
20160228 12:35:04< iceiceice> that would be kind of funny imo :)
20160228 12:38:42-!- Kwandulin [~Miranda@p5DDD036A.dip0.t-ipconnect.de] has quit [Ping timeout: 276 seconds]
20160228 12:39:03-!- gfgtdf [~chatzilla@f054144101.adsl.alicedsl.de] has joined #wesnoth-dev
20160228 12:40:01< gfgtdf> vultraz: you commit https://github.com/wesnoth/wesnoth/commit/ddb873d492dbebe65e7ac23b2b60420bde1236a9 looks like beraks the 'full map/current team' view combobox in the replay ui n case that there are no human sides, did you test that?
20160228 12:41:58< vultraz> gfgtdf: if there are no human sides it's grayed out
20160228 12:42:02< vultraz> gfgtdf: is that the right behavior?
20160228 12:43:04-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Quit: Leaving]
20160228 12:44:04< gfgtdf> vultraz: well not sure the vision comboox has the options 'Human Team, Current Team, All Teams', the fiurst sureleey doesnt make sense in only ai sides, i never watched replay without human sides so i dont know whether poeple want to switch between the 2other ones
20160228 12:44:41< gfgtdf> vultraz: in wesnoth 1.12 it had the options 'Team 1, Current Team, All Teams' so having all sides ai controlled wasnt really a problem
20160228 12:46:40-!- gfgtdf_ [~chatzilla@f054144101.adsl.alicedsl.de] has joined #wesnoth-dev
20160228 12:47:10< gfgtdf_> celmin|sleep: what was teh reason to change bas_enum_cast execption ?
20160228 12:49:17-!- gfgtdf [~chatzilla@f054144101.adsl.alicedsl.de] has quit [Ping timeout: 260 seconds]
20160228 12:49:51-!- gfgtdf_ is now known as gfgtdf
20160228 12:51:39-!- quentinp_ [~quentin@ns363174.ip-91-121-196.eu] has joined #wesnoth-dev
20160228 12:51:44< gfgtdf> Aginor: i think making it imposible to out gui1 gialog on top of gui1 dialogs would be quite bad unless we put all the dialogs to gui2 which is sureley nontrivial
20160228 12:52:11< gfgtdf> Aginor: spoecialyl the elp menu which is gui1 can appear on a lot of other dialogs
20160228 12:52:33< vultraz> gfgtdf: btw do you still have your code for making the hotkey dialog gui2
20160228 12:52:39< gfgtdf> Aginor: like the titlescreen dialog, recuit/recall dialogs, create unit debug dialog, and custom lua gui2 dialogs
20160228 12:53:52-!- quentinp [~quentin@ns363174.ip-91-121-196.eu] has quit [Ping timeout: 264 seconds]
20160228 12:53:52-!- AI0867 [~ai@wesnoth/developer/ai0867] has quit [Ping timeout: 264 seconds]
20160228 12:53:53-!- celmin|sleep [~celmin@unaffiliated/celticminstrel] has quit [Ping timeout: 264 seconds]
20160228 12:53:53-!- Aginor [~andreas@unaffiliated/aginor] has quit [Ping timeout: 264 seconds]
20160228 12:54:28-!- Gambit [~derek@wesnoth/developer/grickit] has quit [Ping timeout: 264 seconds]
20160228 12:54:29-!- higgins [~higgins@105.ip-167-114-152.net] has quit [Ping timeout: 264 seconds]
20160228 12:54:34-!- Grickit [~derek@wesnoth/developer/grickit] has joined #wesnoth-dev
20160228 12:54:46< gfgtdf> vultraz: well i haveparts of what i had wirtten so far, it push it to my branch wait
20160228 12:55:09< gfgtdf> vultraz: here: https://github.com/gfgtdf/wesnoth-old/commit/7b237a7d3b900f9d262b201217f5501c08754d07
20160228 12:56:03-!- travis-ci [~travis-ci@ec2-54-92-141-248.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 12:56:04< travis-ci> wesnoth/wesnoth#8651 (master - ddb873d : Charles Dang): The build was broken.
20160228 12:56:04< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112390947
20160228 12:56:04-!- travis-ci [~travis-ci@ec2-54-92-141-248.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 12:56:06-!- AI0867 [~ai@wesnoth/developer/ai0867] has joined #wesnoth-dev
20160228 12:56:12< gfgtdf> vultraz: the class 'thotkey_dialog' was actually just the 'press the hotkey to bind or press esc' dialog
20160228 12:56:26< vultraz> hmm
20160228 12:56:41< vultraz> well aginor said it was possible to capture the input somehow
20160228 12:56:54< vultraz> blah build broke
20160228 12:57:56< vultraz> "type qualifiers ignored on function return type" I guess remove the second const?
20160228 13:00:19-!- higgins [~higgins@2607:5300:100:200::160d] has joined #wesnoth-dev
20160228 13:00:27< gfgtdf> vultraz: well one of the problem is that the gui2 SDL_KEY_DOWN events only give the keycode not thta scancode which was the horekey system uses
20160228 13:00:50< gfgtdf> vultraz: hmm i dont know which code you are taking about
20160228 13:01:19< gfgtdf> vultraz: hmm ok now isee
20160228 13:01:26< gfgtdf> vultraz: yes never retuign values as 'const'
20160228 13:01:43< gfgtdf> vultraz: so "bool has_human_sides() const { return first_human_team_ != -1; }" is teh correct code
20160228 13:01:44< irker431> wesnoth: Charles Dang wesnoth:master 540df80ff5cb / src/game_state.hpp: Remove superfluous const (fixup ddb873d492db) https://github.com/wesnoth/wesnoth/commit/540df80ff5cb050894de42704b97cf054714a4ea
20160228 13:02:01< vultraz> shit
20160228 13:02:05< vultraz> did i remove the wrong one :|
20160228 13:02:51< vultraz> dammit!
20160228 13:06:02< gfgtdf> vultraz: the sceond const means 'this function is const, it doesnt change the game_state object' whihe the first const means 'the function returns a const bool you cannot change the boolean value returned by this function' this doesnt realy make sense becasue if you returns by value  the value is copied anyway, so people can usually do what they want with it.
20160228 13:06:33< gfgtdf> vultraz: so the compiler warns about that first const.
20160228 13:06:41< vultraz> ok I'll fixup the fixup
20160228 13:08:30< irker431> wesnoth: Charles Dang wesnoth:master 85430507defe / src/game_state.hpp: Remove the correct superfluous const (fixup 540df80ff5cb) https://github.com/wesnoth/wesnoth/commit/85430507defe81ac568d209d14d01eb0750ab644
20160228 13:11:59-!- higgins [~higgins@2607:5300:100:200::160d] has quit [Ping timeout: 268 seconds]
20160228 13:13:57-!- aeonchild [enchilado@defocus/yummy/enchilado] has quit [Quit: Lost terminal]
20160228 13:14:13-!- aeonchild [enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20160228 13:14:27-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Ping timeout: 268 seconds]
20160228 13:15:04-!- Samual [~Samual@xonotic/core-team/Samual] has quit [Ping timeout: 268 seconds]
20160228 13:15:36< zookeeper> this advancement graph needs a bit more refinement, i think being able to theoretically get to 26-4 melee damage is a bit much...
20160228 13:16:37< gfgtdf> zookeeper: which advacament graph you mean ?
20160228 13:17:28-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20160228 13:17:32-!- travis-ci [~travis-ci@ec2-54-92-141-248.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 13:17:33< travis-ci> gfgtdf/wesnoth-old#606 (master - 7b237a7 : gfgtdf): The build is still failing.
20160228 13:17:33< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth-old/builds/112393720
20160228 13:17:33-!- travis-ci [~travis-ci@ec2-54-92-141-248.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 13:18:32-!- Kwandulin [~Miranda@2003:76:f0b:d4e0:e86d:593a:562f:86e9] has joined #wesnoth-dev
20160228 13:19:11-!- Aginor [~andreas@192.184.90.58] has joined #wesnoth-dev
20160228 13:19:11-!- Aginor [~andreas@192.184.90.58] has quit [Changing host]
20160228 13:19:11-!- Aginor [~andreas@unaffiliated/aginor] has joined #wesnoth-dev
20160228 13:19:16-!- higgins [~higgins@105.ip-167-114-152.net] has joined #wesnoth-dev
20160228 13:19:50< gfgtdf> zookeeper: you can tell which advances i need to kae to get 24 melee damage ?
20160228 13:22:46-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160228 13:25:38-!- travis-ci [~travis-ci@ec2-107-22-51-111.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 13:25:39< travis-ci> wesnoth/wesnoth#8652 (master - 540df80 : Charles Dang): The build is still failing.
20160228 13:25:39< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112394263
20160228 13:25:39-!- travis-ci [~travis-ci@ec2-107-22-51-111.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 13:25:58-!- Aginor [~andreas@unaffiliated/aginor] has quit [Read error: Network is unreachable]
20160228 13:26:20-!- Aginor [~andreas@apollo.alternating.net] has joined #wesnoth-dev
20160228 13:26:20-!- Aginor [~andreas@apollo.alternating.net] has quit [Changing host]
20160228 13:26:20-!- Aginor [~andreas@unaffiliated/aginor] has joined #wesnoth-dev
20160228 13:26:56< zookeeper> gfgtdf, yeah... all of them :p
20160228 13:27:18< zookeeper> which, to be fair, should be fairly impossible unless you more or less dedicate your playthrough to it
20160228 13:27:31-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160228 13:27:42< zookeeper> and it's for kaleh's new advancement tree, since i'm scrapping the old one
20160228 13:40:26< vultraz> gfgtdf: do you mind if I rename the 'mp_replay_' variable to 'replay_'
20160228 13:43:46< gfgtdf> zookeeper: so you ak about an advancmenrt tree that wasnt commited yet ?
20160228 13:43:54< gfgtdf> vultraz: y you can do that
20160228 13:44:09< gfgtdf> talk
20160228 13:44:11< gfgtdf> *
20160228 13:44:45< gfgtdf> zookeeper: the odl advancement tree is designed so that whenyn you take smeadvancemnts other bacome impossible
20160228 13:44:54< gfgtdf> zookeeper: is that also the case for the new advancement tree ?
20160228 13:49:47-!- travis-ci [~travis-ci@ec2-54-92-141-248.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 13:49:47< travis-ci> wesnoth/wesnoth#8653 (master - 8543050 : Charles Dang): The build was fixed.
20160228 13:49:47< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112395044
20160228 13:49:47-!- travis-ci [~travis-ci@ec2-54-92-141-248.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 13:55:44< zookeeper> gfgtdf, yes, uncommitted, part of the whole unitset redesign.
20160228 13:56:13< zookeeper> and yes, i currently have mutually exclusive advancements, although maybe less so than in the current one.
20160228 13:58:02< irker431> wesnoth: Charles Dang wesnoth:master 944f68b13683 / src/ (playsingle_controller.cpp playsingle_controller.hpp): Renamed the replay variable since it's no longer mp-exclusive https://github.com/wesnoth/wesnoth/commit/944f68b136831342c97997cbf6d6aa1ddb2c8156
20160228 14:22:54< zookeeper> vultraz, looks like [effect] apply_to=status remove= doesn't actually accept a list unlike the wiki says
20160228 14:23:17< zookeeper> seems like the kind of thing you'd be interested in
20160228 14:33:32< zookeeper> ah, maybe i can resurrect the ancient charisma ability kaleh used to have...
20160228 14:34:00< zookeeper> (that is, adjacent units pay no upkeep)
20160228 14:38:27< irker431> wesnoth: gfgtdf wesnoth:master 39b61a081b66 / src/ (replay.cpp replay.hpp replay_controller.cpp): attempt to fix replay 'play turn' and 'play side turn' https://github.com/wesnoth/wesnoth/commit/39b61a081b664473db778d9fbaa4e37e9601f74e
20160228 14:49:06< vultraz> zookeeper: https://github.com/Vultraz/wesnoth/commit/cefcbfc77f75de4ec2b8c3afeb56877f553a50d8
20160228 14:49:12< vultraz> zookeeper: could you see if that works
20160228 14:51:04-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160228 14:51:55-!- Ivanovic_ [~ivanovic@p579FBD31.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160228 14:52:28-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 244 seconds]
20160228 14:52:42< vultraz> zookeeper: derp >_> typo. https://github.com/Vultraz/wesnoth/commit/4f4d85385e93b044b9d082f7bde228682def84aa
20160228 14:53:29-!- TheJJ [~rofl@ipbcc36ea9.dynamic.kabel-deutschland.de] has quit [Ping timeout: 244 seconds]
20160228 14:53:40-!- TheJJ [~rofl@ipbcc36ea9.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160228 14:53:51-!- Ivanovic_ is now known as Ivanovic
20160228 14:54:41< zookeeper> vultraz, yeah, in a moment
20160228 14:55:02-!- travis-ci [~travis-ci@ec2-107-22-51-111.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 14:55:03< travis-ci> wesnoth/wesnoth#8654 (master - 944f68b : Charles Dang): The build has errored.
20160228 14:55:03< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112400358
20160228 14:55:03-!- travis-ci [~travis-ci@ec2-107-22-51-111.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 15:10:47< gfgtdf> vultraz: i dont think you need the add.empty() and remove.empty() if you put a loop after it anyway
20160228 15:11:01< gfgtdf> vultraz: i man if they are empty then the loop will just be exeuted 0 times
20160228 15:12:14-!- Kwandulin [~Miranda@2003:76:f0b:d4e0:e86d:593a:562f:86e9] has quit [Ping timeout: 268 seconds]
20160228 15:14:37-!- Kwandulin [~Miranda@p200300760F0BD4E0E86D593A562F86E9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160228 15:17:40< vultraz> good point
20160228 15:32:25< zookeeper> vultraz, uh, didn't seem to work.
20160228 15:32:36< vultraz> ....weird
20160228 15:32:57< zookeeper> oh wait
20160228 15:33:04< zookeeper> you didn't commit anything
20160228 15:33:14< vultraz> no it's on my branch :/
20160228 15:33:20< vultraz> I wanted you to test it before I pushed
20160228 15:33:22< vultraz> to master
20160228 15:36:02< zookeeper> might take a bit, pressed ctrl-a in the wrong place and now firefox isn't responding
20160228 15:36:34< zookeeper> and i don't want to kill it either, so i'll wait and see if it comes back to life
20160228 15:36:52< zookeeper> which probably gives you about enough time to test it yourself :>
20160228 15:37:21< zookeeper> (easy to test with default AMLA)
20160228 15:37:21< vultraz> nope :P
20160228 15:37:27< vultraz> I'm not going to come up up with a test case
20160228 15:37:53< zookeeper> edit default AMLA, poison+slow a unit, AMLA it -> done
20160228 15:40:18< zookeeper> ah well, had to kill it. i'll test it.
20160228 15:42:30< zookeeper> yeah, worked
20160228 15:43:53< vultraz> alright, I'll push
20160228 15:44:19< irker431> wesnoth: Charles Dang wesnoth:master 0d2fe9b32251 / src/unit.cpp: Made [effect] apply_to=status modifications take comma separated lists https://github.com/wesnoth/wesnoth/commit/0d2fe9b322513e79770045632d5a12313dbebb65
20160228 15:46:12-!- louis94 [~~louis94@91.178.242.91] has joined #wesnoth-dev
20160228 16:15:23-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160228 16:57:00-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160228 17:23:25< zookeeper> something's wrong with the loyal trait...
20160228 17:25:35< zookeeper> start UtBS, and you have an upkeep cost of 1 from nym, even though she's got the loyal trait. advance her, and the upkeep drops to 0 as it should be.
20160228 17:26:15-!- prkc [~prkc@563BD80A.dsl.pool.telekom.hu] has quit [Remote host closed the connection]
20160228 17:26:52< zookeeper> it's not a side initialization thing apparently, as it doesn't matter whether she's part of the side definition or placed in an event.
20160228 17:28:57< zookeeper> oddly enough, she has upkeep=full in both cases (which isn't wrong), so it's not like someone has changed some code to naively check for that. i dunno why advancement fixes it, when i can't see anything actually changing in the unit.
20160228 17:30:26< zookeeper> oh wait, my bad. it gets changed to upkeep=loyal on advancement.
20160228 17:31:23< zookeeper> so apparently the problem is that it doesn't get applied at unit creation time. or that someone changed the upkeep calculation logic to only look at upkeep=.
20160228 17:31:27< celticminstrel> ...why is there a "merge guifixes into guifixes" commit...
20160228 17:32:48< zookeeper> someone heard you like guifixes
20160228 17:33:06< zookeeper> so they pu... nevermind
20160228 17:33:09< celticminstrel> gfgtdf: I wanted it for some of my AI stuff and figured it might be more generally useful as well.
20160228 17:33:31< celticminstrel> Argh Firefox, why can't you just open the stupid logfile instead of telling me to download it.
20160228 17:33:46< gfgtdf> celticminstrel: hmm ok, it looked strange to be cause i saw no code that used it.
20160228 17:34:07< gfgtdf> celticminstrel: its your fault becasue you use invalid utf8 chracters in chat :)
20160228 17:34:31< celticminstrel> Yes, but it's also Firefox's fault for panicking when it sees those characters.
20160228 17:36:23< gfgtdf> celticminstrel: about the invalid bad_enum_cast: you get bad_enum_cast thrown often, you should consider using enum.parse(string)
20160228 17:36:47< celticminstrel> Eh?
20160228 17:40:12-!- gfgtdf_ [~chatzilla@f054144101.adsl.alicedsl.de] has joined #wesnoth-dev
20160228 17:41:50< zookeeper> gfgtdf, i _think_ the abovementioned loyal/upkeep bug is caused by 5deea2eb1
20160228 17:42:09< zookeeper> haven't verified, but as it changes a lot of the upkeep handling, it seems like a good guess
20160228 17:43:45-!- gfgtdf__ [~chatzilla@x55b19b7e.dyn.telefonica.de] has joined #wesnoth-dev
20160228 17:44:33-!- gfgtdf [~chatzilla@f054144101.adsl.alicedsl.de] has quit [Ping timeout: 268 seconds]
20160228 17:44:35-!- gfgtdf_ [~chatzilla@f054144101.adsl.alicedsl.de] has quit [Ping timeout: 248 seconds]
20160228 17:44:40-!- gfgtdf__ is now known as gfgtdf
20160228 17:44:41< celticminstrel> zookeeper: So Kaleh will become a unit line rather than a single unit type?
20160228 17:45:30< zookeeper> celticminstrel, no, not really. only the level number changes, just like now.
20160228 17:45:35< celticminstrel> Oh.
20160228 17:45:43< celticminstrel> Well, okay.
20160228 17:46:16< celticminstrel> Alright, guifixes testing time.
20160228 17:47:09< celticminstrel> Oh, right, I should make sure my guifixes is up-to-date first. I think I saw another commit in here.
20160228 17:52:55< Aginor> celticminstrel: good idea :)
20160228 17:55:22< celticminstrel> zookeeper: So, do we get to see this new tree at some point?
20160228 17:57:01< zookeeper> yeah, as soon as i stop hitting this or that bug :p
20160228 17:58:38< celticminstrel> Yay.
20160228 18:04:35< zookeeper> it doesn't really have anything that the current one doesn't have, mostly it just has less but in a clearer and nicer format (i think).
20160228 18:18:21< zookeeper> gfgtdf, noticed what i said between your rejoins?
20160228 18:29:11-!- Grickit is now known as Gambit
20160228 18:30:11-!- Kwandulin [~Miranda@p200300760F0BD4E0E86D593A562F86E9.dip0.t-ipconnect.de] has quit [Ping timeout: 268 seconds]
20160228 18:36:02-!- Kwandulin [~Miranda@p200300760F0BD4E0E86D593A562F86E9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160228 18:43:52-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160228 18:44:30-!- irker431 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160228 18:47:57< gfgtdf> zookeeper: no i didn't
20160228 18:48:32< gfgtdf> zookeeper: i assumed the problem was solved by '20160228 17:30:26< zookeeper> oh wait, my bad. it gets changed to upkeep=loyal on advancement.'
20160228 18:48:58< zookeeper> no that was just about the upkeep= attribute
20160228 18:51:18< celticminstrel> I don't suppose anyone has the guifixes PR link readily available.
20160228 18:51:40< zookeeper> gfgtdf, so as i said i think the problem is that when you create a unit with the loyal trait, it doesn't get upkeep=loyal.
20160228 18:52:20-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160228 18:57:20< gfgtdf> zookeeper: hmm k i think isee the problem
20160228 18:57:49< celticminstrel> gfgtdf: You never explained your enum.parse comment.
20160228 18:57:52< gfgtdf> zookeeper: the unit constructor changes upkeep_ to what is set in the cfg even if its not set there.
20160228 18:58:27< gfgtdf> celticminstrel: well, i did but then i got dc so that you didnt receive it.
20160228 18:58:33-!- louis94 [~~louis94@91.178.242.91] has quit [Ping timeout: 268 seconds]
20160228 18:58:34< celticminstrel> Oh.
20160228 18:59:39< gfgtdf> celticminstrel: ill reporst but afte rive comited rthe upkeep_ thing
20160228 19:01:47< gfgtdf> celticminstrel: beacsue that comit migth change related line numbers
20160228 19:01:48< zookeeper> gfgtdf, ok, cool
20160228 19:08:40-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20160228 19:08:46-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20160228 19:15:34-!- irker494 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160228 19:15:34< irker494> wesnoth: Celtic Minstrel wesnoth:guifixes 9965f9e64424 / src/events.hpp: Fix missing prototype errors caused by typos https://github.com/wesnoth/wesnoth/commit/9965f9e644247c8f595b2dc994fa96c74ea7cc3f
20160228 19:17:39< celticminstrel> Aginor: Should I report cases of flicker, or is that going to be handled more generally?
20160228 19:19:01< celticminstrel> Not sure if it's guifixes related, but I'm getting a WML error when clicking the Profile button in the Recruit dialog. The selected recruit was a goblin, I think.
20160228 19:19:18< celticminstrel> (It worked fine for elves, though.)
20160228 19:22:15< celticminstrel> GUI2 dialogs are not repositioned if you change the resolution.
20160228 19:23:03-!- Kwandulin_2 [~Miranda@p200300760F0BD4E0E86D593A562F86E9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160228 19:24:27-!- Kwandulin [~Miranda@p200300760F0BD4E0E86D593A562F86E9.dip0.t-ipconnect.de] has quit [Ping timeout: 268 seconds]
20160228 19:24:57-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160228 19:24:59< irker494> wesnoth: gfgtdf wesnoth:master 4a1b330279ae / src/team.cpp: rewrite team::change_controller_by_wml to use nonthrowing functions https://github.com/wesnoth/wesnoth/commit/4a1b330279ae9462dd70de0e574f23e0339ed726
20160228 19:25:01< irker494> wesnoth: gfgtdf wesnoth:master f60ee0b48f31 / src/unit.cpp: fix upkeep= parsing in unit constructor https://github.com/wesnoth/wesnoth/commit/f60ee0b48f31ca926c947ead39b8b091a0a430dd
20160228 19:25:03< irker494> wesnoth: gfgtdf wesnoth:master efcfe3b38ae6 / src/ (5 files): reset 'ignore oos messages' when starting a new game https://github.com/wesnoth/wesnoth/commit/efcfe3b38ae64041f57841c60f13958e9534916e
20160228 19:25:04< celticminstrel> Ah, that's not actually true. What I'm seeing here is a problem with the preferences dialog specifically.
20160228 19:25:20< celticminstrel> If I resize using the edges, it works.
20160228 19:25:25< gfgtdf> celticminstrel: if you want to convert a string to an enum and an 'invalid' sting is likeley you shoudl use the parse function of the enum class liek it is used here: https://github.com/wesnoth/wesnoth/blob/master/src/team.cpp#L523
20160228 19:26:03< celticminstrel> Hmm.
20160228 19:26:23< celticminstrel> I think for my purpose the exception actually works better, though.
20160228 19:26:49< celticminstrel> I'll keep that in mind, however.
20160228 19:28:15< celticminstrel> Aginor: So, if I change the resolution to a preset value in the preferences dialog, the dialog is not repositioned onscreen. Do you know how to fix that?
20160228 19:29:16-!- louis94 [~~louis94@91.178.242.91] has joined #wesnoth-dev
20160228 19:29:25< Aginor> celticminstrel: That's a limitation of GUI1, I'm not going to try to fix that
20160228 19:29:36< celticminstrel> No, vultraz converted it to GUI2.
20160228 19:29:52< Aginor> hmm
20160228 19:30:04< Aginor> I don't think that has anything to do with my changes though
20160228 19:30:06< celticminstrel> I still have no TC. :(
20160228 19:30:13< Aginor> so I'm loath to take responsibility
20160228 19:30:18< celticminstrel> No, it probably doesn't have anything to do with your changes.
20160228 19:30:19< Aginor> flicker I'll deal with generally
20160228 19:30:24< celticminstrel> But do you know how it would be fixed?
20160228 19:30:31< Aginor> nope :D
20160228 19:30:37< celticminstrel> No ideas even?
20160228 19:30:41< Aginor> I'd have to dig through GUI2 to find out
20160228 19:30:52< celticminstrel> Okay.
20160228 19:30:53< Aginor> it receives the event, but if it's not repositioning the dialog, it should
20160228 19:31:01< Aginor> I know very little about GUI2
20160228 19:31:35< celticminstrel> Is the resize event sent when it's resized by code or only when it's resized by the user?
20160228 19:31:44< celticminstrel> Maybe it needs to manually trigger a resize event.
20160228 19:31:53< celticminstrel> I'll see if that helps.
20160228 19:32:33< celticminstrel> Anyway, I'm not entirely sure what to look for, but apart from flicker I haven't seen anything bad.
20160228 19:32:50< Aginor> it should already receive a resize event
20160228 19:32:58< Aginor> celticminstrel: that means it's good :)
20160228 19:33:23< Aginor> resizing, re-exposing windows and the like
20160228 19:33:53< Aginor> did you see vultraz's and my discussion from ~9h ago about hang when resetting hotkeys?
20160228 19:34:12< celticminstrel> Oh yeah, I forgot to try out the hotkeys dialog. Should I?
20160228 19:34:19< Aginor> it'll hang
20160228 19:34:22< Aginor> it's a bug from master
20160228 19:34:32< celticminstrel> So, I shouldn't?
20160228 19:34:38< Aginor> GUI1 and GUI2 playing poorly with each other
20160228 19:34:41< celticminstrel> I skimmed through that discussion.
20160228 19:34:44< Aginor> no, there's no point
20160228 19:35:30< celticminstrel> Then I'll go back to AI stuff.
20160228 19:35:34< Aginor> ok
20160228 19:35:44< Aginor> do you think the branch is fit for merge to master?
20160228 19:36:19< Aginor> and thank you :)
20160228 19:36:25< celticminstrel> Probably?
20160228 19:36:42< celticminstrel> If you mean to merge it very soon, I'll wait a few minutes before going back to AI.
20160228 19:36:56< Aginor> I'll go press the button
20160228 19:37:16-!- Kwandulin_2 [~Miranda@p200300760F0BD4E0E86D593A562F86E9.dip0.t-ipconnect.de] has quit [Read error: Network is unreachable]
20160228 19:37:30< celticminstrel> It'd be great if the hotkeys bug was fixed by Friday.
20160228 19:37:40< celticminstrel> The prefs resize one isn't as much of a problem.
20160228 19:37:55-!- Kwandulin [~Miranda@p200300760F0BD4E0E86D593A562F86E9.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160228 19:38:03< celticminstrel> But I might try something later, after I've run out of AI things to do, or if I get tired of AI things.
20160228 19:39:50< irker494> wesnoth: Andreas Löf wesnoth:master ac242376e914 / src/ (display.cpp display.hpp sdl/utils.cpp widgets/button.cpp): Add a dynamic layout functionality to the game display https://github.com/wesnoth/wesnoth/commit/ac242376e91463f27b22fde7ca617d4a5c92fbeb
20160228 19:39:52< irker494> wesnoth: Andreas Löf wesnoth:master 5f5a334fba5b / src/ (14 files in 5 dirs): Make window events a separate handler for sdl_events https://github.com/wesnoth/wesnoth/commit/5f5a334fba5b318be5075517307c3f3c5f763f8d
20160228 19:39:54< irker494> wesnoth: Andreas Löf wesnoth:master a5f8d62a75d9 / src/ (9 files in 4 dirs): Add initial support for GUI layering https://github.com/wesnoth/wesnoth/commit/a5f8d62a75d979ed897b4e5a3271016827f68bd0
20160228 19:39:56< irker494> wesnoth: Andreas Löf wesnoth:master a02a8da7ea3b / src/ (7 files in 2 dirs): Continued improvements to layered rendering https://github.com/wesnoth/wesnoth/commit/a02a8da7ea3b3b974b976f9ec2aeae687b1fa880
20160228 19:39:58< irker494> wesnoth: Andreas Löf wesnoth:master b73cc895ae79 / src/ (events.cpp show_dialog.cpp widgets/widget.cpp widgets/widget.hpp): Fix GUI1 layering and redraws. https://github.com/wesnoth/wesnoth/commit/b73cc895ae79843d830f4bdbd58d65b0460802ad
20160228 19:40:00< irker494> wesnoth: Andreas Löf wesnoth:master 5819a3e78d80 / src/gui/widgets/ (window.cpp window.hpp): Make GUI2 windows honor the restore flag passed to show() https://github.com/wesnoth/wesnoth/commit/5819a3e78d803bad5afd039e17056347e44d75d1
20160228 19:40:02< irker494> wesnoth: Andreas Löf wesnoth:master c514f31cff6d / src/gui/dialogs/dialog.hpp: Fix GUI2 dialogs undrawing themselves on the wrong position https://github.com/wesnoth/wesnoth/commit/c514f31cff6d3033dd0bb4b2025cdeff71c48c1a
20160228 19:40:04< irker494> wesnoth: Andreas Löf wesnoth:master ea649ec0a2c2 / src/ (display.cpp display.hpp game_display.cpp game_display.hpp): Fix event-context related assertion failure at exit https://github.com/wesnoth/wesnoth/commit/ea649ec0a2c207aac8afaf225cdf45c319073025
20160228 19:40:06< irker494> wesnoth: Andreas Löf wesnoth:master a275ea9053d8 / src/gui/auxiliary/event/ (dispatcher.hpp distributor.hpp handler.cpp): Ensure that GUI2 windows trigger the redraw-events when closed https://github.com/wesnoth/wesnoth/commit/a275ea9053d886e903b88653a3014247c19b9daa
20160228 19:40:08< irker494> wesnoth: Andreas Löf wesnoth:master b8f18919dd6b / / (1518 files in 158 dirs): Merge remote-tracking branch 'origin/master' into guifixes https://github.com/wesnoth/wesnoth/commit/b8f18919dd6bc1762bd249210b62ee29db1339c6
20160228 19:40:10< irker494> wesnoth: Andreas Löf wesnoth:master 7565ae00d6a9 / src/widgets/widget.cpp: Ensure that GUI1 widgets are drawn on a DRAW_ALL_EVENT https://github.com/wesnoth/wesnoth/commit/7565ae00d6a9b4db24f5a8cb83b3dbcf3ad6e52a
20160228 19:40:12< irker494> wesnoth: Andreas Löf wesnoth:master d00e20720273 / src/ (display.cpp events.cpp storyscreen/render.cpp storyscreen/render.hpp video.cpp): Make storyscreen respect draw-all events https://github.com/wesnoth/wesnoth/commit/d00e207202735498841be866c48f30c3e8858a03
20160228 19:40:14< irker494> wesnoth: Andreas Löf wesnoth:master 12d1fd006fd3 / data/themes/_initial.cfg src/display.cpp src/display.hpp src/widgets/button.cpp: Resize action buttons dynamically on the main display https://github.com/wesnoth/wesnoth/commit/12d1fd006fd359b66d044d0ad36679dbd8e3c892
20160228 19:40:16< irker494> wesnoth: Andreas Löf wesnoth:master c0f7fb6be25e / / (206 files in 39 dirs): Merge remote-tracking branch 'origin/master' into guifixes https://github.com/wesnoth/wesnoth/commit/c0f7fb6be25e9d10ca054d6ea00cc703d520e5b2
20160228 19:40:18< irker494> wesnoth: Andreas Löf wesnoth:master b576e1ae9081 / / (6 files in 6 dirs): Ensure that the map-editor still works https://github.com/wesnoth/wesnoth/commit/b576e1ae9081379222483713f85eb50130140001
20160228 19:40:20< irker494> wesnoth: Andreas Löf wesnoth:master cc760cb33fa4 / src/ (5 files in 2 dirs): Unable undrawing of dialogs used in the title-screen https://github.com/wesnoth/wesnoth/commit/cc760cb33fa49957bc354146bb12027a7b368b25
20160228 19:40:22< irker494> wesnoth: Andreas Löf wesnoth:master 46a0eb0e558f / / (84 files in 37 dirs): Merge remote-tracking branch 'origin/master' into guifixes https://github.com/wesnoth/wesnoth/commit/46a0eb0e558f2d705b6a5dfe8b55dd2a91e59b98
20160228 19:40:24< irker494> wesnoth: Andreas Löf wesnoth:master bdac437efc1e / src/ (video.cpp video.hpp): Fix compilation with SDL1 https://github.com/wesnoth/wesnoth/commit/bdac437efc1e33cad80381b50fc2dde363356c6d
20160228 19:40:26< irker494> wesnoth: Charles Dang wesnoth:master d39bbc23f331 / src/editor/map/context_manager.cpp: Reapply 86746aaaf094ece1540d143c8d58c8f9e551e243 https://github.com/wesnoth/wesnoth/commit/d39bbc23f331bb1d60adb01a0d36989dbccbe116
20160228 19:40:29< irker494> wesnoth: Charles Dang wesnoth:master 716ffa524b20 / src/ (game_launcher.cpp help/help.cpp hotkey/hotkey_preferences_display.cpp): Remove unnecessary calls to CVideo::flip() (and one call to CVideo::delay()) https://github.com/wesnoth/wesnoth/commit/716ffa524b2071525c19cddd0a2e0a36ff1a1d50
20160228 19:40:32< irker494> wesnoth: Andreas Löf wesnoth:master 6978296137d0 / src/events.cpp: Move stray curly bracket to right place https://github.com/wesnoth/wesnoth/commit/6978296137d0e046b6a03c67a9431a5a3405af97
20160228 19:40:34< irker494> wesnoth: Andreas Löf wesnoth:master 2ad8939de7fb / src/ (4 files in 4 dirs): Merge branch 'guifixes' of github.com:wesnoth/wesnoth into guifixes https://github.com/wesnoth/wesnoth/commit/2ad8939de7fb714555b205fc359d90b6a32c4050
20160228 19:40:36< irker494> wesnoth: Celtic Minstrel wesnoth:master 9965f9e64424 / src/events.hpp: Fix missing prototype errors caused by typos https://github.com/wesnoth/wesnoth/commit/9965f9e644247c8f595b2dc994fa96c74ea7cc3f
20160228 19:40:38< irker494> wesnoth: Andreas wesnoth:master 504ed49e01ae / src/ (36 files in 12 dirs): Merge pull request #595 from wesnoth/guifixes https://github.com/wesnoth/wesnoth/commit/504ed49e01aeb9cc46400594f7bac7eeac2dcdf8
20160228 19:40:52< Aginor> celticminstrel: all merged
20160228 19:41:08< Aginor> celticminstrel: I'll think about how to fix the hotkey thing, but I'm a bit stumped
20160228 19:41:16< Aginor> celticminstrel: this is another busy week for me too
20160228 19:46:34-!- travis-ci [~travis-ci@ec2-54-92-141-248.compute-1.amazonaws.com] has joined #wesnoth-dev
20160228 19:46:35< travis-ci> wesnoth/wesnoth#8657 (guifixes - 9965f9e : Celtic Minstrel): The build passed.
20160228 19:46:35< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/112444478
20160228 19:46:35-!- travis-ci [~travis-ci@ec2-54-92-141-248.compute-1.amazonaws.com] has left #wesnoth-dev []
20160228 19:48:09< Aginor> celticminstrel: thank you for fixing my spelling mistake :)
20160228 19:54:43-!- elias [~allefant@allegro/developer/allefant] has quit [Ping timeout: 244 seconds]
20160228 19:55:13-!- elias [~allefant@allegro/developer/allefant] has joined #wesnoth-dev
20160228 20:04:13< gfgtdf> Aginor: does master now have aproblem with gui1 gialog on top o gui2 dialogs ?
20160228 20:04:41< Aginor> gfgtdf: it already did
20160228 20:04:46< celticminstrel> gfgtdf: Apparently - try the hotkeys dialog.
20160228 20:04:58< Aginor> gfgtdf: and try it before the merge
20160228 20:05:16 * Aginor ponders
20160228 20:05:32< Aginor> I think that may always have been broken to be honest
20160228 20:05:45< Aginor> but I'm too lazy to do the legwork to find out
20160228 20:05:46< gfgtdf> Aginor: what exactly is broken ?
20160228 20:06:03< Aginor> gfgtdf: read the backlog
20160228 20:06:17< Aginor> I explained the problem there
20160228 20:06:31< Aginor> It comes down to the event-system's implementation
20160228 20:07:04< Aginor> I will be away for a bit, cleaning cat-puke
20160228 20:08:30-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160228 20:11:12< gfgtdf> Aginor: i cant see in teh backlog how the problem shows up all i found were expires image links posted by vultraz
20160228 20:11:58-!- Ivanovic [~ivanovic@p579FBD31.dip0.t-ipconnect.de] has quit [Changing host]
20160228 20:11:58-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20160228 20:12:45< celticminstrel> gfgtdf: The image shows the preferences dialog, with the "Hotkeys reset" popup over it, and it's clear that the hotkeys dialog itself is behind the prefs dialog.
20160228 20:13:40< mattsc> celticminstrel: Creating_Custom_AIs is ready
20160228 20:13:59< Aginor> gfgtdf: press the "reset to default" button
20160228 20:14:04< celticminstrel> mattsc: Were there others that I didn't proofread yet?
20160228 20:14:06< Aginor> in the hotkeys preferences
20160228 20:14:07< mattsc> except for the one section marked as missing, but that one will require more work than I will likely have time for before traveling starts again mid-week.
20160228 20:14:20< mattsc> celticminstrel: let me check ...
20160228 20:14:29< celticminstrel> How long is the travelling this time?
20160228 20:15:02< mattsc> just a week; Wed - Thu of the following week
20160228 20:15:26< mattsc> celticminstrel: I don’t think you have checked Experimental_AI and Micro_AIs yet, but I am also not sure if you need to
20160228 20:15:47< gfgtdf> vultraz: do you have an optinion on my gui2 hotkeys dialog (the windows layout i mean) ? I thought tht we could maybe put that part of the hotkeys dialog to gui2 so that only the 'tress the hotkeys now' part ramins gui1
20160228 20:15:57< mattsc> I did not re-read all of Micro_AIs, just removed all the version warning stuff from the into section
20160228 20:16:25< Aginor> gfgtdf: I have an idea how to move the "press the hotkeys" into gui2 too, but that should work fine
20160228 20:16:26< mattsc> s/into/intro
20160228 20:16:40< Aginor> it's invoking gui2::show_message that's the breaker
20160228 20:18:57-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160228 20:20:48-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160228 20:24:03< gfgtdf> Aginor: well but gui2::show_message is used quite often
20160228 20:24:39< gfgtdf> Aginor: specially it can be drawn on top of any ingame gui1 dialog by traing to quit wesnoth
20160228 20:25:53< Aginor> gfgtdf: that's less of a problem
20160228 20:26:09< Aginor> it's the GUI1 dialog that's put on top of GUI2 that breaks stuff
20160228 20:26:34< Aginor> because GUI1 adds its own event context, causing GUI2 to no longer receive events
20160228 20:27:02< gfgtdf> Aginor: hmm ok but it still easy to reproduce that by opening the help froma  gui2 dialog and then pressing abort
20160228 20:27:06< Aginor> having GUI2 on top of GUI1 (like in ingame UI) is fine, it follows the assumptions of GUI1
20160228 20:27:16< gfgtdf> quit the game i mean
20160228 20:39:08< Aginor> gfgtdf: interestingly enough, that works
20160228 20:39:33< Aginor> I'm not suire if the help dialog adds its own event context though
20160228 20:51:06< celticminstrel> I wonder what the point of engine_fai::evaluate is...
20160228 20:53:23< celticminstrel> Okay, so the Lua [engine] runs some code at startup time. I guess this code is only run once, so there's no real reason to compile it and store it somewhere?
20160228 20:55:03< celticminstrel> The engine needs stores two bits of state - one is returned from the startup code and the other is the contents of [data]/
20160228 20:55:07< celticminstrel> ^.
20160228 21:02:12-!- louis94 [~~louis94@91.178.242.91] has quit [Read error: Connection reset by peer]
20160228 21:03:04< zookeeper> gfgtdf, random question: if i use [on_undo] in a moveto event, $unit will still refer to the correct unit when undoing, right?
20160228 21:05:23< gfgtdf> zookeeper: hmm [on_undo] currently doesn't do delayed variable substitution, so that $unit.x or similar will be evaluated at the time when [on_undo] is used.
20160228 21:06:12< zookeeper> gfgtdf, right... so it could point to an arbitrary unit? how about $x1 and $y1?
20160228 21:06:27< zookeeper> (or, if not arbitrary unit, then be empty)
20160228 21:08:58< gfgtdf> zookeeper: hmm im not sure you understood, if you use for example [on_undo] [tag] unit_id = "$unit.id" [/tag][on_undo] then something like [tag] unit_id = "Spearman-15" [/tag] will be stored in the undo stack becasue the variable substitution doesnt happen delayed.
20160228 21:09:22< celticminstrel> I don't understand what's going on in ai/lua/core.cpp. :|
20160228 21:09:39-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Ping timeout: 248 seconds]
20160228 21:10:46< celticminstrel> All sorts of things are being dumped into the Lua registry so that they don't get collected.
20160228 21:10:51< zookeeper> gfgtdf, okay... but doesn't that mean that it will refer to the correct unit when undoing?
20160228 21:11:06< zookeeper> i mean, assuming you don't do something particularly weird
20160228 21:12:41< celticminstrel> I think the comments in lua_ai_context::create are wrong.
20160228 21:13:01< gfgtdf> zookeeper: well yes it does, but it does also mean tthat things like {VARIABLE a 5} {VARIABLE b $a} might behave unexpected becasue $a will be evaluaed at the time when [on_undo] is used.
20160228 21:13:18< zookeeper> gfgtdf, sure, sure. ok, all good then
20160228 21:13:44< celticminstrel> Okay, this makes a bit more sense if I change the comments.
20160228 21:13:54< gfgtdf> zookeeper: the reccmented way is basically  to  keep the code inside [on_undo] rather simple for exampel by setting some vairables and then usinf [fire_event] or something
20160228 21:14:33< zookeeper> gfgtdf, when adding it to the wiki, would be a good idea to mention that variable substitution happens when the event is originally executed
20160228 21:15:06< zookeeper> and just to clarify, i don't think there's anything wrong with that, it's probably the only sensible way it can work anyway
20160228 21:15:25< celticminstrel> mattsc: From the code it sounds like the data field is in the ai table.
20160228 21:15:57< celticminstrel> Which is different from what I think you've told me.
20160228 21:16:05< mattsc> celticminstrel: it is not; it’s in ‘self’ (the engine)
20160228 21:17:23-!- Yaiyan [~Yaiyan@46.101.48.31] has quit [Ping timeout: 244 seconds]
20160228 21:17:23-!- clavii [~clavi@163.172.10.77] has quit [Ping timeout: 244 seconds]
20160228 21:17:50< celticminstrel> mattsc: Is there a possibility that it's in both?
20160228 21:18:02< mattsc> let me check
20160228 21:19:22< mattsc> I don’t see it in the ai table
20160228 21:19:28< mattsc> at least not in this ai table:  wesnoth.debug_ai(wesnoth.current.side).ai
20160228 21:19:36< mattsc> got to be afk, sorry
20160228 21:19:40< celticminstrel> Hm, alright. I could be misreading the code.
20160228 21:19:43< celticminstrel> It's confusing.
20160228 21:20:09< celticminstrel> It almost makes me want to rip it all out and rewrite it so that I can at least understand what's going on. >_>
20160228 21:20:15 * celticminstrel probably won't do that though.
20160228 21:22:39-!- Yaiyan [~Yaiyan@46.101.48.31] has joined #wesnoth-dev
20160228 21:23:17< celticminstrel> So, a candidate action is passed the ai context table only, according to the code.
20160228 21:23:37< celticminstrel> Ohhh...
20160228 21:24:11< vultraz> celticminstrel: what is this as about a gobo in unit attack?
20160228 21:24:40< celticminstrel> This weirdness is locked into the external candidate actions implementation.
20160228 21:24:48< vultraz> gfgtdf: I haven't tried compiling yet, sorry
20160228 21:24:57< vultraz> (the hotkeys dialog yet)
20160228 21:25:08< celticminstrel> They're passed three parameters. The first is ai_context:get_ai(). The second is the exec_params or eval_params. The third is the raw ai_context.
20160228 21:25:08< vultraz> Aginor: glad everything's merged :)
20160228 21:25:38< celticminstrel> So, I think I was indeed reading the code correctly - what's passed as the first parameter for regular Lua candidate actions ends up as the third parameter for external ones.
20160228 21:25:54< celticminstrel> vultraz: Sorry, what?
20160228 21:26:11< vultraz> er, recruit
20160228 21:26:13< vultraz> [06:19:00]	celticminstrel	Not sure if it's guifixes related, but I'm getting a WML error when clicking the Profile button in the Recruit dialog. The selected recruit was a goblin, I think.
20160228 21:26:27< vultraz> recruit is still gui1
20160228 21:26:33< celticminstrel> Oh. Start a game with Northerners, recruit, click Profile.
20160228 21:26:55< celticminstrel> It may be unrelated to GUI stuff.
20160228 21:27:00-!- clavi [~clavi@163.172.10.77] has joined #wesnoth-dev
20160228 21:27:28< vultraz> I'll take a look
20160228 21:28:21< celticminstrel> Wait, the dummy Lua engine is actually quite weird...
20160228 21:28:33< celticminstrel> And implies that the first parameter will have an empty data table.
20160228 21:28:48< celticminstrel> While the third parameter won't have any.
20160228 21:30:12< celticminstrel> I can't figure out where the actual ai context table gets injected into this.
20160228 21:31:33< celticminstrel> Oh wait, duh. (...):get_ai() is equivalent to (...).get_ai((...)), so it's indeed the ai context parameter which fills the ai parameter in the dummy Lua engine.
20160228 21:31:39< celticminstrel> ^context table
20160228 21:32:48< celticminstrel> Sigh. I guess I'll have to actually write a dummy Lua external candidate action to really see what's going on here.
20160228 21:33:52< Aginor> celticminstrel: I take no responsibility for WML errors :p
20160228 21:37:50< zookeeper> gfgtdf, problem: moveto event triggers on a move. on_undo triggers when undoing the move. however, if you redo, the moveto event doesn't trigger.
20160228 21:40:57< vultraz> Aginor: I just noticed a small weird issue: if you have a transient message like Objectives up, and drag the window (not resize it), the background of that window will be full opauqe black for a second before restoring to reansparency
20160228 21:41:31< vultraz> (that = the dialog like Objectives)
20160228 21:42:38< Aginor> vultraz: file a bug so it doesn't get forgotten about, but I don't think it's not a priority at the moemnt
20160228 21:42:51< vultraz> yeah, it's minor
20160228 21:42:54< Aginor> we need to come up with a way to fix your hang instead
20160228 21:43:41< celticminstrel> So uh... I was wrong about one thing. The [engine] code seems to be executed with distressing frequency.
20160228 21:44:21< celticminstrel> Ah, wait, this is a completely different problem, isn't it.
20160228 21:44:32< Aginor> vultraz: I'll tidy up the issuetracker and add a commit to the changelog
20160228 21:44:34< celticminstrel> It's the CA code being recompiled every time, which is both good and bad.
20160228 21:44:55< Aginor> (updating status from "ready for test" to "fixed") later
20160228 21:47:55< celticminstrel> Or maybe I was wrong?
20160228 21:47:59 * celticminstrel confused now.
20160228 21:49:08-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160228 21:50:41< celticminstrel> Is it my imagination, or did the appearance of menus change at some point?
20160228 21:51:13< vultraz> no it is not :)
20160228 21:51:18< vultraz> their backgrounds are slightly lighter
20160228 21:51:50< celticminstrel> They do have a lighter "ring" around them.
20160228 21:52:05< celticminstrel> I'm pretty sure that wasn't the case before.
20160228 21:52:33< vultraz> no, it's the other part that's lighter
20160228 21:57:23< celticminstrel> ...what was the Lua multiline comment syntax again...
20160228 21:57:46< celticminstrel> Oh right, --[[
20160228 21:59:32< celticminstrel> Okay, so my dummy testing Lua candidate action is not working at all.
20160228 21:59:39-!- Kwandulin [~Miranda@p200300760F0BD4E0E86D593A562F86E9.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160228 21:59:43-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160228 21:59:51< celticminstrel> It says that ai.data and self.data are both nil.
20160228 22:01:08< celticminstrel> According to the dummy engine code, ai should have a data field and a get_ai function, but the latter is also nil.
20160228 22:01:33-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Excess Flood]
20160228 22:02:32-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20160228 22:02:34< celticminstrel> Oh, wait, I wasn't accounting for one thing. Hmm...
20160228 22:03:57< irker494> wesnoth: Charles Dang wesnoth:master 800d2d8ab04a / src/dialogs.cpp: Fixed dialogs using unit_preview_pane having their Profile button invisible https://github.com/wesnoth/wesnoth/commit/800d2d8ab04abfff2777fccb65fa7d2b496bc218
20160228 22:04:03< vultraz> Aginor: ^
20160228 22:20:08< Aginor> good :)
20160228 22:21:26< mattsc> celticminstrel: I’m back, at least for a short while.  Anything I can help with?
20160228 22:21:43< celticminstrel> Well, for some reason, ai.side is nil.
20160228 22:21:50< celticminstrel> I can't figure out why.
20160228 22:21:56< celticminstrel> (This is in an external Lua CA.)
20160228 22:22:28< celticminstrel> In fact I suspect ai is an empty table.
20160228 22:23:24< vultraz> Aginor: unfortunately, the Credits screen seems to have been broken... not sure when, but it wasn't broken recently, I don't think.. (segfault in gui::dialog_frame::handle_event(SDL_Event const&) when you click the button) :(
20160228 22:23:30< mattsc> You can use my testing setup, if you want.  It’s in the AI-demos add-ons.
20160228 22:23:40< mattsc> not plural, just one of them  :P
20160228 22:24:19< celticminstrel> Okay, hmm.
20160228 22:24:40< celticminstrel> So it turns out that the ai table is the table being returned by my external CA. Maybe I'm not supposed to do that.
20160228 22:25:05< mattsc> Umm, no, the CA should not return that
20160228 22:25:12< mattsc> It’s one of the inputs to it, in fact.
20160228 22:25:23< celticminstrel> Your example returns the same thing though...
20160228 22:25:26< mattsc> Sounds like reading the wiki page I just finished should help.
20160228 22:25:35< mattsc> hmmm?
20160228 22:25:42< celticminstrel> mattsc: I mean that the ai argument in my evaluation function is the table containing the evaluation function.
20160228 22:26:12-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160228 22:26:27< ancestral> vultraz: I’m still getting a crash
20160228 22:26:30< mattsc> uh, no, that’s not like it is supposed to be, I think
20160228 22:26:36< celticminstrel> At the end of the CA file, you return a table containing evaluation and execution. For some reason, it seems like that table is then being passed to my evaluation function.
20160228 22:26:38< vultraz> ancestral: whattttt???? D:
20160228 22:26:46< ancestral> Hang on
20160228 22:27:05< mattsc> yes, but in the ‘self’ argument, not the ‘ai’ argument
20160228 22:27:13< celticminstrel> By the way, is it possible to not hide chat while a menu is open?
20160228 22:27:35< celticminstrel> mattsc: Huh? Is that supposed to be the case? But no, it's being passed as the first argument in my case.
20160228 22:27:36< mattsc> ‘self’ is the Lua engine, which contains the CA functions; ‘ai’ is the table containing all the Lua AI functions
20160228 22:28:06< ancestral> vultraz: Yep. AI-only still crashes, but I don’t get an OS crash report
20160228 22:28:19< ancestral> Let me try through terminal
20160228 22:28:21< celticminstrel> I'm probably either doing something wrong somewhere or broke something elsewhere.
20160228 22:28:52< vultraz> ancestral: but ai only doesn't crash for me :( ... thought there are some images now not found... wtf
20160228 22:28:53< celticminstrel> Hmm, I have on idea.
20160228 22:28:56< celticminstrel> ^one
20160228 22:29:14< ancestral> ancestral: Hang on
20160228 22:29:18< ancestral> I might have figured it out
20160228 22:29:30< vultraz> ffs gui1
20160228 22:30:15< mattsc> celticminstrel: let me do a quick test in-secnario, rather than with the testing setup
20160228 22:30:29< vultraz> god dammit
20160228 22:30:33< ancestral> Well, maybe not
20160228 22:30:44< ancestral> vultraz: Is there logging I can do?
20160228 22:30:50< vultraz> uhhh
20160228 22:30:52< vultraz> not sure
20160228 22:30:56< vultraz> ancestral: what happens, exactly?
20160228 22:31:00< ancestral> I load the game
20160228 22:31:03< ancestral> I see the screen briefly
20160228 22:31:07-!- fendrin [~quassel@wesnoth/developer/fendrin] has joined #wesnoth-dev
20160228 22:31:07< ancestral> Then crashes to desktop
20160228 22:31:11< celticminstrel> vultraz / Aginor: Is it possible to show chat text behind the context menu?
20160228 22:31:11< vultraz> ...weird
20160228 22:31:13< ancestral> I’ll link you the file
20160228 22:31:16< ancestral> You give it a try
20160228 22:31:16< vultraz> that does not happen to me
20160228 22:31:18< vultraz> ok
20160228 22:31:19< ancestral> Morituri 3p
20160228 22:31:39< ancestral> (I created it new and still has issues)
20160228 22:31:52< vultraz> Aginor: also, this: https://www.dropbox.com/s/91ys6fbdc6dnris/replayBLAGH.PNG?dl=0
20160228 22:31:52< celticminstrel> I use the context menu sort of as a "pause" feature to prevent scrolling when I switch to a different app, so it can be a little annoying that the chat text disappears.
20160228 22:32:40< vultraz> Aginor: I suspect the ui is being drawn before the theme wml is loaded, since the 'button names' are missing from the log error
20160228 22:33:09< ancestral> vultraz: http://mproud.com/wesnoth/3p_Morituri_replay-3.gz
20160228 22:33:40< mattsc> celticminstrel: http://pastebin.com/pPQ8jMqz
20160228 22:33:48< celticminstrel> mattsc: I doubt it's relevant, but I have exec_parms and eval_parms both set to "".
20160228 22:34:00< mattsc> celticminstrel: this is from inside the Guardian Micro AI test scenario
20160228 22:34:11< vultraz> ancestral: does not crash for me
20160228 22:34:17< vultraz> ancestral: are you sure you have latest master
20160228 22:34:26< ancestral> I git pulled
20160228 22:34:26< celticminstrel> I also added wesnoth.message calls inside the dummy Lua engine itself which confirms that everything is properly set at that point.
20160228 22:34:28< ancestral> I cleaned
20160228 22:34:32< ancestral> I mean, I can do it again
20160228 22:34:43< celticminstrel> So I don't understand why it's not working.
20160228 22:34:44-!- fabi [~quassel@wesnoth/developer/fendrin] has quit [Ping timeout: 260 seconds]
20160228 22:35:15< celticminstrel> Unless there's a problem with the generated candidate action code.
20160228 22:35:33< mattsc> celticminstrel: is what I just sent you consistent with what you are seeing?
20160228 22:36:20< ancestral> What is the dialog.cpp fix?
20160228 22:36:25< mattsc> All I did was insert the debug out code into one of the Micro AIs and run the test scenario.  No change to the code otherwise at all (well, I had to add the ‘self’ field as that was not being used)
20160228 22:36:26< ancestral> I just pulled that doen
20160228 22:36:28< ancestral> *down
20160228 22:36:30< celticminstrel> mattsc: No; I'm not using that method, but based on pairs() iteration I find that ai = {execution = function: 0x..., evaluation = function: 0x...}
20160228 22:36:41< celticminstrel> And self is empty - pairs() iteration prints nothing.
20160228 22:37:05< vultraz> ancestral: the Profile button in unit_preview_pane wasn't visible
20160228 22:37:35< mattsc> celticminstrel: then something is different about your setup than it is in mine.  I can’t really say anything else without more information.  :(
20160228 22:37:59< celticminstrel> Yeah, maybe I broke something somewhere.
20160228 22:38:05-!- Appleman1234 [~Appleman1@KD106161153065.au-net.ne.jp] has quit [Ping timeout: 268 seconds]
20160228 22:38:53< celticminstrel> Trying to figure it out.
20160228 22:39:46< celticminstrel> The weird thing is that it's only messed up when passed to my functions - the actual direct CA code has the correct values. (I think)
20160228 22:40:00< celticminstrel> By which I mean, the generated code that loads and calls the external CA.
20160228 22:40:33< mattsc> celticminstrel: btw, self.get_ai() returns the ai table
20160228 22:40:41< mattsc> I just tested that.
20160228 22:41:22< celticminstrel> I know. The aforementioned generated code calls it to get the ai table for the first argument.
20160228 22:41:39< mattsc> So, if you use external CAs, ‘self’ only contains ‘data’ and that function.
20160228 22:41:47< celticminstrel> That's why I can't understand what's going on here.
20160228 22:41:49< mattsc> Whereas if you use the old message, it also contains the CA code.
20160228 22:42:09< celticminstrel> The generated code is return wesnoth.require(location):evaluation ... 
20160228 22:42:10< celticminstrel> ...
20160228 22:42:11< mattsc> s/message/method
20160228 22:42:13< mattsc> :P
20160228 22:42:16< celticminstrel> Wait.
20160228 22:42:19< celticminstrel> That colon.
20160228 22:42:35< celticminstrel> I suddenly understand what's going on here.
20160228 22:42:44< mattsc> You need a colon, not a period?
20160228 22:42:54< celticminstrel> I defined my CA a little differently from your example, you see.
20160228 22:43:25< celticminstrel> I wrote "function evaluation(ai, cfg, self)" and at the bottom "return {evaluation=evaluation,execution=execution}".
20160228 22:43:39< ancestral> Still crashes, shucks
20160228 22:43:47< celticminstrel> But that colon in the generated code means that the function is actually being passed four arguments, not three.
20160228 22:44:01< celticminstrel> So, with how I have it, my ai table is in fact called cfg.
20160228 22:44:14< celticminstrel> And the self table is an unnamed argument.
20160228 22:44:16< ancestral> mattsc: You don’t buy chance have a binary built of master?
20160228 22:44:45< mattsc> ancestral: I don’t think I’ve built master this year yet  :(
20160228 22:44:55 * ancestral nods
20160228 22:45:26< celticminstrel> Okay yeah, it looks like I have it working now.
20160228 22:45:34< mattsc> Sorry; been gone for most of Dec and Jan, and since I’ve gotten back, I’ve only done this wiki nonsense
20160228 22:45:56< celticminstrel> "this wiki nonsense" is extremely important work, so don't worry about it.
20160228 22:46:35< mattsc> celticminstrel: well, good;  that’s what you get for trying to be smart.  ;)
20160228 22:47:18< celticminstrel> I was working from memory of your example, but missed that detail.
20160228 22:47:36< celticminstrel> I do not like the external CA setup though.
20160228 22:47:56< mattsc> Well, you’re a lot smarter than I then, if you can get that close from memory alone
20160228 22:48:13< mattsc> you’ve said that before
20160228 22:50:18< mattsc> If you get rid of the external CAs, the Micro AIs will stop working as they do  (and I don’t only mean that they need to be rewritten; it won’t be possible to combine them at will if we need an engine definition that is required to be set at the beginning of the scenario)
20160228 22:51:15< celticminstrel> My problems are more with the implementation and fiddly details than the overall concept of them.
20160228 22:52:03< celticminstrel> For example, there's no reason to use the colon rather than the dot for the evaluation and execution functions, and in fact doing so conflicts with the recommendation of calling the third parameter "self".
20160228 22:52:46< celticminstrel> (Using the colon generally means that there is an invisible first parameter which is called "self".)
20160228 22:53:46< celticminstrel> So, instead of "local ca = {}; function ca:evaluation(...)" I'd prefer "local ca = {}; function ca.evaluation(...)".
20160228 22:54:11< mattsc> okay, no issues with that
20160228 22:54:14< celticminstrel> Or, alternatively - keep the colon, but use that invisible "self" argument to store the data.
20160228 22:54:48< celticminstrel> ...actually, that's a separate issue, I think. If we did that, you'd probably still define them with a dot.
20160228 22:54:49-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20160228 22:55:36< celticminstrel> Or maybe not. Perhaps I should actually give this a try, just to see how it works.
20160228 22:59:43-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160228 23:08:04< celticminstrel> So, [state] is valid in [stage], [goal], and [aspect], but not [candidate_action] for some reason. (Lua ones only of course.)
20160228 23:09:01< celticminstrel> Also weird is that candidate actions are expected to return 5 values, though I think 4 of those are just ignored.
20160228 23:10:47< celticminstrel> It seems like it probably serves the same purpose as [engine][data], but local to the specific component...
20160228 23:11:48< celticminstrel> Oh, wait, it is supported by candidate actions, but only for the execution, not for the evaluation.
20160228 23:13:46< celticminstrel> Oh, I was wrong, [aspect] doesn't support it.
20160228 23:13:56< celticminstrel> It passes a dummy config instead.
20160228 23:14:23< celticminstrel> (The thing about 5 values only applies to evaluation.)
20160228 23:15:01< celticminstrel> And [goal] also passes a dummy config. So I guess it's only [stage] and [candidate_action].
20160228 23:15:23< celticminstrel> I wonder if it would be useful to enable it for [aspect] and [goal], too.
20160228 23:23:11-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160228 23:24:04< vultraz> ancestral: it's weird that it crashes for you and not me
20160228 23:25:44< celticminstrel> ...wait. What does enabled=yes|no mean in a [candidate_action]?
20160228 23:25:45< ancestral> I want to log, but I don’t know what to log for
20160228 23:27:07< vultraz> can't you get a stacktrace?
20160228 23:27:11< vultraz> or just try log=all
20160228 23:27:16< ancestral> I’ll try that
20160228 23:27:27< vultraz> it will flood your console but you might get something
20160228 23:27:35< ancestral> —log=all
20160228 23:27:43< ancestral> What’s the syntax?
20160228 23:28:07< ancestral> —log-error all?
20160228 23:29:16< vultraz> --log-debug=all
20160228 23:29:30< vultraz> and I don'tkid when I say it will *flood* your console
20160228 23:30:08< vultraz> .... in fact, it makes wesnoth impossible to even open
20160228 23:30:15< vultraz> so forget that
20160228 23:30:24< ancestral> Ran fine for me
20160228 23:30:50< vultraz> oh wait
20160228 23:30:57< vultraz> log-error=all does allow wesnoth to start
20160228 23:30:58< vultraz> ok
20160228 23:31:03< vultraz> do you get anything?
20160228 23:31:28< ancestral> lots of things
20160228 23:31:42< ancestral> 20160228 17:30:03 info display: calling SDL_Quit()
20160228 23:31:43< ancestral> 20160228 17:30:03 info display: called SDL_Quit()
20160228 23:32:57< ancestral> Earlier I got
20160228 23:33:22< ancestral> 20160228 16:29:00 error display: could not open image 'buttons/.png'
20160228 23:33:24< vultraz> wait...
20160228 23:33:25< vultraz> huh
20160228 23:33:33< ancestral> I know I was messing with themes
20160228 23:33:34< vultraz> if I run with --log-error=all it does crash for me ;|
20160228 23:33:36< vultraz> :| *
20160228 23:33:39< ancestral> Aha
20160228 23:33:49< ancestral> (I know I switched to Default though)
20160228 23:33:53-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has joined #wesnoth-dev
20160228 23:34:03< vultraz> i get the theme thing too
20160228 23:34:26-!- Appleman1234 [~Appleman1@KD106161150109.au-net.ne.jp] has joined #wesnoth-dev
20160228 23:35:25< vultraz> I got the callstack
20160228 23:35:37< vultraz> ahhh
20160228 23:35:40< vultraz> yes, I see
20160228 23:39:53< celticminstrel> mattsc: Do you have lots of working examples of aspects implemented in Lua? Particularly the more complex aspects.
20160228 23:40:14< vultraz> ancestral, Aginor: this crash is related to the missing images I posted earlier
20160228 23:42:00< ancestral> 20160228 17:41:49 error display: could not open image 'buttons/.png'
20160228 23:42:01< ancestral> 20160228 17:41:49 error display: could not open image 'buttons/-pressed.png'
20160228 23:42:02< ancestral> 20160228 17:41:49 error display: could not open image 'buttons/-active.png'
20160228 23:42:03< ancestral> 20160228 17:41:49 error display: error initializing button images! file name: buttons/.png
20160228 23:42:07< vultraz> yes
20160228 23:42:13< vultraz> somehow that's causing the crash
20160228 23:43:26< vultraz> layout_buttons() calling set_image
20160228 23:44:14-!- mjs-de [~mjs-de@x4db5c93f.dyn.telefonica.de] has quit [Remote host closed the connection]
20160228 23:46:42-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 260 seconds]
20160228 23:47:18< Aginor>  hmmm
20160228 23:47:46< Aginor> which is needed to be called to resize the buttons without having to recreate them
20160228 23:47:52< Aginor> although fixable with a simple check
20160228 23:48:03< Aginor> if (!image.empty())
20160228 23:48:06< Aginor> or somesuch
20160228 23:52:06< vultraz> yes, that fixes it
20160228 23:53:39< irker494> wesnoth: Charles Dang wesnoth:master 0babe82b5b0f / src/display.cpp: Don't load button images if none were provided https://github.com/wesnoth/wesnoth/commit/0babe82b5b0f32dfbde904bc42fc0f1e3d83a909
20160228 23:53:42< vultraz> ancestral: ^ see if that fixes your crash
20160228 23:55:23< vultraz> though perhaps I should move that check to set_image itself
20160228 23:55:29< vultraz> but first see if that fixes your problem
20160228 23:55:50< ancestral> building
20160228 23:57:26< mattsc> celticminstrel: no, since I mostly work with custom AIs (not with customized version of the default AI), I barely use aspects at all
20160228 23:57:34< celticminstrel> Ah.
20160228 23:57:34< mattsc> The only Lua aspect examples I have are those on the wiki
--- Log closed Mon Feb 29 00:00:12 2016