--- Log opened Sat Apr 08 00:00:08 2017
20170408 00:07:47-!- vultraz_iOS [uid24821@wesnoth/developer/vultraz] has joined #wesnoth-dev
20170408 00:12:56< vultraz_iOS> celticminstrel: could this become an enum? https://github.com/wesnoth/wesnoth/blob/master/src/gui/widgets/grid.hpp#L39-L60
20170408 00:13:55-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Read error: Connection reset by peer]
20170408 00:15:09-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 00:19:11< celticminstrel> It could, but I wouldn't recommend it.
20170408 00:19:18< celticminstrel> Those are bitmasks for a bitfield.
20170408 00:24:49-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20170408 00:24:52-!- irker070 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20170408 00:24:52< irker070> wesnoth: Charles Dang wesnoth:master 2053362c97e4 / src/gui/widgets/ (grid.cpp grid.hpp): GUI2/Grid: added child getter by contained widget https://github.com/wesnoth/wesnoth/commit/2053362c97e4ce4424cf2503c636dbd4a66c824a
20170408 00:25:00-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20170408 00:25:02-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20170408 00:25:06< vultraz_iOS> formatting only it is then
20170408 00:58:46-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20170408 01:03:05-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20170408 01:07:26-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 01:09:28-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 240 seconds]
20170408 01:51:04-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20170408 02:40:47-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20170408 02:40:53-!- janebot [~Gambot@grickit.us] has joined #wesnoth-dev
20170408 02:40:57-!- janebot [~Gambot@grickit.us] has quit [Changing host]
20170408 02:40:57-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20170408 02:53:45< vultraz_iOS> so, it seems my set_origin usage for the storyscreen isn't optimal
20170408 02:53:50< vultraz_iOS> I'm working on a new method
20170408 02:54:34< vultraz_iOS> two reasons:
20170408 02:54:48< vultraz_iOS> 1. if the *initial* panel position is not "bottom", the panel won't be in the right place
20170408 02:55:02< vultraz_iOS> 2. if you resize when the panel is not on the bottom, it will revert to the bottom
20170408 02:55:31< vultraz_iOS> 3. this was also a problem with the old screen, but you cannot have a title AND text at the top - the box would cover it
20170408 02:56:10< vultraz_iOS> my new method solves all problems
20170408 03:04:39< celticminstrel> "all" :P
20170408 03:10:03< pydsigner> celticminstrel: Even cancer!
20170408 03:11:22< vultraz_iOS> even Trump!
20170408 03:15:27 * celticminstrel was thinking Trump, yeah. :P
20170408 03:15:37< celticminstrel> Cancer's a good one too.
20170408 03:30:38< vultraz_iOS> im glad i went back to rethink my API
20170408 03:30:43< vultraz_iOS> this is what i had last night : https://github.com/wesnoth/wesnoth/compare/master...Vultraz:master
20170408 03:30:56< vultraz_iOS> (mostly from moving the grid child stuff into its own file)
20170408 03:40:06< vultraz_iOS> and ^ here is the new code
20170408 03:40:06< irker070> wesnoth: Charles Dang wesnoth:master d7b564e2a71b / src/gui/widgets/ (grid.cpp grid.hpp widget.cpp widget.hpp): GUI2/Widget: added ability to set widget alignment on the fly https://github.com/wesnoth/wesnoth/commit/d7b564e2a71bf473a666f184302273ec23be2856
20170408 03:40:09< irker070> wesnoth: Charles Dang wesnoth:master 5fbd4a727bce / / (3 files in 3 dirs): Story Viewer: deployed new alignment set API https://github.com/wesnoth/wesnoth/commit/5fbd4a727bce7cabbe80e9da17cc44de7a67ddad
20170408 03:40:13< vultraz_iOS> much cleaner
20170408 03:45:42< vultraz_iOS> celticminstrel: is the WFL NOT op ~, !=, or !?
20170408 03:50:54< celticminstrel> IIRC it's "not"
20170408 03:51:07< celticminstrel> Assuming you mean boolean not.
20170408 03:52:45< vultraz_iOS> so
20170408 03:52:51< vultraz_iOS> foo not = bar?
20170408 03:53:06< celticminstrel> That's why I said "assuming you mean boolean not".
20170408 03:53:12< celticminstrel> Obviously you don't mean boolean not.
20170408 03:53:17< celticminstrel> != is the not equal operator.
20170408 03:53:24< vultraz_iOS> ok
20170408 03:54:22< vultraz_iOS> (we really do need to add handling of a bool type)
20170408 03:54:38< celticminstrel> In future, instead of asking me, you can check http://wiki.wesnoth.org/Wesnoth_Formula_Language :P
20170408 03:54:51< celticminstrel> We don't need to add handling of a bool type, but I've considered doing so.
20170408 03:55:10< celticminstrel> It means modifying the tokenizer to accept "true" and "false" as keywords.
20170408 03:55:26< celticminstrel> Which shouldn't be too hard.
20170408 03:55:43< celticminstrel> And then they need to be parsed as expression keywords, similar to how "functions" is currently handled.
20170408 03:57:48< celticminstrel> Uh, something about your bit arithmetic looks like it might be off...
20170408 03:58:27< vultraz_iOS> hm?
20170408 03:59:56< celticminstrel> Ah, wait, maybe it's fine after all?
20170408 04:00:38< celticminstrel> Clears the bits set in mode_mask, then sets the bits set in set_flag - that's the intention?
20170408 04:01:06< vultraz_iOS> basically, yes
20170408 04:01:42< celticminstrel> I wonder how many places interpolate macros into the middle of formulas like that.
20170408 04:02:11< celticminstrel> You randomly changed the mode mame from middle to center for absolutely no reason. Fun.
20170408 04:02:54< vultraz_iOS> because center is the name of the vertical_alignment key
20170408 04:03:01< vultraz_iOS> it allows me to set both things with a single string
20170408 04:03:05< vultraz_iOS> key value*
20170408 04:03:37< celticminstrel> Oh. I see.
20170408 04:07:56< vultraz_iOS> celticminstrel: https://github.com/wesnoth/wesnoth/commit/03c7b3debd2c85fe14629ecde20d726de68540ce why we need a bool type (dunno why it's not getting reported)
20170408 04:09:20< celticminstrel> text_box_default.cfg:172-3
20170408 04:09:47< celticminstrel> I sorta feel like that colour should check text = '', but I could just be misunderstanding what that is.
20170408 04:09:47< irker070> wesnoth: Charles Dang wesnoth:master 03c7b3debd2c / data/gui/widget/panel_story_viewer.cfg src/gui/dialogs/story_viewer.cpp: Story Viewer: hide title if there's no title to show, and display text panel bor https://github.com/wesnoth/wesnoth/commit/03c7b3debd2c85fe14629ecde20d726de68540ce
20170408 04:10:33< celticminstrel> vultraz_iOS: Why what's not getting reported?
20170408 04:10:48< celticminstrel> vultraz_iOS: Also can you explain that GUI_FILTER macro?
20170408 04:11:25< celticminstrel> vultraz_iOS: Also, in that commit you just pushed, the "proper" way is to set title_present to an int, not a string.
20170408 04:12:48< vultraz_iOS> hm?
20170408 04:12:49< vultraz_iOS> explain
20170408 04:12:56< vultraz_iOS> 0/1?
20170408 04:13:00< celticminstrel> Yes.
20170408 04:13:09< vultraz_iOS> and check against 0/1?
20170408 04:13:11< celticminstrel> Please answer my two questions too?
20170408 04:13:12< celticminstrel> Yes.
20170408 04:13:19< celticminstrel> Or just don't compare at all.
20170408 04:13:31< celticminstrel> "panel_position != 'top' or title_present"
20170408 04:13:35< vultraz_iOS> ahhh
20170408 04:13:36< vultraz_iOS> ok
20170408 04:14:06< vultraz_iOS> what GUI_FILTER macro?
20170408 04:14:16< celticminstrel> text_box_default.cfg:172-3
20170408 04:15:01< vultraz_iOS> uh...
20170408 04:15:20< vultraz_iOS> do you mean _GUI_FILTER_BOX_STUFF?
20170408 04:15:32< celticminstrel> That sounds right.
20170408 04:16:02< vultraz_iOS> for the magnifying class icon in filter text boxes
20170408 04:16:47< celticminstrel> There's text="Search" too though
20170408 04:16:55< celticminstrel> And disabled colour?
20170408 04:17:24< vultraz_iOS> yes
20170408 04:17:34< vultraz_iOS> you get gray "Search" text when it's not selected
20170408 04:17:58< celticminstrel> Oh wait, is that definition only used when the textbox is not focused?
20170408 04:18:10< vultraz_iOS> yes
20170408 04:19:29< irker070> wesnoth: Charles Dang wesnoth:master 3a3e6fe72531 / data/gui/widget/panel_story_viewer.cfg src/gui/dialogs/story_viewer.cpp: Fixup 03c7b3d https://github.com/wesnoth/wesnoth/commit/3a3e6fe72531db8a124c76cdfc636fceda66d87a
20170408 04:19:30< celticminstrel> Hmm, but also, if it's unfocused and there is text in it, that text is not drawn.
20170408 04:19:38< vultraz_iOS> yes
20170408 04:19:39-!- JyrkiVesterinen [~JyrkiVest@87-100-188-214.bb.dnainternet.fi] has joined #wesnoth-dev
20170408 04:19:50< celticminstrel> How is that a good idea?
20170408 04:20:03< celticminstrel> The text would return when you focus it?
20170408 04:20:08< vultraz_iOS> because it's "background" text
20170408 04:20:27< vultraz_iOS> it only shows if there's no text and it's not focused
20170408 04:20:34< celticminstrel> I'm talking about the text you type into it.
20170408 04:20:40< vultraz_iOS> celticminstrel: should we add a variant ctor that takes bool and casts to int?
20170408 04:20:48< vultraz_iOS> celticminstrel: what about it
20170408 04:21:02< celticminstrel> I haven't actually tried it, but it looks like it'll vanish when the text box loses focus? Am I just imagining this?
20170408 04:21:27< celticminstrel> I'd say sure to a variant ctor taking bool, but it might cause problems.
20170408 04:21:40< celticminstrel> Try it and see, I guess?
20170408 04:21:54< vultraz_iOS> you're imagining it
20170408 04:21:58< vultraz_iOS> go try any filter box
20170408 04:22:00< vultraz_iOS> like in game load
20170408 04:26:16< celticminstrel> Next question: toggle_panel_default.cfg:126
20170408 04:26:20< celticminstrel> What's the intent?
20170408 04:26:53< celticminstrel> And also, where is that definition used?
20170408 04:28:00< vultraz_iOS> commented out experiment for procedural gradient
20170408 04:28:12< celticminstrel> What kind of gradient?
20170408 04:28:13< vultraz_iOS> fancy is used in campaigns, prefs, and the addon manager
20170408 04:28:17< vultraz_iOS> alpha fadeout
20170408 04:29:17< celticminstrel> So it's the one with the yellow line on the left?
20170408 04:29:34< vultraz_iOS> Yes
20170408 04:29:53< vultraz_iOS> What are you doing that requires looking through all these definitions?
20170408 04:30:19< celticminstrel> Just minor cleanup.
20170408 04:30:24< vultraz_iOS> Good, good
20170408 04:30:36< vultraz_iOS> Book variant ctor doesn't work
20170408 04:30:40< vultraz_iOS> Bool
20170408 04:30:46< celticminstrel> Define "doesn't work".
20170408 04:30:47< vultraz_iOS> Ambitious with int ctor
20170408 04:30:52< vultraz_iOS> Ambiguous
20170408 04:30:59< celticminstrel> Yeah, that's more or less what I suspected could happen.
20170408 04:31:28< celticminstrel> The fix would probably be adding a constructor for every int type.
20170408 04:31:41< celticminstrel> But I guess as long as we don't have a bool type, there's no point.
20170408 04:31:51< celticminstrel> Because constructing from a bool will call the int constructor ATM.
20170408 04:32:20< vultraz_iOS> Oh?
20170408 04:32:28< vultraz_iOS> You mean I don't have to static cast manually?
20170408 04:33:00< celticminstrel> You don't have to, no.
20170408 04:33:15< celticminstrel> Converting between int and bool could give warnings, but I think that's normally for the other direction.
20170408 04:51:56< celticminstrel> vultraz_iOS: What do I need to do to preview the new storyscreen?
20170408 04:58:43< vultraz_iOS> celticminstrel: comment out src/storyscreen/interface.cpp:42 and 60
20170408 04:58:43< irker070> wesnoth: Celtic Minstrel wesnoth:master 348d2e1bb2d1 / data/gui/widget/ (9 files): Some cleanup of GUI2 formulas https://github.com/wesnoth/wesnoth/commit/348d2e1bb2d1d5e03b908c394403f96636dd071a
20170408 04:59:53< vultraz_iOS> I approve of these cleanups
20170408 05:12:47< celticminstrel> 255 * (1.0 - as_decimal(x) / width)
20170408 05:13:06< celticminstrel> That gives you a simple horizontal alpha fadeout.
20170408 05:13:43< celticminstrel> If applied to solid color, it gives much the same result as the image you're currently using there.
20170408 05:13:47-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20170408 05:13:57< celticminstrel> But the image you were applying it to was not solid colur in the slightest.
20170408 05:14:47-!- Appleman1234 [~Appleman1@pl32445.ag1212.nttpc.ne.jp] has quit [Ping timeout: 268 seconds]
20170408 05:17:25< celticminstrel> JyrkiVesterinen: Do you have any idea why line 100 of variant_value.cpp might be causing "incompatible iterator assertions" in my MSVC 2013 build?
20170408 05:17:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20170408 05:21:58< celticminstrel> vultraz_iOS: Storyscreen title seems to be missing its background? Apart from that it looks good, though the text area seems to be a little smaller.
20170408 05:22:11< vultraz_iOS> it is
20170408 05:22:18< vultraz_iOS> and the title no longer has a background
20170408 05:22:25< celticminstrel> Yes, that's the problem.
20170408 05:22:28< vultraz_iOS> it is not
20170408 05:22:30< celticminstrel> It is.
20170408 05:22:35< celticminstrel> It is a gigantic problem.
20170408 05:22:38< vultraz_iOS> I added a subtle gradient effect to the top
20170408 05:22:52< vultraz_iOS> both to frame the background and provide a dark backdrop for the title
20170408 05:22:56< JyrkiVesterinen> celticminstrel: No idea. Quick source code analysis suggests that it should be impossible.
20170408 05:23:07< celticminstrel> Yeah, that's what I thought too!
20170408 05:24:20< vultraz_iOS> celticminstrel: why is it such a problem otherwise?
20170408 05:26:02< celticminstrel> I didn't notice the gradient effect until you mentioned it, and now that I do notice, I don't think I like it.
20170408 05:26:12< celticminstrel> Anyway, it doesn't solve the problem that the background did.
20170408 05:26:18< celticminstrel> It's a problem because of contrast.
20170408 05:26:34< celticminstrel> There's no way to reliable ensure that the text has decent contrast against the background.
20170408 05:26:39< celticminstrel> ^reliably
20170408 05:26:50< celticminstrel> By which I mean, by choose an appropriate text colour.
20170408 05:27:20< celticminstrel> Thus, you provide a dark background which then dominates and provides the needed contrast, even if the actual background shows through.
20170408 05:27:40< celticminstrel> Contrast against the typical bigmaps probably won't be a major problem, but you can't assume that the title will only appear over a conventional bigmap.
20170408 05:27:44< vultraz_iOS> that's the point of the gradient :|
20170408 05:27:59< celticminstrel> The gradient doesn't solve it. The text is well below the gradient's darkness.
20170408 05:28:20< vultraz_iOS> ok, simple fact
20170408 05:28:20< celticminstrel> That is, the darkness of the gradient has pretty much ended by the point where the text is drawn.
20170408 05:28:25< vultraz_iOS> I can't draw the background :|
20170408 05:28:36< celticminstrel> Plus IMO the gradient looks ugly.
20170408 05:28:43< vultraz_iOS> :/
20170408 05:28:50< vultraz_iOS> but it provides nice framing
20170408 05:31:36-!- Kwandulin [~Kwandulin@p200300760F3E7DF149CEF7081D588286.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170408 05:32:42< vultraz_iOS> and i can't figure out how to draw the background
20170408 05:34:08< vultraz_iOS> maybe new label definition...
20170408 05:34:14< vultraz_iOS> ill tackle that later
20170408 05:34:18< vultraz_iOS> right now i need to do...
20170408 05:34:28< vultraz_iOS> tiling
20170408 05:58:35< irker070> wesnoth: Celtic Minstrel wesnoth:story_title_padding_poc e02bcc735934 / data/gui/ (widget/label_default.cfg window/story_viewer.cfg): Story viewer: Add translucent background behind title https://github.com/wesnoth/wesnoth/commit/e02bcc73593444844b2bd1698705f6c2d9e8a082
20170408 05:58:57< celticminstrel> I disagree on the nice framing.
20170408 05:59:17< celticminstrel> Anyway, there's a partial implementation of a background. ^
20170408 05:59:28< celticminstrel> Still missing something though.
20170408 05:59:59< celticminstrel> The canvas needs to be about 20 pixels wider and higher for this to work as it should.
20170408 06:00:07< celticminstrel> But I have no idea if there's a way to make that happen.
20170408 06:04:59-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20170408 06:11:03-!- Appleman1234 [~Appleman1@pl26331.ag1212.nttpc.ne.jp] has joined #wesnoth-dev
20170408 06:13:57< vultraz_iOS> celticminstrel: ill take a look
20170408 06:20:54< vultraz_iOS> celticminstrel: I literally see no difference :/
20170408 06:21:03< celticminstrel> Did you miss the fact that it's on a branch?
20170408 06:21:23< celticminstrel> It is a little hard to see though since there's no padding...
20170408 06:21:43< vultraz_iOS> oh
20170408 06:21:44< vultraz_iOS> dur
20170408 06:21:49< vultraz_iOS> not i have to checkout a branch
20170408 06:22:43< vultraz_iOS> um
20170408 06:22:45< vultraz_iOS> rather... 
20170408 06:22:46< vultraz_iOS> uh...
20170408 06:22:50< vultraz_iOS> alright :/
20170408 06:28:00< vultraz_iOS> I do not like
20170408 06:28:46< celticminstrel> Nor do I though, because there's no padding.
20170408 06:28:53< celticminstrel> That's why I pushed to a branch instead of master.
20170408 06:29:17< celticminstrel> The background needs to be larger than the text in order to look good.
20170408 06:37:10< vultraz_iOS> but a block is still bad
20170408 06:38:17-!- atarocch [~atarocch@93.56.160.28] has joined #wesnoth-dev
20170408 06:41:26< celticminstrel> No.
20170408 06:41:26< vultraz_iOS> at least if we could do sideways fadeout
20170408 06:41:45< celticminstrel> Well, it'd be fine if it faded out at the edges, sure.
20170408 06:41:52< celticminstrel> As long as the fadeout is entirely in the padding area.
20170408 06:42:00< celticminstrel> ie, the currently-nonexistent area. >_>
20170408 06:42:21< celticminstrel> I don't think a block is bad, but fading out the edges is acceptable too.
20170408 06:42:54< vultraz_iOS> sadly, we can't do fading proceduraly without a surface
20170408 06:42:56< vultraz_iOS> we should change that
20170408 06:44:03< celticminstrel> Well, I think the first step would be to actually add the padding.
20170408 06:44:23< celticminstrel> Padding could be something generic to styled_widget, actually.
20170408 06:55:14< vultraz_iOS> celticminstrel: should I remove the gradient?
20170408 06:57:22< celticminstrel> ???
20170408 06:57:29< vultraz_iOS> the one you didn't like
20170408 06:57:31< vultraz_iOS> behind the title area
20170408 06:57:35< celticminstrel> Oh, that.
20170408 06:57:52< celticminstrel> You could, or you could get some further opinions on it first.
20170408 06:58:13< vultraz_iOS> I think I'll add if there's no title
20170408 06:58:34< vultraz_iOS> it looks a *lot* better with it than w/o it, trust me
20170408 07:00:03< celticminstrel> If you say so.
20170408 07:00:23< celticminstrel> Though if the background works out, the gradient may become obsolete?
20170408 07:02:07-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20170408 07:04:27< irker070> wesnoth: Celtic Minstrel wesnoth:master 88420357b07d / data/gui/ (widget/label_story_screen_title.cfg window/story_viewer.cfg): Story Viewer: add translucent background behind title https://github.com/wesnoth/wesnoth/commit/88420357b07d79764d2f9066624ec6eb19173b2a
20170408 07:04:29< irker070> wesnoth: Charles Dang wesnoth:master 512723055953 / src/gui/dialogs/story_viewer.cpp: Story Viewer: only show upper gradient if no title is present https://github.com/wesnoth/wesnoth/commit/512723055953580bc67c2772e928e475cd3cba2f
20170408 07:04:30< vultraz_iOS> celticminstrel: ^
20170408 07:07:01-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 268 seconds]
20170408 07:07:40< celticminstrel> Is the only change to my commit the removal of the borders?
20170408 07:07:49< vultraz_iOS> no
20170408 07:07:54< vultraz_iOS> default height was set
20170408 07:07:59< celticminstrel> Oh, I see a 20 and a vertically centred too.
20170408 07:08:03< vultraz_iOS> and that
20170408 07:08:12< celticminstrel> Default height?
20170408 07:08:16< celticminstrel> What do you mean?
20170408 07:08:26< vultraz_iOS> default height for the widget
20170408 07:08:29< vultraz_iOS> set to 50
20170408 07:09:45< vultraz_iOS> else it would just shrink to text size
20170408 07:09:57< irker070> wesnoth: Charles Dang wesnoth:master 1ba98d6e42b5 / src/gui/dialogs/story_viewer.cpp: Story Viewer: tweak upper gradient opacity https://github.com/wesnoth/wesnoth/commit/1ba98d6e42b5ee324660bb2678b2f89b982fc0ab
20170408 07:10:08< celticminstrel> Oh, I see it.
20170408 07:10:33< celticminstrel> And of course you added all the useless spacing too.
20170408 07:10:42< vultraz_iOS> useless spacing?
20170408 07:10:44< vultraz_iOS> ah yes
20170408 07:10:45< vultraz_iOS> :D
20170408 07:10:53< celticminstrel> Like spacing between tags at different indentation levels.
20170408 07:11:04< vultraz_iOS> indeed
20170408 07:11:07< vultraz_iOS> that's the GUI2 format
20170408 07:11:13< celticminstrel> It's ridiculous.
20170408 07:11:17< celticminstrel> A waste of space.
20170408 07:11:22< vultraz_iOS> I like it
20170408 07:11:31< celticminstrel> Well good for you.
20170408 07:18:40-!- aidanhs [~aidanhs@2a00:d880:6:1ad::8e27] has quit [Ping timeout: 260 seconds]
20170408 07:25:44< vultraz_iOS> celticminstrel: any other thing I should address?
20170408 07:28:11< vultraz_iOS> (Thanks for doing the label bg )
20170408 07:37:35< irker070> wesnoth: Celtic Minstrel wesnoth:master e46b10af0eb9 / data/gui/widget/label_story_screen_title.cfg: Story viewer: Shrink translucent background behind title to a box https://github.com/wesnoth/wesnoth/commit/e46b10af0eb9d1ab3558179d78ca630267e1e7e3
20170408 07:37:38< celticminstrel> Not that I can think of?
20170408 07:37:58< celticminstrel> Should I try to find a way to fade out the edges now?
20170408 07:51:37< Aginor> wedge009: https://github.com/wesnoth/wesnoth/pull/961/files#diff-5eac1472101c1f3f68f6c9ce39068934L890 is also a bug, the original author is doing integer-style comparison on a floating point value
20170408 07:52:40< celticminstrel> But if it was explicitly assigned the value 1.0, the comparison should be successful.
20170408 07:52:46< Aginor> given the nature of a floating point number, that's fraught with problems, it's better to check whether the result is close enough to 1.0 that you don't care about any inaccuracies
20170408 07:52:53< Aginor> celticminstrel: that's true
20170408 07:53:27< celticminstrel> So it's probably "safe" in practice here, even though it's technically a bad idea.
20170408 07:53:43< celticminstrel> Given that the engine_val is AFAIK only assigned one of 1.0, 0.5, 1.5.
20170408 07:54:37< Aginor> I was admittedly assuming that it too would have arithmitics performed on it
20170408 07:54:48< Aginor> but if it's only ever assigned contstants...
20170408 07:55:01< celticminstrel> I'm not 100% sure.
20170408 08:03:20< vultraz_iOS> celticminstrel: sure, see if you can implement dynamic fade out in rectangles without a surface
20170408 08:05:58< vultraz_iOS> celticminstrel: I imagine it would have to do with allowing color as a formula (which I think we do), and allow sub-formulas for the component keys
20170408 08:06:43< celticminstrel> Pretty sure colors never support formulas in GUI2.
20170408 08:07:09< celticminstrel> I'd have to look at the code to see if supporting a formula is sufficient.
20170408 08:07:45< celticminstrel> But I suspect it would not be.
20170408 08:09:14< vultraz_iOS> Hmm
20170408 08:09:19< vultraz_iOS> Probably not actually..
20170408 08:09:27< vultraz_iOS> Not unless the color were evaluated per pixel
20170408 08:10:01< celticminstrel> Which is exactly the problem - it's probably not.
20170408 08:10:13< celticminstrel> Generally, the GUI2 formulas are only evaluated once each.
20170408 08:10:37< celticminstrel> One could probably do something using multiple layered rects, I suppose...
20170408 08:11:28< vultraz_iOS> This is where ogl/shaders comes in :/
20170408 08:11:52< celticminstrel> Speaking of which, if we were using OGL, we'd get tiling for free. :P
20170408 08:12:06< Aginor> or just a sensible rendering pipeline and alpha layering
20170408 08:12:23< celticminstrel> Hmm?
20170408 08:13:11< Aginor> it'd be easy enough to define a primitive to apply a alpha-overlay with linear interpolation on a rectangle
20170408 08:13:23< Aginor> ie, a fade in over time
20170408 08:13:49< Aginor> but the way rendering is done, it'd have to be hacked in to work neatly
20170408 08:14:17< vultraz_iOS> Indeed..
20170408 08:14:23< vultraz_iOS> We need better rendering :/
20170408 08:14:45< vultraz_iOS> Thankfully we're slowly consolidating the rendering areas
20170408 08:19:42< Aginor> it's still governed by the main game display, which is the hard bit to tackle
20170408 08:20:44-!- zookeeper [zookeeper@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20170408 08:21:51< vultraz_iOS> Aginor: what do you mean?
20170408 08:22:04< vultraz_iOS> Aginor: that the main game display is still the barrier?
20170408 08:22:06< vultraz_iOS> well, yes
20170408 08:23:44< vultraz_iOS> celticminstrel: ....I wanted to avoid the box just around the text :|
20170408 08:23:46< vultraz_iOS> why did you restore it
20170408 08:24:01< vultraz_iOS> also, the x = 20 was deliberate
20170408 08:24:11< vultraz_iOS> to line up with the 20 border in the text
20170408 08:24:17< celticminstrel> Because it looks better? Also, what's the point of fading out around the edges if there are no edges?
20170408 08:24:21< vultraz_iOS> since the title no longer matches the base rect
20170408 08:24:44< celticminstrel> Well, you could restore the x=20 if you don't mind the text not being centred.
20170408 08:25:01< vultraz_iOS> (it's to be debated if the text should match the base rect)
20170408 08:25:32< vultraz_iOS> celticminstrel: I imagined horizontal fadeout only, based on alignment
20170408 08:25:47< vultraz_iOS> left = full on the right, gradually fade out to the right
20170408 08:25:55< vultraz_iOS> vice-versa for right
20170408 08:26:00< vultraz_iOS> and fadeout on both sides for centered
20170408 08:26:22< celticminstrel> I was thinking of something along the lines of the caption in the middle of this: http://celmin.pwcsite.com/oboe/boe-cutscene-sample.png
20170408 08:26:48< celticminstrel> Which reminds me that you wanted rounded rectangles...
20170408 08:28:31< vultraz_iOS> neh, forget that
20170408 08:28:33< vultraz_iOS> meh*
20170408 08:29:55< vultraz_iOS> fuzzy borders are not modern UX
20170408 08:30:31< celticminstrel> Says the person who wants alpha fades everywhere.
20170408 08:30:52< vultraz_iOS> alpha fades are accaptable
20170408 08:31:04< celticminstrel> How is that different than an alpha fade out from the centre?
20170408 08:32:01< vultraz_iOS> it just is
20170408 08:32:14< celticminstrel> It's still an alpha fade.
20170408 08:32:22< vultraz_iOS> but very fuzzy
20170408 08:32:36< celticminstrel> Yes, alpha fades are fuzzy.
20170408 08:32:39< celticminstrel> Obviously.
20170408 08:33:22< celticminstrel> I just thought of something that could give it a slightly faded border - set border_width = 1 or 2 and set border_color = 0,0,0,100 or so.
20170408 08:33:52< vultraz_iOS> perhaps, perhaps
20170408 08:41:39 * vultraz_iOS still pondering the tiling...
20170408 08:42:17< celticminstrel> Tiling is relatively easy TBH, assuming you have direct access to a surface and can blit to it.
20170408 08:42:33< vultraz_iOS> oh, sure, it's easy
20170408 08:42:41< vultraz_iOS> the question is how to interface it with the canvas
20170408 08:43:06< vultraz_iOS> esp since it already has tiling
20170408 08:43:11< vultraz_iOS> but it's the wrong tiling!
20170408 08:43:12< celticminstrel> Eh?
20170408 08:43:25< vultraz_iOS> the canvas has resize_mode = tile
20170408 08:43:37< vultraz_iOS> but that's centered-outward no-scale tiling in the provided rect
20170408 08:43:50< celticminstrel> So resize_mode = tile means...
20170408 08:43:54< celticminstrel> Tile from top left?
20170408 08:44:05< vultraz_iOS> er, top-left tiling, yes
20170408 08:44:18< vultraz_iOS> the storyscreen wants possibly scaled centered tiling
20170408 08:44:35< vultraz_iOS> now, since i made tile_surface take a centered argument, that can be changed
20170408 08:44:35< celticminstrel> So add a few more values for resize_mode?
20170408 08:44:56< vultraz_iOS> possibly
20170408 08:45:05< vultraz_iOS> or im pondering if multiple resize modes could be added
20170408 08:45:07< vultraz_iOS> er
20170408 08:45:09< vultraz_iOS> combined
20170408 08:45:24< vultraz_iOS> so scale, tile would scale, then tile
20170408 08:45:38< vultraz_iOS> but I'd also have to add a [tile_properties] tag
20170408 08:45:39< celticminstrel> Makes sense.
20170408 08:45:48< vultraz_iOS> and i dunno if that's reasonable..
20170408 08:45:50< celticminstrel> You could add tile_properties
20170408 08:45:52< vultraz_iOS> or...
20170408 08:45:54< celticminstrel> Or you could just add tile_centered
20170408 08:46:02< celticminstrel> resize_mode = scale, tile_centered
20170408 08:46:23< celticminstrel> The other problem is that the story viewer wants different modes horizontally and vertically.
20170408 08:46:26< vultraz_iOS> perhaps tiling occupies the entire screen for the storyscreen
20170408 08:46:33< vultraz_iOS> not the image rect
20170408 08:46:38< vultraz_iOS> like the canvas expects
20170408 08:47:16< vultraz_iOS> perhaps resize_mode can be made a [resize] subtag
20170408 08:48:07< wedge009> Aginor: Good to see you're still around. I know that floating-point comparisons are generally bad. Also recognise that the ratio is usually a constant as celticminstrel mentioned. What would be a sensible compromise? < 0.99 && > 1.01 ?
20170408 08:48:14< vultraz_iOS> then again, that would conflict with IPFs :P
20170408 08:48:21< vultraz_iOS> or, not really conflict
20170408 08:48:31< vultraz_iOS> just seems like a little bit of a duplication..
20170408 08:48:59< Aginor> wedge009: I'm mostly not around nowadays to be honest
20170408 08:49:22< wedge009> Aginor: I probably meant 'around' as in 'alive'. ;)
20170408 08:49:29< vultraz_iOS> Aginor: hoping you can work on the hotkey PR before the next release
20170408 08:50:06< Aginor> vultraz_iOS: I bet you can never guess what I'm doing right now
20170408 08:50:28-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20170408 08:50:29< vultraz_iOS> are you working on the hotkey stuff? 
20170408 08:50:39< Aginor> yes
20170408 08:50:57< vultraz_iOS> :O
20170408 08:51:07< vultraz_iOS> \o/
20170408 08:51:08 * wedge009 is laughing at this little exchange
20170408 08:51:43< vultraz_iOS> celticminstrel: i wonder if would be worth adding some of the storyscreen stuff to the canvas...
20170408 08:51:49< vultraz_iOS> preserve ratio, etc..
20170408 08:52:30< vultraz_iOS> probably not
20170408 08:54:16< celticminstrel> The stuff you've already implemented, probably not, I guess.
20170408 08:54:55< vultraz_iOS> yeah...
20170408 08:54:58-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 256 seconds]
20170408 08:55:00< Aginor> refactor it and keep it with something that's responsible for that kind of stuff instead of somewhere mostly unrelated
20170408 08:55:25< vultraz_iOS> Aginor: I don't do any manual surface calculations in my new storyscreen
20170408 08:55:56< vultraz_iOS> everything is delegated to the canvas via the construction of formulas
20170408 08:56:13< vultraz_iOS> was just wondering if any of that should be controlled by single wml key
20170408 09:21:01< zookeeper> hmh, problem 1 seems to still exist in master: https://forums.wesnoth.org/viewtopic.php?f=21&t=45210
20170408 10:02:33-!- gfgtdf [~chatzilla@x4e363f37.dyn.telefonica.de] has joined #wesnoth-dev
20170408 10:12:30< vultraz_iOS> zookeeper: you can probably franken some nice elves from here https://www.wesnoth.org/jetrel/Wesnoth/elf-archer-comp.png
20170408 10:13:19< celticminstrel> Speaking of which, are we going to use that sylph ever? Or the shyde?
20170408 10:13:31-!- JyrkiVesterinen [~JyrkiVest@87-100-188-214.bb.dnainternet.fi] has quit [Quit: .]
20170408 10:13:36< vultraz_iOS> when he's done :P
20170408 10:13:47< celticminstrel> Sure looks done to me. Barring animations obviously.
20170408 10:13:55< vultraz_iOS> of course
20170408 10:14:02< celticminstrel> Oh wait, there's also a new druid and enchantress?
20170408 10:14:08< celticminstrel> Or is it a new sorceress?
20170408 10:14:11< vultraz_iOS> there's new everyone
20170408 10:14:15< vultraz_iOS> well, there eventually will be
20170408 10:14:18< celticminstrel> Aha.
20170408 10:14:34< celticminstrel> I guess there is indeed.
20170408 10:14:53< celticminstrel> So by "done" you mean "when the seven black boxes are filled"?
20170408 10:15:06< vultraz_iOS> yes
20170408 10:15:19< celticminstrel> The druid also looks unfinished. Her staff is flat.
20170408 10:15:24< vultraz_iOS> though i did do a Lord sprite with help from .. enchi?
20170408 10:15:25< vultraz_iOS> i think
20170408 10:15:26< vultraz_iOS> and yes
20170408 10:15:39< celticminstrel> I think it's the only one though.
20170408 10:16:00< vultraz_iOS> https://www.dropbox.com/s/e1ok2bfk3yc6nl9/Vult%20Elves.png?dl=0
20170408 10:18:31< vultraz_iOS> it is, sadly impossible to get him to do anything
20170408 10:18:35< vultraz_iOS> :)
20170408 10:18:37< celticminstrel> ...why the heck is the high lord wielding a glaive?
20170408 10:18:45< vultraz_iOS> why not?
20170408 10:18:55< celticminstrel> Because polearms aren't very elvish?
20170408 10:19:20< celticminstrel> Oh, there's a finished version of the druid from before.
20170408 10:19:32< vultraz_iOS> i added to the staff
20170408 10:20:00< zookeeper> yes and then when we use those i bet someone is going to biggerize all other units too, right?
20170408 10:20:28< vultraz_iOS> he means to decrease the biggerism
20170408 10:20:32< vultraz_iOS> then again, I heard that 6 years ago
20170408 10:20:45 * vultraz_iOS is salty
20170408 10:21:09< zookeeper> maybe we should appoint an art director
20170408 10:21:43< zookeeper> oh wait, we don't even have any candidates
20170408 10:22:12< vultraz_iOS> :) :) :) :)
20170408 10:24:09< vultraz_iOS> what's frustrating is...
20170408 10:24:12< vultraz_iOS> I know he's busy
20170408 10:24:17< vultraz_iOS> and has other stuff to do
20170408 10:24:38 * zookeeper maintains that the only sensible long-term choice is to eliminate sprite animations
20170408 10:24:41 * zookeeper goes afk
20170408 10:24:47< vultraz_iOS> but... i mean, he committed to at least the desert elves and i cannot for the *Life* of me get an answer as to whether he'll finish or wehn
20170408 10:38:46-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20170408 10:42:48-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20170408 10:46:18-!- aidanhs [~aidanhs@81.4.110.234] has joined #wesnoth-dev
20170408 11:16:23< celticminstrel> So any objections to merging wedge009's PR?
20170408 11:18:24< irker070> wesnoth: Celtic Minstrel wesnoth:wml_tag_porting 3c5dfe76da9e / / (12 files in 5 dirs): Properly port [music] to Lua https://github.com/wesnoth/wesnoth/commit/3c5dfe76da9e803a686e7a8402d5cb3d4c23c75c
20170408 11:18:26< irker070> wesnoth: Celtic Minstrel wesnoth:wml_tag_porting 411f4d9954b2 / src/scripting/ (game_lua_kernel.cpp lua_common.hpp lua_unit.cpp): Lua: Add float mutator macros and some fixups/cleanup to other mutator/accessor  https://github.com/wesnoth/wesnoth/commit/411f4d9954b218a49ffc7fab85de2be220e80ad1
20170408 11:18:28< irker070> wesnoth: Celtic Minstrel wesnoth:wml_tag_porting 4efcfb5bf974 / / (6 files in 4 dirs): Port [volume] to Lua https://github.com/wesnoth/wesnoth/commit/4efcfb5bf974257911f4a05b1dd3196434f54efe
20170408 11:18:30< irker070> wesnoth: Celtic Minstrel wesnoth:wml_tag_porting 953da8b80c47 / data/lua/wml-tags.lua src/scripting/lua_audio.cpp: fixup! Properly port [music] to Lua https://github.com/wesnoth/wesnoth/commit/953da8b80c47e92c607029e342cefb4adb7c4369
20170408 11:21:23-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 11:23:15< irker070> wesnoth: Celtic Minstrel wesnoth:master 9b2eac1920bb / data/gui/widget/label_story_screen_title.cfg: Simplify storyscreen title definition https://github.com/wesnoth/wesnoth/commit/9b2eac1920bbf21fca8a274939b2cb8d1411278c
20170408 11:37:13< irker070> wesnoth: Andreas Löf wesnoth:composed_hotkeys d4da5fe9f37e / src/ (5 files in 2 dirs): Fix 24681: Proper filtering of modifiers when handling hotkeys https://github.com/wesnoth/wesnoth/commit/d4da5fe9f37e8d98dc19c434fd61c358a58b0687
20170408 11:41:51< vultraz_iOS> Aginor: could you potentially rebase (or probably should just make a new branch to avoid travis spam)? especially hotkey_preference_display doesn't exist any more and that is now handled in a GUI2 dialog
20170408 11:41:59< vultraz_iOS> of which I need to fix handling off...
20170408 11:42:08< vultraz_iOS> not sure of the is_hotkeyable_event will help.
20170408 11:42:10< vultraz_iOS> perhaps
20170408 11:42:37< vultraz_iOS> and the stuff in create_hotkey is also in said dialog too
20170408 11:42:47< vultraz_iOS> er
20170408 11:42:49< vultraz_iOS> wait, no
20170408 11:42:51< vultraz_iOS> ignore that
20170408 11:42:58< vultraz_iOS> I just split it into two functions in the same file
20170408 11:43:31< vultraz_iOS> (gui dialog is gui/dialogs/hotkey_bind.cpp*
20170408 11:48:33< Aginor> I don't follow what you're talking about
20170408 11:51:19< vultraz_iOS> The hockey bind dialog is now gui2 in master (and suffering from the gui2 issue of text input and key down events handled by the same gui2 event )
20170408 11:52:27< Aginor> by the looks of it, it's suffering from more than that
20170408 11:52:28< vultraz_iOS> I'm asking if A: you could rebate, since hotkey_preferences_display.hpp no longer exists in master and some of your most recent changes to hotkey_item.cpp (those in the create_hotkey function) would be places slightly differently in master as I split that function into two
20170408 11:53:13< vultraz_iOS> And B tell me if your new is_hotkeyable_event function could be used to filter out modifier key events for the dialog
20170408 11:53:26< Aginor> B: yes
20170408 11:53:29< vultraz_iOS> In order to fix the bug where hotkeys with modifiers can no longer be set
20170408 11:53:49< Aginor> you should use B then listening to events while binding
20170408 11:53:49< vultraz_iOS> Or whether the gui2 pipeline has to be changes to handle key down and text input differently
20170408 11:53:55< vultraz_iOS> As you spoke of a few weeks ago
20170408 11:53:59< Aginor> It still should
20170408 11:54:25< vultraz_iOS> But this function can serve as a temp fix, yes?
20170408 11:54:41< Aginor> it's orthogonal
20170408 11:54:55< Aginor> you need that bit for your hotkey handling
20170408 11:55:04< Aginor> text input handling in GUI2 is different
20170408 11:55:15< Aginor> that should also be fixed
20170408 11:56:10< vultraz_iOS> Yes
20170408 11:56:12< vultraz_iOS> It should
20170408 11:56:39< Aginor> let's resurrect the GUI1 dialog instead
20170408 11:56:53< vultraz_iOS> I'd rather not
20170408 11:56:58< vultraz_iOS> And just fix gui2
20170408 11:57:45< vultraz_iOS> Which I will get to
20170408 11:58:28< vultraz_iOS> But right now I want to know if I can somehow make the new dialog allow setting of hotkeys with modifier keys with your latest changes
20170408 11:58:39< vultraz_iOS> THEN I can split the gui2 input handling later
20170408 11:59:04< Aginor> yeah, you just need to make the new dialog do the same thing as the old dialog
20170408 11:59:08< Aginor> and you're sorted
20170408 12:00:07< vultraz_iOS> I can't do it exactly the same (nor do i even understand what it was doing) because the event API is different!
20170408 12:00:09< Aginor> pass the events to the hotkey subsystem and let it decide what's appropriate events or not
20170408 12:00:47< celticminstrel> vultraz_iOS: Wasn't there a way for GUI2 dialogs to subscribe to the raw SDL events too?
20170408 12:01:03< vultraz_iOS> No
20170408 12:01:39< vultraz_iOS> There are some gui2 events that correspond to the raw SDL events but the raw event is never passed through
20170408 12:01:53< vultraz_iOS> So I cannot do as Aginor suggested
20170408 12:02:15< Aginor> that makes my PR useless then
20170408 12:02:22< Aginor> I'll go close it
20170408 12:02:46< vultraz_iOS> All I can do is subscribe to KEY_DOWN, which is fired by SDL_KEY_DOWN and TEXT__INPUT
20170408 12:03:07< vultraz_iOS> Why in hell do we keep going around in circles here
20170408 12:03:46< vultraz_iOS> I have a feeling we have at least 3 different issues here but I cannot seem to get a straight damn answer as to WHAT THEY ARE and HOW TO FIX THEM
20170408 12:04:38< vultraz_iOS> What was the issue that promoted the PR originally? How does it relate to the GUI 2 event system? How should the GUI2 event system interact with the hotkey system?
20170408 12:04:59< vultraz_iOS> Included in the second point are things like whether I should add forwarding or the raw SDL event to dialogs!
20170408 12:05:10< Aginor> vultraz_iOS: go and re-read the PR and the bug referenced :)
20170408 12:05:47< vultraz_iOS> I'm not asking you to close your PR I'm just trying to understand this mess!
20170408 12:05:58< Aginor> ok
20170408 12:06:29 * celticminstrel sighs.
20170408 12:06:35< vultraz_iOS> There are *multiple* design changes that have to be made here and they're all mixed up
20170408 12:06:41< Aginor> I'm trying hard to not be too frustrated here right now, but it's just after midnight and I'm pretty tired
20170408 12:07:20< vultraz_iOS> I'm trying to fix the gui2 dialog and you're talking about scancodes and keycodes and events and composed and not and I have no idea what to do anymore!
20170408 12:07:24-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170408 12:07:35< vultraz_iOS> I don't know what affects what or who or where when how anything!
20170408 12:07:36< celticminstrel> I'll take a look at Aginor's PR soon, okay?
20170408 12:08:30< Aginor> my take on the problem is that GUI2 is lacking some of the event handling stuff because it's masking the raw events, which the hotkey system consumes
20170408 12:09:03-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 12:09:20< Aginor> my PR relies on the raw event handling to deal with text composition so we can go back to having a locale-independant ":" as a hotkey
20170408 12:09:21< vultraz_iOS> I split the create_hotkey function in two and made it take a scan code
20170408 12:09:36< vultraz_iOS> Using SDL_GetScancodeFromKey
20170408 12:10:15< Aginor> As a part of that, I also went away from scancodes and used keysyms instead, since that also means that keys won't change between languages (z,y on a german vs us keyboard for example)
20170408 12:10:32< vultraz_iOS> GUI2 uses keycodes internally
20170408 12:10:58< vultraz_iOS> https://github.com/wesnoth/wesnoth/blob/master/src/gui/dialogs/hotkey_bind.cpp#L39 as you can see I hook into KEY_DOWN (the GUI2 event) here
20170408 12:11:13-!- Shiki [~Shiki@p54855738.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170408 12:11:29< vultraz_iOS> the dispatcher pulls the keycode from the SDL_KEY_DOWN event data and passes it to the event
20170408 12:11:32< vultraz_iOS> so I *have it*
20170408 12:11:38< vultraz_iOS> BUT
20170408 12:11:44< Aginor> vultraz_iOS: that's insufficient
20170408 12:11:54< vultraz_iOS> gui2 *also* handles TEXT_INPUT with the *same gui2 functions*
20170408 12:12:00< vultraz_iOS> so my handler here is fired *twice*
20170408 12:12:08< vultraz_iOS> that's why I filter out SDLK_UNKNOWN
20170408 12:12:11< Aginor> vultraz_iOS: as I told you, that should change
20170408 12:12:26< vultraz_iOS> since that's the code set when this event fired by TEXT_INPUT events
20170408 12:12:27< vultraz_iOS> yes
20170408 12:12:27< Aginor> add a proper event handler for it
20170408 12:12:41< vultraz_iOS> I can do that easily
20170408 12:12:50< Aginor> and make the text field do textinput events for text inputs while you're at it
20170408 12:13:08< vultraz_iOS> *if that is done*, will this dialog once again be able to set composed hotkeys!
20170408 12:13:30< vultraz_iOS> that is, I assume I need to change it to be fired by TEXT_INPUT
20170408 12:13:35-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 12:13:36-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20170408 12:13:46< Aginor> vultraz_iOS: not entirely
20170408 12:14:06< Aginor> the best thing you can do there is to simply let the hotkey system consume the raw events
20170408 12:14:37< Aginor> although I guess you could always go and plumb through all hotkeyable event classes
20170408 12:14:46< vultraz_iOS> here are the new create_hotkey functions from master 
20170408 12:14:47< vultraz_iOS> https://github.com/wesnoth/wesnoth/blob/master/src/hotkey/hotkey_item.cpp#L158-L188
20170408 12:15:12< vultraz_iOS> I had *though* it was enough to pass the data I got from the GUI2 events
20170408 12:15:15< Aginor> yeah, I looked at them
20170408 12:15:49< vultraz_iOS> I cannot fully understand *why* the hotkey system needs the raw event
20170408 12:16:45< Aginor> keyboard events, mouse click events, joystick events, joyhat events, potentially gamepad events in the future
20170408 12:16:58< Aginor> and the text input events
20170408 12:17:11< vultraz_iOS> the heck is a joyhat
20170408 12:17:12< Aginor> but yes, you can make your dialog listen to all of that
20170408 12:17:24< Aginor> vultraz_iOS: an input device
20170408 12:17:49< vultraz_iOS> I *can* make a special event that simply forwards the raw data to the hotkey system
20170408 12:18:05< Aginor> http://marsxplr.com/_upl/aviatorlogitechjse-orig.jpg
20170408 12:18:12< Aginor> someone made a picture already
20170408 12:18:38< Aginor> incidentially, those are usually mapped as the d-pad on a gamepad
20170408 12:19:05< Aginor> alternatively, just rip out *all* of the joystick code
20170408 12:19:42< irker070> wesnoth: gfgtdf wesnoth:master fe098418f781 / src/config.cpp: ignore blank (unset) values when generating config checkums. https://github.com/wesnoth/wesnoth/commit/fe098418f78189aa847037f804d3ad18e9472acb
20170408 12:19:58< vultraz_iOS> I can easily add additional listeners to this dialog as necessary
20170408 12:20:09< vultraz_iOS> it already hooks separate events for mouse and keyboard 
20170408 12:20:52< vultraz_iOS> and simply call the appropriate create_hotkey overload
20170408 12:21:35< vultraz_iOS> but I'm ignoring joysticks and gamepads and controllers for now
20170408 12:21:48< vultraz_iOS> my *immediate* concern is fixing the handling of composed hotkeys with this dialog
20170408 12:22:58< vultraz_iOS> so im disregarding *any other input types*
20170408 12:23:05< vultraz_iOS> besides mouse and keyboard
20170408 12:23:13< vultraz_iOS> as I said, I can add handlers for them when the time comes
20170408 12:24:07< vultraz_iOS> we're not handling any hotkey events here
20170408 12:24:18< vultraz_iOS> GUI2 has to get to that eventually but not now
20170408 12:24:55< Aginor> I'm going to bed now, I'll see if I can have a look next weekend
20170408 12:26:12-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20170408 12:26:32< celticminstrel> I have round rectangle! \o/
20170408 12:26:34< vultraz_iOS> Aginor: barring that, what I'm understanding is that, say, pressing SHIFT would be a SDL_KEY_DOWN event but NOT TEXT_INPUT. Pressing, say, 'a', would be TEXT_INPUT, NOT SDL_KEY_DOWN. But somehow that doesn't seem right. *if that is not so*, I can also add filtering, possibly using your code, to the dispatcher system. But that's basically what I seem to
20170408 12:26:34< vultraz_iOS> understand you advocating
20170408 12:26:46< celticminstrel> The corners don't really look good with a border_thickness > 1, but whatever.
20170408 12:27:14< Aginor> vultraz_iOS: the event system shouldn't filter,t he recipient should
20170408 12:27:43< Aginor> shift+a would generate two keydown events and one textinput event (for 'A')
20170408 12:28:03< Aginor> ctrl+a would generate two keydown events
20170408 12:28:14< vultraz_iOS> ...oh
20170408 12:28:18< vultraz_iOS> ok see I didn't know that
20170408 12:28:30< vultraz_iOS> I thought there was built-in filtering between non-text keys and text keys
20170408 12:28:33< vultraz_iOS> and their events
20170408 12:29:28< Aginor> so did I, until I finished the work to deal with it, but you seemed to be more interested in ranting than talking about it
20170408 12:29:51< irker070> wesnoth: Celtic Minstrel wesnoth:master 76dcfcb32f68 / data/gui/schema.cfg src/gui/core/canvas.cpp: GUI2 Canvas: Implement rounded rectangle shape https://github.com/wesnoth/wesnoth/commit/76dcfcb32f68f20a2fd75554a0b7c973a2dd34b8
20170408 12:29:52< Aginor> I'll however finish this work
20170408 12:30:13< vultraz_iOS> We just need to coordinate.
20170408 12:30:17< Aginor> but I'm not going to commit to doing any future work
20170408 12:30:39< vultraz_iOS> You're dealing with the hotkey code, I'm dealing with the GUI code, and we've hit a point where they intersect and everyone's getting confused.
20170408 12:30:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 260 seconds]
20170408 12:31:51< Aginor> vultraz_iOS: you went and replaced the gui code only after I raised my original PR, did not coordinate with me and now you're upset that we're not on the same page. I'm finding that frustrating.
20170408 12:32:21< vultraz_iOS> I didn't realize your PR affected that area 
20170408 12:32:27< celticminstrel> ...
20170408 12:32:27< vultraz_iOS> I'm sorry.
20170408 12:32:45< vultraz_iOS> I thought it had to do with hotkey *handling* and key vs scancodes
20170408 12:32:49< Aginor> I had changes in the dialogue you removed.
20170408 12:33:05< Aginor> (I think)
20170408 12:33:36-!- Shiki [~Shiki@p54855738.dip0.t-ipconnect.de] has left #wesnoth-dev ["Verlassend"]
20170408 12:33:37< vultraz_iOS> You only just added changes there.
20170408 12:33:45< Aginor> fair enough
20170408 12:33:53-!- Shiki [~Shiki@p54855738.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170408 12:36:42< vultraz_iOS> again, I'm sorry.
20170408 12:37:33< vultraz_iOS> if I had realized how much my change would affect your changes and vice versa I wouldn't have converted the dialog
20170408 12:37:41< vultraz_iOS> until after
20170408 12:38:57< Aginor> all good, I'll look at cleaning up the mess next weekend
20170408 12:50:45< Aginor> good night
20170408 12:51:18-!- JyrkiVesterinen [~JyrkiVest@87-100-188-214.bb.dnainternet.fi] has joined #wesnoth-dev
20170408 13:00:36-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170408 13:02:16-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 13:05:22-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20170408 13:08:10-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 264 seconds]
20170408 13:43:10-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20170408 13:45:27-!- Kwandulin [~Kwandulin@p200300760F3E7DF149CEF7081D588286.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20170408 13:47:10-!- Shiki [~Shiki@p54855738.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20170408 13:54:06< celticminstrel> So vultraz_iOS, are you going to try doing the buttons with the new round_rectangle shape?
20170408 13:54:16< vultraz_iOS> I can try
20170408 13:54:35< celticminstrel> It's the same as [rectangle], plus corner_radius
20170408 14:02:48< celticminstrel> Currently I'm attempting to implement filled circles.
20170408 14:02:58< celticminstrel> Because this is a prerequisite for implementing filled rounded rects.
20170408 14:03:33< vultraz_iOS> would be nice if the borders were aliased 
20170408 14:04:03< celticminstrel> Aliasing is hard.
20170408 14:04:11< vultraz_iOS> indeed
20170408 14:04:21< celticminstrel> Or at least... harder.
20170408 14:04:28< vultraz_iOS> maybe we shouldn't bother
20170408 14:04:51< celticminstrel> It's definitely doable.
20170408 14:05:25< celticminstrel> You just need to discover the algorithm and apply it. Which is probably harder than it sounds, but still.
20170408 14:07:17< vultraz_iOS> we should probably wait until we have OGL or something 
20170408 14:07:53< celticminstrel> I mean, it's not like you need OGL for it. IIRC it's a relatively straightforward extension of a non-aliased algorithm.
20170408 14:08:27< celticminstrel> Roughly, having some concept of the "percentage" of a pixel the line passes through.
20170408 14:08:46< celticminstrel> But I don't really see any need for aliasing, either.
20170408 14:09:14< celticminstrel> I wonder if SDL does aliasing on its line drawing.
20170408 14:09:37< celticminstrel> If so, you could probably get aliasing on the circle by duplicating the first point at the end and using DrawLines instead of DrawPoints.
20170408 14:09:58< celticminstrel> Although that also assumes the points are sorted in a geometric order.
20170408 14:10:10< celticminstrel> Which I'm pretty sure isn't actually the case.
20170408 14:12:40< vultraz_iOS> I actually don't use rectangles for the buttons
20170408 14:12:46< celticminstrel> I know.
20170408 14:12:55< celticminstrel> Because there were no rounded rectangles, right?
20170408 14:13:14< vultraz_iOS> yes, so i refactored to avoid "erasing" corners with 0 alpha points
20170408 14:13:18< vultraz_iOS> but...
20170408 14:13:28< celticminstrel> You used different colours for each side or something?
20170408 14:13:30< vultraz_iOS> given the shaded sides I'd still have to then draw over the sides with lines
20170408 14:13:53< vultraz_iOS> though...
20170408 14:13:54-!- stikonas_ is now known as stikonas
20170408 14:14:00< vultraz_iOS> at least some might be able to use it
20170408 14:14:01< vultraz_iOS> I'll see
20170408 14:14:02< celticminstrel> I'd think you wouldn't draw right over them, more like next to them.
20170408 14:14:28< celticminstrel> But anyway, it's up to you really.
20170408 14:14:49< vultraz_iOS> thanks for adding it though
20170408 14:14:59< celticminstrel> I intend to try out the round rectangle in the storyscreen, which is why I added it (and why I'm now trying to figure out filled circles).
20170408 14:15:20< vultraz_iOS> rounded rectangles in the storyscreen?
20170408 14:15:24< vultraz_iOS> this might not look good...
20170408 14:15:27< celticminstrel> For the title background?
20170408 14:15:49< vultraz_iOS> we'll have to see though
20170408 14:15:52< celticminstrel> Anyway, I intend to try it out. If it doesn't look good we don't need to keep it.
20170408 14:17:01< celticminstrel> Oh yeah, it should be obvious from things I've said, but [round_rectangle] currently ignores fill_color.
20170408 14:18:42< vultraz_iOS> naturally
20170408 14:21:20< vultraz_iOS> honestly, im quite anxious to get HW 
20170408 14:21:25< vultraz_iOS> accl
20170408 14:21:27< vultraz_iOS> and OGL
20170408 14:21:33< vultraz_iOS> do you think we can do it before 2018?
20170408 14:26:47< stikonas> hi, is this a known bug? https://paste.kde.org/potjmfbrc
20170408 14:27:48< stikonas> probably not... Can't find anything similar amongts bug reports
20170408 14:28:21< stikonas> ok, found it... https://gna.org/bugs/?23834
20170408 14:28:40< stikonas> had to look for older bugs...
20170408 14:38:40< vultraz_iOS> agh
20170408 14:40:06< vultraz_iOS> i remember that
20170408 14:41:47< vultraz_iOS> celticminstrel: i think the unit.... something or another was missing there?
20170408 14:44:37< vultraz_iOS> i can't remember
20170408 14:45:38< vultraz_iOS> the unit map?
20170408 14:46:53< vultraz_iOS> (why was it so again? something to do with the game display.. or game map..)
20170408 14:53:54< mattsc> celticminstrel: I’ve done some poking around in/with FAI and everything seems to be working as before, AFAICT.
20170408 14:56:46< vultraz_iOS> :D
20170408 14:59:35-!- Kwandulin [~Kwandulin@p200300760F3E7DF149CEF7081D588286.dip0.t-ipconnect.de] has joined #wesnoth-dev
20170408 14:59:52< gfgtdf> propabilty resources::game_board, at lest tht the onyl resoruce that function seesm to be using
20170408 15:00:49< vultraz_iOS> ah
20170408 15:04:17< vultraz_iOS> should there be a game_board in the editor?
20170408 15:05:54< gfgtdf> propabily not since most of game_boards functions are quite game related like end_turn() check_victory(), not sure though
20170408 15:15:47< vultraz_iOS> hmm
20170408 15:15:55< vultraz_iOS> the map context owns a unit map
20170408 15:21:21-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20170408 15:31:03-!- irker070 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20170408 15:45:56-!- irker263 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20170408 15:45:56< irker263> wesnoth: Jyrki Vesterinen wesnoth:master fb3cae64dfed / src/gui/widgets/ (grid.cpp window.hpp): Fix a severe performance regression on filtering add-ons https://github.com/wesnoth/wesnoth/commit/fb3cae64dfed2275f18a5e19273b9ca2b1fef1f2
20170408 15:54:06-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20170408 15:54:13-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20170408 17:06:37-!- mkdroid [~null@unaffiliated/matthiaskrgr] has joined #wesnoth-dev
20170408 17:06:39-!- SigurdFD [~SigurdFD@dynamic-acs-72-23-110-196.zoominternet.net] has joined #wesnoth-dev
20170408 17:29:11< SigurdFD> I found a discrepancy/bug with store_sides that seems to have been around at least 6 years
20170408 17:29:39< SigurdFD> the wiki https://wiki.wesnoth.org/InternalActionsWML#.5Bstore_side.5D says:
20170408 17:29:56< SigurdFD> "income: Base income for this side (without villages)"
20170408 17:30:50< SigurdFD> but that's not what it provides, it gives base + village income
20170408 17:31:09< SigurdFD> what's described is not provided in any way
20170408 17:31:31< SigurdFD> according to the wiki logs, it's been like that since 2011
20170408 17:31:52< SigurdFD> and looking at the porting store_side to lua commit: https://github.com/wesnoth/wesnoth/commit/12093d87846740b3b5271804469a5fb53a9df20f
20170408 17:32:00< SigurdFD> it looks like it was that way before as well
20170408 17:33:14< SigurdFD> I'm thinking just to leave income as is, and add a key for base_income, and update the wiki accordingly
20170408 17:33:44< SigurdFD> thoughts?
20170408 17:33:56< SigurdFD> zookeeper^
20170408 17:34:02 * zookeeper knows nothing
20170408 17:34:58< zookeeper> but if it's true, then your plan sounds like the right one
20170408 17:35:08< SigurdFD> do you think the number of add-ons that would actually be looking for this info is small?
20170408 17:35:33< SigurdFD> ran across this with one of my add-ons a few years ago & again recently
20170408 17:36:21< zookeeper> couldn't say, it's not something i can really check
20170408 17:36:31< SigurdFD> ok
20170408 17:38:25< celticminstrel> vultraz_iOS: I've said several times before that the editor doesn't have a game_board, only a map_context.
20170408 17:38:36< vultraz_iOS> yes
20170408 17:38:47< SigurdFD> zookeeper: I'll make a pr for it then
20170408 17:38:48< vultraz_iOS> im working on the problem 
20170408 17:54:22-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 264 seconds]
20170408 18:03:33-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 18:06:33< irker263> wesnoth: Charles Dang wesnoth:master f9dadbb80aa5 / src/ (5 files in 3 dirs): Restored resources::units in a limited context https://github.com/wesnoth/wesnoth/commit/f9dadbb80aa5aaaa0cba44a5c865c16fa959ef4c
20170408 18:06:36< irker263> wesnoth: Charles Dang wesnoth:master 7364d728600e / src/editor/map/context_manager.cpp: Editor: properly set resources::filter_con https://github.com/wesnoth/wesnoth/commit/7364d728600ecd518053e09ac5495551042cb4a9
20170408 18:06:39< irker263> wesnoth: Charles Dang wesnoth:master 4bc0ac0041c3 / src/variable.cpp: Variable: remove asserts on resources::gamedata absence https://github.com/wesnoth/wesnoth/commit/4bc0ac0041c3e7bf928d8057ac94fc508e6ef0e4
20170408 18:06:42< irker263> wesnoth: Charles Dang wesnoth:master eaea9be11773 / src/units/abilities.cpp: Unit: re-deployed resources::units for ability handling (bug #23834) https://github.com/wesnoth/wesnoth/commit/eaea9be1177398b2fe1d06739c2eb43d1bd204e0
20170408 18:06:45< irker263> wesnoth: Charles Dang wesnoth:master 0d9806df59b1 / src/gui/core/canvas.cpp: GUI2/Canvas: fixed comment formatting causing a compiler warning https://github.com/wesnoth/wesnoth/commit/0d9806df59b17048b216fe5d7553afef47f96ce1
20170408 18:07:23< vultraz_iOS> celticminstrel: ^
20170408 18:13:52< irker263> wesnoth: Charles Dang wesnoth:master 887a99ecee14 / src/editor/controller/editor_controller.cpp: Fixup 7364d72 https://github.com/wesnoth/wesnoth/commit/887a99ecee1444210035c86d27cccf6fdf359ea3
20170408 18:14:35< vultraz_iOS> celticminstrel: crash seems to be fixed
20170408 18:16:46< vultraz_iOS> Not sure why so much stuff in variable.cpp relies on game data
20170408 18:30:42-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Read error: Connection reset by peer]
20170408 18:31:06-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20170408 18:31:10< stikonas> vultraz_iOS: yeah, I can confirm that it is fixed
20170408 18:31:47< vultraz_iOS> \ o /
20170408 18:33:50-!- SigurdFD [~SigurdFD@dynamic-acs-72-23-110-196.zoominternet.net] has quit []
20170408 18:41:52-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170408 18:42:04-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 19:13:48< gfgtdf> SigurdFD: i think [store_side] income shoudl returns the same as the the [side]income= key, everythign else sounds just confusing
20170408 19:14:45< gfgtdf> vultraz_iOS: well because variale.cpp get/sets the game varibles which are stores in game_data...
20170408 19:14:55< gfgtdf> vultraz_iOS: i mean there is not reall something else it does
20170408 19:38:05-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170408 19:40:16-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 19:44:05-!- mjs-de [~mjs-de@x4e304e6e.dyn.telefonica.de] has joined #wesnoth-dev
20170408 20:10:46< celticminstrel> vultraz_iOS: Isn't units accessible from the map_context though?
20170408 20:11:01< celticminstrel> Yeah, it is.
20170408 20:12:22< celticminstrel> Not sure how stuff in variable.cpp wouldn't rely on game data.
20170408 20:13:41< celticminstrel> Ugh, you had to touch canvas.cpp... :/
20170408 20:13:59< celticminstrel> Oh well, at least no conflicts.
20170408 20:14:31< celticminstrel> It's pretty lucky that that comment misformatting didn't break the build though.
20170408 20:24:21< mattsc> Build succeeded.  2348 warnings.  Wow.
20170408 20:24:45< celticminstrel> What kinds of warnings are these?
20170408 20:25:57< mattsc> I am not going to look at each of >2k warnings, but the vast majority seems to be of this type:
20170408 20:25:59< mattsc> 'wfl::variant_value_base' has virtual functions but non-virtual destructor
20170408 20:26:05< celticminstrel> Uh-oh.
20170408 20:26:09< celticminstrel> I'll fix that.
20170408 20:27:15< irker263> wesnoth: Celtic Minstrel wesnoth:master d9f771b5b91a / src/formula/variant_value.hpp: Fix non-virtual destructor in variant_value_base https://github.com/wesnoth/wesnoth/commit/d9f771b5b91ac25a7d5816ac5967e0d1d2059d84
20170408 20:27:43-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20170408 20:27:50< celticminstrel> Just for future reference, vultraz_iOS, always add a virtual destructor to any base class with virtual functions. Otherwise things may not be deleted properly.
20170408 20:27:55-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 20:28:11< celticminstrel> In the case of variant_value_base, it would've meant the actual values it allocated are leaked.
20170408 20:28:28< celticminstrel> (ie, the values allocated by some subclass of it)
20170408 20:29:06< celticminstrel> The problem is that, if you have a non-virtual destructor and call "delete base_class_ptr", only the base-class portion is destroyed.
20170408 20:32:08< mattsc> celticminstrel: thanks, recompiling; same seems to be happening with progressive_*, but much less frequently (so far)
20170408 20:32:52< celticminstrel> That surprises me a little, actually.
20170408 20:33:10< celticminstrel> BTW, can you tell me the warning flag for that warning?
20170408 20:33:36< celticminstrel> Usually the warning message includes the flag that would be needed to enable/disable that individual warning.
20170408 20:34:59< mattsc> -Wnon-virtual-dtor
20170408 20:36:34< mattsc> Down to 278 warnings.  First one is this:
20170408 20:36:41< mattsc> src/units/frame_private.hpp:27:7: warning: 'progressive_base<image::locator, image::locator>' has virtual functions but non-virtual destructor [-Wnon-virtual-dtor]
20170408 20:36:41< mattsc> class progressive_base
20170408 20:37:04< celticminstrel> Any warnings not in frame_private.hpp?
20170408 20:37:10< celticminstrel> Or not involving progressive_base?
20170408 20:43:26< mattsc> src/formula/formula.cpp:34:13: warning: no previous prototype for function 'tokens_to_string' [-Wmissing-prototypes]
20170408 20:43:26< mattsc> std::string tokens_to_string(const tk::token* i1, const tk::token* i2)
20170408 20:43:57< mattsc> ld: warning: direct access in return_to_play_side_exception::execute() to global weak symbol typeinfo for lua_jailbreak_exception means the weak symbol cannot be overridden at runtime. This was likely caused by different translation units being compiled with different visibility settings.
20170408 20:44:20< mattsc> Everything else seems to be frame_private.hpp.
20170408 20:44:40< celticminstrel> IIRC, the formula.cpp warning can be suppressed by using the static keyword. I have no idea how to fix the ld warnings (I was getting tons of those two when I still built on XCode).
20170408 20:44:53< irker263> wesnoth: Celtic Minstrel wesnoth:master 2eed69822bdf / SConstruct src/units/frame_private.hpp: Travis: Consider non-virtual destructors an error in base classes https://github.com/wesnoth/wesnoth/commit/2eed69822bdfa221d1787c600950b8170d7dfd2c
20170408 20:45:17< celticminstrel> Ah, come to think of it, that should've been tagged "scons" rather than "Travis". Oh well!
20170408 20:46:00< mattsc> recompiling ...
20170408 20:47:18< celticminstrel> Also, if someone wants to apply the same to CMake, that'd probably be good.
20170408 20:47:42< mattsc> Yes, those two types are the only left now, 14 warnings total (6 in formula.cpp; 8 ld warnings)
20170408 20:48:09-!- Kwandulin [~Kwandulin@p200300760F3E7DF149CEF7081D588286.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20170408 20:49:41< celticminstrel> Does the static keyword fix the formula.cpp warnings? Or I guess, wrapping the declarations in an anonymous namespace. (I can't remember if Wesnoth has Opinions on that.)
20170408 20:51:25-!- trewe [~trewe@2001:8a0:d130:e901:faa4:1f2e:419d:1e1c] has joined #wesnoth-dev
20170408 20:54:21< mattsc> Yes, it does (needed 6 times total).
20170408 20:54:34< celticminstrel> I guess you used static then.
20170408 20:54:49< mattsc> sorry?
20170408 20:54:59< celticminstrel> Well, I don't mind, at least.
20170408 20:55:03< mattsc> Oh, I see.
20170408 20:55:04< celticminstrel> As long as the warnings are gone.
20170408 20:55:13< mattsc> Yes, I used the static keyword.
20170408 20:55:20< celticminstrel> I think vultraz_iOS also started using static in some places for this purpose.
20170408 20:55:28< mattsc> Since I don’t know how to do the other one.  :P
20170408 20:56:29< mattsc> I can commit it, if you think that’s the right thing to do, but you’ll have to tell me what commit message I should use.
20170408 20:56:56< celticminstrel> Something along the lines of "fix warnings" should suffice. Presumably specifying what the warning was too.
20170408 20:57:19< celticminstrel> Alright, I finally successfully filled one octant of the circle!
20170408 20:57:21< celticminstrel> \o/
20170408 20:58:21< mattsc> Woot!
20170408 20:58:58< irker263> wesnoth: mattsc wesnoth:master e0d221fa4bc0 / src/formula/formula.cpp: Fix "no previous prototype" warnings https://github.com/wesnoth/wesnoth/commit/e0d221fa4bc0e925ecff904050165e7184108f25
20170408 21:03:11-!- JyrkiVesterinen [~JyrkiVest@87-100-188-214.bb.dnainternet.fi] has quit [Quit: .]
20170408 21:17:46-!- vultraz_iOS [uid24821@wesnoth/developer/vultraz] has quit [Quit: Connection closed for inactivity]
20170408 21:28:21-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20170408 21:29:10-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 21:29:43-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20170408 21:29:57-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 21:41:30< gfgtdf> celticminstrel: any idea on how to fix https://gna.org/bugs/?25639 ?
20170408 21:44:56-!- zookeeper [zookeeper@wesnoth/developer/zookeeper] has quit [Ping timeout: 260 seconds]
20170408 21:52:59< celticminstrel> And now I have the full circle filled with no overlaps! \o/
20170408 22:10:34-!- mjs-de [~mjs-de@x4e304e6e.dyn.telefonica.de] has quit [Remote host closed the connection]
20170408 22:14:58-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170408 22:16:34-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 22:19:50< celticminstrel> Fun fact: x + + r is valid code.
20170408 22:25:59-!- trewe [~trewe@2001:8a0:d130:e901:faa4:1f2e:419d:1e1c] has quit [Quit: quit]
20170408 22:46:44-!- mkdroid [~null@unaffiliated/matthiaskrgr] has quit [Quit: I'll be back!]
20170408 22:52:38-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20170408 22:56:25-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20170408 22:56:44-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 23:00:27< irker263> wesnoth: Celtic Minstrel wesnoth:master 2dbdc7ec66a1 / data/gui/schema.cfg data/gui/window/debug_clock.cfg src/gui/core/canvas.cpp: GUI2 Canvas: Enable drawing filled circles https://github.com/wesnoth/wesnoth/commit/2dbdc7ec66a114d2fce13067fb5fe86744deae04
20170408 23:00:29< irker263> wesnoth: Celtic Minstrel wesnoth:master f016ed96f418 / src/gui/core/canvas.cpp: GUI2 Canvas: Enable filled rounded rectangles https://github.com/wesnoth/wesnoth/commit/f016ed96f4187ba4eaf08be09546a258ed7c8aa3
20170408 23:00:31< irker263> wesnoth: Celtic Minstrel wesnoth:master 508a4937361b / src/gui/core/canvas.cpp: GUI2 Canvas: Draw rectangle fill before border https://github.com/wesnoth/wesnoth/commit/508a4937361b500d1fffc518cc0cc7b94bf2db55
20170408 23:00:33< irker263> wesnoth: Celtic Minstrel wesnoth:master 753df47137aa / data/gui/widget/label_story_screen_title.cfg: Story viewer: Use a slightly fuzzy rounded rectangle as the title background https://github.com/wesnoth/wesnoth/commit/753df47137aa3f23a365bd22f7fa5a8075abd8ef
20170408 23:04:49< irker263> wesnoth: Wedge009 wesnoth:master f8fa22ccc085 / src/units/frame.cpp: Prevent alpha underflow (bug #14503) https://github.com/wesnoth/wesnoth/commit/f8fa22ccc085d7dad238bb0510a1c94b4ae7b14c
20170408 23:04:51< irker263> wesnoth: Wedge009 wesnoth:master 9078ec5c55c9 / src/units/frame.cpp: Avoid floating-point equality comparisons. https://github.com/wesnoth/wesnoth/commit/9078ec5c55c95a0ccd94fe53265501887d78b6d3
20170408 23:04:53< irker263> wesnoth: Celtic Minstrel wesnoth:master 2df28aa3cb52 / src/units/frame.cpp: Merge pull request #961 from Wedge009/bug_14503_fix https://github.com/wesnoth/wesnoth/commit/2df28aa3cb52943d02dc60e1c00ec605b1af2c50
20170408 23:05:51< irker263> wesnoth: Jeffrey 'Sigurd' Westcoat wesnoth:master ec2d734acda3 / changelog data/lua/wml-tags.lua: WML: Add base_income key to [store_side] (#976) https://github.com/wesnoth/wesnoth/commit/ec2d734acda3b9ee8416db583a048fc09e612423
20170408 23:07:23-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20170408 23:07:36-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20170408 23:11:08< celticminstrel> So gfgtdf, is the [upload_log] just some random data injected into the save file?
20170408 23:11:22< celticminstrel> It doesn't seem to be used for anything.
20170408 23:11:36< gfgtdf> celticminstrel: afaik yes, it insde [replay] and somhow citains ai related data
20170408 23:11:57< gfgtdf> contains*
20170408 23:13:14< celticminstrel> It's basically only used by src/ai/testing.?pp...
20170408 23:13:23< celticminstrel> mattsc: I don't suppose you know what that file is for?
20170408 23:13:29< celticminstrel> ^those files
20170408 23:15:29< celticminstrel> It also seems like that the class in that file is referenced nowhere.
20170408 23:18:07-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20170408 23:36:49< mattsc> celticminstrel: well, those files produce debug output about the AI to stdour when the game is started with the --log-debug=ai/testing CL option
20170408 23:36:50< gfgtdf> celticminstrel: ye maybe it meant to be hand read by people
20170408 23:37:26< gfgtdf> celticminstrel: unfrtuneteley the people who wrote the original ai c++ code aren't there anymore
20170408 23:37:45< celticminstrel> mattsc: Well yes, they would do that... if those functions were ever called from anywhere.
20170408 23:38:05< mattsc> celticminstrel: I just tried in master, it still works
20170408 23:38:11< gfgtdf> celticminstrel: the commit which ads most of these logs seems to be this one https://github.com/wesnoth/wesnoth/commit/6e3ec07d968dff34aba5ca1e1aaa6c27ddbcd346
20170408 23:39:35< mattsc> celticminstrel: I am talking about testing.?pp, not upload_log  (but I assume that was clear)
20170408 23:40:15-!- vultraz_iOS [uid24821@wesnoth/developer/vultraz] has joined #wesnoth-dev
20170408 23:40:42< mattsc> E.g.: “20170408 16:35:51 debug ai/testing: TURN_START_UNITS_COST3: 48”
20170408 23:41:20< mattsc> and the only place where UNITS_COST appears in the code is in testing.cpp
20170408 23:42:00< mattsc> Those files were originally introduced by Crab when he did his refactoring of the AI as part of his GSoC project:
20170408 23:42:01< mattsc> https://github.com/wesnoth/wesnoth/commit/ab540dca1517e82d0f93153604cdb00f256becd6
20170408 23:42:44< celticminstrel> Hmm...
20170408 23:43:55< celticminstrel> Okay, so I assume removing them is not a good idea then?
20170408 23:44:12< mattsc> I don’t know …
20170408 23:44:21< mattsc> Is there a reason to remove them?
20170408 23:44:34< celticminstrel> Not really, TBH.
20170408 23:44:48< mattsc> I currently don’t see that information as particularly useful, but one never knows what might come in handy at some point.
20170408 23:44:53< celticminstrel> Hmm.
20170408 23:45:00< celticminstrel> Well, okay.
20170408 23:45:14< mattsc> Plus, once the code is there, it is much easier to add new information that might be more useful than if you have to do it from scratch.
20170408 23:45:28< mattsc> But having said that, I personally currently see no need for it …
20170408 23:45:52< mattsc> For example, I output similar information for my Fred AI, but I produce that in the Lua code.
20170408 23:46:22< vultraz_iOS> So that's what a virtual dtor is for
20170408 23:54:19< gfgtdf> vultraz_iOS: you need a virtual dtor whenever you have delete toa  pointer to a differnt class
20170408 23:54:47< gfgtdf> vultraz_iOS: in: class A : B {}; B* b = new A(); delete b;
20170408 23:55:08< gfgtdf> vultraz_iOS: without a virtual dtor it woudl just call B::~B() instead of a A::~A()
20170408 23:55:19< vultraz_iOS> ah, I see
20170408 23:58:07< gfgtdf> actually it is a little more complicated since the memeory of the object itself also needs to be freed
20170408 23:59:11< celticminstrel> If you had A* a = new A() though, delete a would do what you want. Just FTR.
20170408 23:59:35< celticminstrel> Basically a non-virtual destructor means the destructor is resolved at compile-time based on the type of the pointer.
--- Log closed Sun Apr 09 00:00:19 2017