--- Log opened Tue Oct 04 00:00:45 2016
20161004 00:10:56-!- Duthlet [~Duthlet@dslb-188-104-253-155.188.104.pools.vodafone-ip.de] has quit [Quit: leaving]
20161004 00:18:12< irker795> wesnoth: mattsc wesnoth:master 6d7680494de0 / src/ai/default/ (ca_move_to_targets.cpp ca_move_to_targets.hpp): Default AI MtT CA: do not evaluate movemaps until needed https://github.com/wesnoth/wesnoth/commit/6d7680494de0b9f24dc7614fedb53cf9d58e7473
20161004 00:18:14< irker795> wesnoth: mattsc wesnoth:master eefc28c13873 / src/ai/default/ca_move_to_targets.cpp: Default AI MtT CA: remove guardian moves before moving other units https://github.com/wesnoth/wesnoth/commit/eefc28c1387309e881e06d0e2ea46c57031d7a5a
20161004 00:18:16< irker795> wesnoth: mattsc wesnoth:master 5e005007bd55 / src/ai/default/ (ca_move_to_targets.cpp ca_move_to_targets.hpp): Merge pull request #807 from wesnoth/ai-improvements-guardians https://github.com/wesnoth/wesnoth/commit/5e005007bd5520a7c4fc00324ec8865876950139
20161004 00:24:40-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 265 seconds]
20161004 00:25:22< tad_> shadowm, It's nothing more than an opinion but I poked around the SDL wiki. SDL_Rect refers to SDL_BlitRect(?) and the remarks there are that the blit function verifies the rect .. which makes me think it's up to the programmer to see a rect is degenerate (0 high or 0 wide) and 'do the right thing' for the specific case.
20161004 00:28:55-!- gfgtdf [~chatzilla@x4e36af65.dyn.telefonica.de] has joined #wesnoth-dev
20161004 00:29:43< shadowm> You mean SDL_BlitSurface()?
20161004 00:30:35 * tad_ nods. Read it and forgot the name in the time it took to typo it :P
20161004 00:31:28< shadowm> It doesn't really say anything specific either, just "performs rectangle validation and clipping".
20161004 00:33:14< shadowm> In particular it doesn't specify anything about how the source rectangle is handled if it's weird.
20161004 00:34:32< shadowm> Still, in 1.12 the code in question uses SDL_SomethingSomethingRect with SDL 1.2 and apparently it doesn't paint 0 px-wide rectangles as if they were 1 px-wide.
20161004 00:34:35< tad_> shadowm, I am looking for the source for SDL_BlitSurface() to see what it does.
20161004 00:35:11< shadowm> That's not what concerns me here, this is SDL_RenderFillRect() not SDL_BlitSurface(
20161004 00:35:14< shadowm> ).
20161004 00:35:30< tad_> shadowm, I would say 'filling' a 0-wide rect should do nothing .. 'stroking' should give a line the width of the pen
20161004 00:36:11< tad_> shadowm, I'm thinking fill rect eventually goes to blit surface so it may punt the issue ...
20161004 00:36:42< shadowm> Generating a source surface for a rectangle fill operation seems like a tremendous waste.
20161004 00:37:25< shadowm> So, hopefully not.
20161004 00:39:42-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 00:41:39< shadowm> So I guess this render stuff uses a hardware-accelerated backend?
20161004 00:43:31-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 265 seconds]
20161004 00:43:56< celticminstrel> Not quite, IIUC.
20161004 00:43:59-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has quit [Ping timeout: 244 seconds]
20161004 00:44:34< celticminstrel> The renderer is set up to be hardware-accelerated, but vultraz used SDL_CreateSoftwareRenderer, which I take to mean it's not actually being accelerated here.
20161004 00:44:49< celticminstrel> ...or was it CreateSurfaceRenderer?
20161004 00:45:10< celticminstrel> It's entirely possible that I'm wrong though.
20161004 00:45:33< shadowm> Regardless, I guess SDL_RenderFillRect doesn't use SDL_FillRect.
20161004 00:45:58< shadowm> It seems far more abstract. *goes back to messing with GUI2 instead*
20161004 00:48:17< tad_> Arg. Not finding a precise answer but I wonder what's up .. the code I'm looking at does a divide-by-zero if you give it a 0-wide or 0-high rectange.  **shudders
20161004 00:50:03< shadowm> Yes, SDL isn't precisely designed to prevent you from shooting yourself in the foot. There are libraries built on top of it that do that.
20161004 00:56:39< gfgtdf> vultraz: i think https://github.com/wesnoth/wesnoth/commit/de8a4270a4aee6cd405d5a9abf4a5034c96e451e is the bad commit for http://gna.org/bugs/?25144: sicne we might reload the game config after mp configure/create eras_[era_index]->cfg mighrt now point to and old and already deleted cfg object from the old game_config.
20161004 00:57:45< gfgtdf> vultraz: cannot build wesnoth currently so i cannot test that but form looking at the code this assumption makes sense.
20161004 00:58:25< gfgtdf> matthiaskrgr: coudl you test whther http://gna.org/bugs/?25144 disappears hen you revert https://github.com/wesnoth/wesnoth/commit/de8a4270a4aee6cd405d5a9abf4a5034c96e451e ?
20161004 01:01:04< shadowm> Annoyingly, the GUI1 file chooser would allow using the character keys for entering filenames and arrows for moving through the directory view, but I can't do that in GUI2 because only one widget can have the focus at a time.
20161004 01:01:25< shadowm> Maybe I should start the file dialog focused on the input box instead of the list view...
20161004 01:04:55< gfgtdf> shadowm: hmm yes afaik we ahve thi problem on other places too (mp lobby, addons list)
20161004 01:04:58< celticminstrel> There is a way for both the listbox and text field to receive input.
20161004 01:05:08< celticminstrel> The debug unit create uses this, I believe.
20161004 01:05:11< celticminstrel> Possibly others.
20161004 01:09:19< shadowm> Huh, yes, it does.
20161004 01:09:51< shadowm> gui2::twindow::add_to_keyboard_chain(), huh.
20161004 01:12:18< shadowm> celticminstrel: Yes, that works. Thanks.
20161004 01:15:22< shadowm> It seems like the text selection isn't highlighted if the textbox doesn't have the focus... hm.
20161004 01:16:14< shadowm> Or maybe I'm misusing all these methods I just made public. Probably just that.
20161004 01:22:06-!- mkdroid [~null@unaffiliated/matthiaskrgr] has quit [Quit: I'll be back!]
20161004 01:28:16-!- mkdr0id [~null@unaffiliated/matthiaskrgr] has quit [Remote host closed the connection]
20161004 01:41:47< matthiaskrgr> 03:25 < gfgtdf> [02:58:25] matthiaskrgr: coudl you test whther http://gna.org/bugs/?25144 disappears hen you revert https://github.com/wesnoth/wesnoth/commit/de8a4270a4aee6cd405d5a9abf4a5034c96e451e ?
20161004 01:41:53< matthiaskrgr> gfgtdf: yes, that seems to fix it 
20161004 01:43:43< matthiaskrgr> gn
20161004 01:43:53-!- gfgtdf_ [~chatzilla@x4e369d23.dyn.telefonica.de] has joined #wesnoth-dev
20161004 01:46:50-!- gfgtdf [~chatzilla@x4e36af65.dyn.telefonica.de] has quit [Ping timeout: 252 seconds]
20161004 01:47:03-!- gfgtdf_ is now known as gfgtdf
20161004 01:59:32-!- vultraz_iOS [uid24821@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161004 02:18:26-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161004 02:31:22< gfgtdf> vultraz: you saw my comment earlier about http://gna.org/bugs/?25144?
20161004 02:43:26< vultraz> what about it?
20161004 02:44:10< vultraz> hm, actually, I think I know what the problem likely is
20161004 02:44:13-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161004 02:45:34< vultraz> gfgtdf: it's probably de8a4270a4aee6cd405d5a9abf4a5034c96e451e
20161004 02:45:55< gfgtdf> vultraz: yes thats what i said
20161004 02:46:18< gfgtdf> gfgtdf vultraz: i think https://github.com/wesnoth/wesnoth/commit/de8a4270a4aee6cd405d5a9abf4a5034c96e451e is the bad commit for http://gna.org/bugs/?25144: sicne we might reload the game config after mp configure/create eras_[era_index]->cfg mighrt now point to and old and already deleted cfg object from the old game_config.
20161004 02:46:42< vultraz> what do you propose as a fix?
20161004 02:47:04< vultraz> it seems to me the cfg should always be valid
20161004 02:47:54< gfgtdf> vultraz: calling get_parameters before relaing the game config might be teh esiest solutoj, not sur though
20161004 02:48:35< gfgtdf> vultraz: also did you see my comment in https://gna.org/bugs/?25142 about ot shiing teh chatbox in m_staging for local multiplayer games?
20161004 02:49:30< vultraz> hm
20161004 02:49:36< vultraz> ill deal with that
20161004 02:49:53< vultraz> i could just make the chatbox take a pointer
20161004 02:50:04< vultraz> idk if i want to hide the chatbox for local games
20161004 02:50:06< vultraz> ill see how it looks
20161004 02:51:16< gfgtdf> ok have to go now.
20161004 02:52:47-!- gfgtdf [~chatzilla@x4e369d23.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 49.0.1/20160922113459]]
20161004 03:18:32-!- irker795 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20161004 03:18:39-!- VultCave [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20161004 03:18:51-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20161004 03:19:03-!- VultCave is now known as vultraz
20161004 03:20:19-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20161004 03:20:19-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161004 03:26:11-!- tad_ [~tadcarluc@173.217.65.103] has quit [Quit: Leaving]
20161004 04:00:07< shadowm> vultraz: Any opinions on what I was talking about earlier on the textbox highlight and SDL_RenderFillRect?
20161004 04:01:45< vultraz> shadowm: I did not see that
20161004 04:02:15< vultraz> let me pull up the logs
20161004 04:03:48< vultraz> 20161004 01:15:22< shadowm> It seems like the text selection isn't highlighted if the textbox doesn't have the focus... hm.
20161004 04:03:57< vultraz> gui2 has no visible text selection i think
20161004 04:04:29< vultraz> 20161003 23:27:08< shadowm> Aginor: Do you know if it's correct for SDL_RenderFillRect() to paint a 1 pixel wide rectangle if width is zero?
20161004 04:04:36< vultraz> I'd say no
20161004 04:04:41< vultraz> this didn't work
20161004 04:04:57< vultraz> as such, i added an explicit check for width and made the code use SDL_RenderDrawLine instead
20161004 04:05:08< vultraz> s/width/thickness
20161004 04:05:37< vultraz> shadowm: is that what you wanted to know?
20161004 04:06:02-!- JyrkiVesterinen [~JyrkiVest@87-100-201-154.bb.dnainternet.fi] has joined #wesnoth-dev
20161004 04:06:18< vultraz> oh I see you restored some stuff
20161004 04:07:11< vultraz> shadowm: oh, I see
20161004 04:07:33< vultraz> i didn't realize i had screwed up
20161004 04:07:45< vultraz> i honestly though gui2 textboxes had no visible selection
20161004 04:08:33< vultraz> what's with that 1px line, though
20161004 04:09:10< celticminstrel> vultraz: Basically, the canvas needs to skip drawing rects that have 0 width or height due to what appears to be a minor bug in SDL2.
20161004 04:09:21< celticminstrel> I'm sure you can fix this easily.
20161004 04:09:38-!- vultraz_iOS [uid24821@wesnoth/developer/vultraz] has quit [Quit: Connection closed for inactivity]
20161004 04:10:26< vultraz> so it seems the background stuff i removed in https://github.com/wesnoth/wesnoth/commit/67cf95a5af3caefb3123da806126282997c1121d
20161004 04:10:32< vultraz> has a minor effect on the text...
20161004 04:10:39< vultraz> i guess that issue wasn't resolved
20161004 04:10:54< shadowm> celticminstrel:  I have the "fix" in my working copy since hours ago.
20161004 04:11:16< shadowm> I'm just unsure if SDL's behavior is correct or not.
20161004 04:11:18< celticminstrel> Ah okay.
20161004 04:11:35< vultraz> ok i misunderstood everything you said completely
20161004 04:11:58< celticminstrel> My personal opinion is that SDL's behaviour is not correct. If I wanted to draw a line using FillRect (which is weird in the first place) I'd request a line of width or height 1.
20161004 04:17:19-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 272 seconds]
20161004 04:26:18< vultraz> shadowm: for the record, no my changes to the gui2 canvas do not make it hw accelerated, but it *is* faster than the old method.
20161004 04:27:32< vultraz> shadowm: also note that even though i don't think it's an issue since you don't have more than one text box in your dialog, but you should use keyboard_capture on the textbox and add_to_keyboard_chain on the list
20161004 04:27:44< vultraz> s/but//
20161004 04:28:34-!- ancestral [~ancestral@75-168-189-115.mpls.qwest.net] has joined #wesnoth-dev
20161004 04:28:35< vultraz> shadowm: or else the text box may capture input from other text boxes/text input widgets
20161004 04:35:58< shadowm> I copied off the unit create dialog.
20161004 04:37:24< vultraz> good, that's correct.
20161004 04:43:57-!- irker141 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20161004 04:43:57< irker141> wesnoth: Ignacio R. Morelle wesnoth:master 7fa70375e1a0 / src/gui/core/canvas.cpp: gui2/canvas: Fix zero-width rectangles being painted 1 px-wide https://github.com/wesnoth/wesnoth/commit/7fa70375e1a00c1ada1cbf782bc52b812349a0d0
20161004 04:51:15-!- JyrkiVesterinen [~JyrkiVest@87-100-201-154.bb.dnainternet.fi] has quit [Quit: .]
20161004 05:06:45-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 05:11:18-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 264 seconds]
20161004 05:42:39< shadowm> :\ It doesn't seem like the text box highlight box's presence depends on the focused state.
20161004 05:45:08< vultraz> no, it's in every state
20161004 05:59:31-!- nore [~ncourant@sas.eleves.ens.fr] has quit [Quit: WeeChat 1.4]
20161004 06:14:16< shadowm> UGH. WHY.
20161004 06:14:33< shadowm> The selection span is correctly set internally, it just doesn't get displayed.
20161004 06:21:34< vultraz> shadowm: what do you mean?
20161004 06:21:53< shadowm> Okay, update_canvas() isn't getting called at the right time.
20161004 06:22:36< shadowm> And this is why I should make my own public method instead of making these internals public.
20161004 06:23:46< vultraz> what are you doing
20161004 06:24:50-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20161004 06:25:26< shadowm> I'm trying to make the textbox selecty thingy that the GUI1 filechooser does for suggesting file extensions on save.
20161004 06:26:02< celticminstrel> Is this like eg Firefox's autocomplete in the URL bar?
20161004 06:26:19< shadowm> ... I guess?
20161004 06:26:21< vultraz> why just firefox
20161004 06:26:36< shadowm> Because celmin knows me better than you evidently. <3
20161004 06:26:46< celticminstrel> Where you start typing and some text is added to the end of the textbox already selected so that typing further replaces it (but then it's added again).
20161004 06:27:02< shadowm> (If he'd said anything else I'd just have said I have no idea because I only really use Firefox.)
20161004 06:27:03< celticminstrel> Firefox because it's the first thing I thought of that did what I thought shadowm seems to be implying he's doing.
20161004 06:27:20 * celticminstrel could have said that with a lot less words.
20161004 06:27:25< shadowm> Well, sort of, but it's more primitive.
20161004 06:27:54< shadowm> The GUI1 filechooser can be made to stick the ".map" extension, for example, whenever the textbox is cleared (e.g. after navigating to a new directory).
20161004 06:28:26< shadowm> Or it can select the name portion of a filename after selecting an entry in the listview, leaving the extension unselected, so that when you press a key only the name portion is replaced.
20161004 06:29:22< shadowm> So I'm just struggling exposing the required functionality in gui2::ttext_box and gui2::ttext_ to the world so that it works.
20161004 06:30:27< shadowm> Apparently the existing methods are protected and not public for a reason, and the reason is that while they mark the widget as dirty, they do not update the canvas variables, therefore the new selection offset and length aren't ever applied when drawing the highlight rectangle.
20161004 06:30:51< shadowm> So I'll just make me my own wrapper around those that properly takes care of updating those variables.
20161004 06:31:58< vultraz> why not make them update the canvas?
20161004 06:32:08< shadowm> Because they probably don't do that for a reason.
20161004 06:32:29< shadowm> They _are_ protected, as I said.
20161004 06:33:00< shadowm> That usually implies that the callers know all the nitty-gritty and don't need extra hand-holding for their task.
20161004 06:35:53< vultraz> I think it's more one isn't supposed to use ttext_ directly
20161004 06:36:16< shadowm> Yes, that too.
20161004 06:36:36< shadowm> Although that doesn't explain why it has public methods and a public ctor anyway.
20161004 06:37:27< vultraz> because it allows for stuff like textbox.text()
20161004 06:37:28< shadowm> Okay, the public methods can probably be attributed to providing the base public API for text widgets. The public ctor, though?
20161004 06:37:38< vultraz> which should not be in the text box widget, but in ttext_
20161004 06:37:39< celticminstrel> I was about to say that, yeah. :P
20161004 06:37:46< vultraz> Blah
20161004 06:37:47< celticminstrel> Maybe the constructor should be made protected.
20161004 06:37:52-!- atarocch [~atarocch@ip-145-193.sn3.clouditalia.com] has joined #wesnoth-dev
20161004 06:37:58 * celticminstrel by "that" was referring to shadowm not vultraz 
20161004 06:37:59< shadowm> ttext_ incidentally seems to contain part of the logic for a multiline textbox.
20161004 06:38:05< vultraz> why do you people say this stuff so much more eloquently 
20161004 06:38:09< celticminstrel> Oh fun!
20161004 06:38:41< shadowm> So e.g. it's aware of "lines" being a concept that can happen.
20161004 06:39:28< vultraz> it seems in this case moving certain functions into the public interface isn't that bad
20161004 06:39:44< celticminstrel> They're already in the public interface, silly. :P
20161004 06:39:53< celticminstrel> The ones he was talking about, at least.
20161004 06:40:05< celticminstrel> When he said "public methods" in ttext_.
20161004 06:40:23 * celticminstrel suspects the trailing underscore on class names is intended to convey "abstract base class".
20161004 06:40:36< vultraz> yes
20161004 06:40:42< vultraz> like tselectable_
20161004 06:40:59< vultraz> though in this case there's also ttext which makes it confusing :|
20161004 06:41:59< shadowm> Well, if you pay attention to the underscore, or aren't in the gui2 namespace or using it into yours, it's pretty obvious which is which.
20161004 06:42:42< shadowm> But yes, a naming convetion like text_widget_base or abstract_text_widget would be preferable.
20161004 06:43:13< celticminstrel> I was planning to mass-rename GUI2 classes after the 1.13.6 release.
20161004 06:43:18< celticminstrel> So I'll keep that in mind.
20161004 06:43:32< shadowm> Are you planning to drop the t's? :p
20161004 06:43:38< celticminstrel> Exactly.
20161004 06:44:00< celticminstrel> Kinda still want a common prefix for widgets, and for dialogs, though.
20161004 06:44:08< celticminstrel> Something explanatory.
20161004 06:44:19< celticminstrel> Might not bother. I dunno.
20161004 06:49:08-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20161004 06:54:59-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 06:56:10-!- atarocch [~atarocch@ip-145-193.sn3.clouditalia.com] has quit [Remote host closed the connection]
20161004 06:59:14-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 252 seconds]
20161004 07:07:49-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20161004 07:11:53< shadowm> sizeof(unsigned short) <= sizeof(int), right?
20161004 07:12:12< celticminstrel> Yes.
20161004 07:14:21-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161004 07:15:00< vultraz> celticminstrel: why not just widget_, widget_base, and dialog_
20161004 07:15:15< shadowm> How about namespaces?
20161004 07:15:20< celticminstrel> It's a bit long, but I suppose I could do that. (Not the middle one though.)
20161004 07:15:27< shadowm> gui2::widget::foo, gui2::dialog::foo.
20161004 07:15:30< celticminstrel> I dunno about namespaces. I suppose it's an option.
20161004 07:15:45< shadowm> (gui2::tdialog would become gui2::dialog::base or something.)
20161004 07:15:59< celticminstrel> Namespaces bug me a bit. Maybe because XCode insists on indenting them no matter what.
20161004 07:16:36< celticminstrel> I think maybe tdialog under that schema could become something like gui2::dialog::modal.
20161004 07:16:47< shadowm> I'm not a fan of using the :: operator very often.
20161004 07:16:49< celticminstrel> And popup would become gui2::dialog::modeless.
20161004 07:16:59< celticminstrel> And tip would be gui2::dialog::tooltip or something.
20161004 07:17:41< shadowm> So in other words I very rarely use nested namespaces.
20161004 07:17:54-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 264 seconds]
20161004 07:17:54-!- wedge010 is now known as wedge009
20161004 07:17:55< shadowm> Or nested classes etc.
20161004 07:18:06< celticminstrel> Somehow I missed the "not" the first time I read that.
20161004 07:18:44< shadowm> The fact that in C++03 (and IIRC C++11 too?) you can't define nested namespaces in one go probably doesn't help.
20161004 07:19:15< shadowm> (i.e. `namespace foo::bar {}` instead of `namespace foo { namespace bar {} }`)
20161004 07:20:08< shadowm> Oh boy. `length = -start`.
20161004 07:20:17< shadowm> length is an int, start is a size_t.
20161004 07:21:42< JyrkiVesterinen> I had a quick look. Nested namespace definition is coming in C++17. :(
20161004 07:22:08< vultraz> was just checking that
20161004 07:22:11< vultraz> what JyrkiVesterinen said
20161004 07:22:18< shadowm> How would I go about truncating start to fit int without feeling bad about my life decisions?
20161004 07:22:42< shadowm> std::min(start, INT_MAX) or so?
20161004 07:23:01< shadowm> Which is from the C library's limits.h IIRC.
20161004 07:23:16< celticminstrel> Uhhh...
20161004 07:23:23< vultraz> why does start need to be size_t?
20161004 07:23:30< celticminstrel> Is int(start) not sufficient?
20161004 07:23:50< shadowm> Because ttext_ already assumes the selection offset is a size_t and the length a signed int.
20161004 07:24:06< shadowm> I mean, it's designed like that. It doesn't assume. It just is that way.
20161004 07:24:23-!- Duthlet [~Duthlet@dslb-188-104-253-155.188.104.pools.vodafone-ip.de] has joined #wesnoth-dev
20161004 07:24:27< shadowm> celticminstrel: What's int(size_t(-1))?
20161004 07:24:46< shadowm> I'm guessing it's not -1.
20161004 07:24:50< vultraz> I don't particularly see the logic in that design
20161004 07:25:08< shadowm> vultraz: The length carries the selection direction.
20161004 07:25:12< celticminstrel> Uhh. I guess size_t(-1) is something like 0xffffffff.
20161004 07:25:34< shadowm> Yes, size_t(-1) is certainly whatever the largest value for size_t is.
20161004 07:25:39< celticminstrel> So that would be truncated to 0xffff.
20161004 07:25:44< shadowm> And the largest value for size_t is greater than the largest value for int.
20161004 07:25:46< celticminstrel> Which when cast to int would be -1.
20161004 07:25:48< celticminstrel> I think.
20161004 07:25:52< shadowm> Potentially, I mean. Usually is.
20161004 07:26:00 * celticminstrel using specific values as an example only.
20161004 07:26:10-!- boucman_work [~boucman@gre92-5-82-237-199-7.fbx.proxad.net] has joined #wesnoth-dev
20161004 07:26:20 * celticminstrel suspects that the examples aren't actually the correct ones for int and size_t.
20161004 07:26:38 * vultraz really doesn't enjoy this stuff
20161004 07:26:51< JyrkiVesterinen> The relevant quote from the standard:
20161004 07:26:52< shadowm> I don't think... 0xFFFFU is -1 when cast to int.
20161004 07:26:54< JyrkiVesterinen> "If the destination type is signed, the value is unchanged if it can be represented in the destination type (and
20161004 07:26:54< JyrkiVesterinen> bit-field width); otherwise, the value is implementation-defined."
20161004 07:27:05< JyrkiVesterinen> int(size_t(-1)) is implementation-defined.
20161004 07:27:12< shadowm> Isn't the sign bit the most significant bit on x86?
20161004 07:28:39< shadowm> Okay, so basically there's no transitivity in int(size_t(-1)).
20161004 07:29:07< celticminstrel> Yes it's the most significant bit on x86.
20161004 07:29:08< shadowm> Which is what I suspected. My point being that no, int(start) is clearly not sufficient for sufficiently large values of start.
20161004 07:29:12< JyrkiVesterinen> If you want 0xffffffffu to convert to -1, you aren't really truncating.
20161004 07:29:31< JyrkiVesterinen> How about detecting that value and making it -1, and using static_cast<int> otherwise?
20161004 07:30:15< shadowm> Well, back to the actual usecase then the best solution would be int(std::min<size_t>(start, INT_MAX)).
20161004 07:30:15-!- Nobun [~nobun@host219-23-dynamic.50-82-r.retail.telecomitalia.it] has joined #wesnoth-dev
20161004 07:30:33 * vultraz blinks
20161004 07:30:52< celticminstrel> I generally prefer to use <numeric_limits> over <climit>
20161004 07:31:08< celticminstrel> So instead of INT_MAX it's std::numeric_limits<int>::max().
20161004 07:31:10< JyrkiVesterinen> Are you worried that "start" would exceed two billion in normal operation? A textbox should never have anywhere that many characters.
20161004 07:31:34< shadowm> JyrkiVesterinen: This is Wesnoth we're talking about. Not expecting the unexpected has gotten us in trouble before.
20161004 07:31:35< vultraz> I wonder how many words 2 billion characters is anyway
20161004 07:31:48< celticminstrel> Depends on the words, duh.
20161004 07:32:36< vultraz> let's say... 5 letter words
20161004 07:32:42< shadowm> Also, the real turning point would be 2^16 - 1 characters.
20161004 07:32:58< vultraz> 400,000,000 words
20161004 07:33:11< shadowm> On 32-bit x86 anyway.
20161004 07:33:12< celticminstrel> I assume you divided by 6 to get that.
20161004 07:33:26< celticminstrel> If you divided by 5 you're assuming 4-letter words.
20161004 07:33:45< shadowm> Wait.
20161004 07:33:57< shadowm> Never mind my bullshit numbers. You know how I am with numbers. <_<
20161004 07:34:09< vultraz> celticminstrel: why do you say that
20161004 07:34:18< celticminstrel> Space is a character too.
20161004 07:34:22< vultraz> ahh
20161004 07:34:24< shadowm> 2^32 - 1 is more like it.
20161004 07:34:24< vultraz> right
20161004 07:34:39< vultraz> well in which case it's 333,333,333.333....
20161004 07:34:56< JyrkiVesterinen> 2^31 -1 because int is signed. About two billion as I said.
20161004 07:35:25< shadowm> Anyway, the point is, I don't want to give some future lunatic (or an automated fuzz tester) a hard time.
20161004 07:35:38< shadowm> So might as well future-proof this crap.
20161004 07:35:39< vultraz> ... how does 2^31 = 2 billion.
20161004 07:35:46< JyrkiVesterinen> For reference, here's a handly list of word counts of some epic literature (for the purpose of comparing it with the longest video game scripts): http://gaming.wikia.com/wiki/List_of_longest_video_game_scripts#Comparison
20161004 07:35:56< celticminstrel> vultraz: It doesn't.
20161004 07:35:56< JyrkiVesterinen> "Mahabharata - 1.8 million words"
20161004 07:36:03< shadowm> vultraz: 2,147,483,648. Over 2 billion.
20161004 07:36:04< JyrkiVesterinen> Nowhere near 300 million.
20161004 07:36:16-!- nore [~ncourant@sas.eleves.ens.fr] has joined #wesnoth-dev
20161004 07:36:26< celticminstrel> If I recall correctly, isn't 2^31 actually more like 20 billion? I could be making this up.
20161004 07:36:35< celticminstrel> Ah, yeah, okay.
20161004 07:36:39< celticminstrel> Timing. <_<
20161004 07:36:44< celticminstrel> But yeah.
20161004 07:36:55< celticminstrel> What Jyrki said. Just ignore me.
20161004 07:37:05< shadowm> (Don't get me started on long vs. short scale. It's even worse than metric vs. imperial system because even closely-related countries differ on scales for big numbers.)
20161004 07:37:20< vultraz> no, 2 billion
20161004 07:37:21< vultraz> huh
20161004 07:37:25< vultraz> i'd have thought it'd be more
20161004 07:37:42 * celticminstrel was wondering if it might not actually be better to change it to "unsigned int", though that's a more extensive change.
20161004 07:37:48< celticminstrel> Uhh...
20161004 07:37:50< celticminstrel> ^-not
20161004 07:38:32< shadowm> So yeah, <numeric_limits>, huh.
20161004 07:38:53< shadowm> It feels like I've never heard of this before somehow.
20161004 07:39:31< JyrkiVesterinen> I prefer the C macros anyway. They are shorter to type.
20161004 07:40:04< shadowm> I guess I have to admit I'm somewhat fond of macros for this kind of thing because they have ALL_UPPERCASE_IMPORTANT_LOOKING_NAMES.
20161004 07:40:39< shadowm> But if numeric_limits is preferred here, I can use that.
20161004 07:41:11< JyrkiVesterinen> It's up to you. I think no one would reject either option.
20161004 07:44:23-!- irker141 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20161004 07:47:05< shadowm>     src/gui/widgets/text.cpp:217:26: error: comparison between signed and unsigned integer expressions [-Werror=sign-compare]
20161004 07:47:11< shadowm>       if(backwards && -length > start) {
20161004 07:47:32 * shadowm facekeyboards.
20161004 07:48:25< shadowm> I'm trying to add functionality, not rewrite ttext_.
20161004 07:55:37-!- boucman_work [~boucman@gre92-5-82-237-199-7.fbx.proxad.net] has quit [Ping timeout: 265 seconds]
20161004 07:55:39< shadowm> Eh.
20161004 07:56:00< shadowm> Well... calling update_canvas() at the end didn't do the trick.
20161004 07:56:02-!- Nobun [~nobun@host219-23-dynamic.50-82-r.retail.telecomitalia.it] has quit [Quit: Salve a tutti]
20161004 07:56:21-!- atarocch [~atarocch@5.92.70.169] has joined #wesnoth-dev
20161004 07:56:33< shadowm> Also, um, ttext_box believes the selection start is an unsigned int.
20161004 07:56:43< shadowm> ttext_ stores it as a size_t.
20161004 07:57:16< shadowm>     const unsigned start = get_selection_start();
20161004 07:57:27< shadowm>     size_t get_selection_start() const;
20161004 07:57:40< shadowm> Shruuuuuug.
20161004 07:58:22< JyrkiVesterinen> I guess that whoever wrote that line simply figured that text boxes with over 4 billion characters aren't worth worrying about.
20161004 07:58:58< shadowm> Or wasn't keeping track of his types.
20161004 07:59:13< JyrkiVesterinen> As I demonstrated above, a content author would be able to stuff an entire book, *any* book, in there. And there would still be plenty of integer range left.
20161004 07:59:21< shadowm> Yes, you can keep saying that.
20161004 07:59:27< shadowm> <shadowm> Anyway, the point is, I don't want to give some future lunatic (or an automated fuzz tester) a hard time.
20161004 08:00:10< JyrkiVesterinen> I'm simply not at all surprised that there is already code that truncates the selection start.
20161004 08:01:32< shadowm> I'm bviously calling update_canvas() at the right time but it believes the offset and length are 0?
20161004 08:02:47< shadowm> Did I just waste my time doing all that?
20161004 08:04:22< vultraz> possibly
20161004 08:04:37< shadowm> *I* am setting them to zero.
20161004 08:06:20< shadowm> Bluh.
20161004 08:06:32< shadowm> I botched a copy-paste.
20161004 08:06:45< shadowm> Now it works.
20161004 08:06:51< vultraz> praise be
20161004 08:06:55< shadowm> Exactly as intended.
20161004 08:07:17< shadowm> Except that clicking on list items appears to cancel the event chain thingy.
20161004 08:07:39< vultraz> LE SIGH
20161004 08:08:06< shadowm> You have the same problem in the unit create dialog.
20161004 08:08:32< shadowm> I can probably work around that in my dialog by stuffing the capture & chain sequence into the selection handler.
20161004 08:08:48< vultraz> oh yeah
20161004 08:09:03< vultraz> i disregarded it in other cases
20161004 08:09:09< vultraz> but i guess here it's needed
20161004 08:09:12< shadowm> Hacks.
20161004 08:13:44< shadowm> wesnoth: src/gui/core/event/distributor.cpp:684: void gui2::event::tdistributor::keyboard_add_to_chain(gui2::twidget*): Assertion `std::find(keyboard_focus_chain_.begin(), keyboard_focus_chain_.end(), widget) == keyboard_focus_chain_.end()' failed.
20161004 08:14:01 * shadowm throws keyboard wallwards.
20161004 08:14:46< vultraz> heh
20161004 08:14:49< shadowm> Turns out doing it from the row select event handler makes everything go to hell.
20161004 08:15:17< shadowm> Hm.
20161004 08:15:49< shadowm> I could just focus the textbox and leave the other thingy alone.
20161004 08:16:01< shadowm> Yesssss that works nicely.
20161004 08:16:17< shadowm> So I guess the focus chain thing is permanent.
20161004 08:16:40< vultraz> It is
20161004 08:16:44< shadowm> Oh yes, I've finally achieved feature parity with the GUI1 filechooser.
20161004 08:16:51< Aginor> shadowm: no, I don't know if it's correct behaviour. I suspect that it is expected behaviour because it's silly to try to paint a nonexistant rectangle
20161004 08:16:54< vultraz> I just confirmed this can also be achieved by adding the filter to the keyboard chain
20161004 08:17:11< vultraz> though the box won't retain it's 'focused' status
20161004 08:17:13< shadowm> Now I can start working on improving it UX-wise.
20161004 08:17:17< vultraz> ie, thecursor won't show
20161004 08:17:25< vultraz> I take it you manually set focus?
20161004 08:17:34< vultraz> instead of adding the list to the keyboard chain?
20161004 08:17:39< vultraz> s/list/filter
20161004 08:18:07< shadowm> preshow captures on the textbox and adds the listbox to the chain. The select handler simply recaptures the textbox.
20161004 08:18:34< vultraz> ok 
20161004 08:18:42< vultraz> honestly that's probably a better solution anyway
20161004 08:18:44< shadowm> This means that I can change the listbox selection with either the mouse or the keyboard and the textbox is refocused.
20161004 08:19:02< shadowm> This would look weird if the listbox looked different when focused.
20161004 08:19:19< shadowm> e.g. the listbox would never attain the focused state and that'd look weird to the user.
20161004 08:19:36< vultraz> adding the filter to the keyboard chain as well... could cause problems
20161004 08:19:41< vultraz> plus it doesn't show it as focuses
20161004 08:19:43< vultraz> focused
20161004 08:19:50< vultraz> so your solution is optimal 
20161004 08:20:18< shadowm> As long as listboxes don't have a visually-distinct focused state, anyway.
20161004 08:20:32< vultraz> they shouldn't
20161004 08:20:43< shadowm> They would on every other widget toolkit.
20161004 08:20:50< vultraz> the toggle panel is what has the variable states
20161004 08:21:02< shadowm> Including Windows' native controls
20161004 08:21:33< shadowm> The key distinction being that on those, list items aren't major widgets or able to appear outside lists.
20161004 08:21:43< shadowm> I don't think they are usually widgets at all.
20161004 08:22:45< shadowm> Our toggle panel doesn't have a distinct focused state at the moment either.
20161004 08:22:54< shadowm> (Note: focused, not selected.)
20161004 08:23:01< vultraz> ah, true
20161004 08:23:14< shadowm> Otherwise it'd also look weird and raise eyebrows.
20161004 08:23:19< celticminstrel> It does darken or something on mouseover though...?
20161004 08:23:49< shadowm> That's hovered, not focused.
20161004 08:23:50< vultraz> oh yeah
20161004 08:23:58< vultraz> no, that's focused
20161004 08:24:04< vultraz> according to the wml
20161004 08:24:28< shadowm> I don't care about the WML, the WML is all wrong. :p
20161004 08:24:38< shadowm> (See also: "border" instead of "margin".)
20161004 08:24:51< shadowm> For the purpose of this conversation, when I say focused I mean "has the keyboard focus".
20161004 08:25:02< vultraz> oh
20161004 08:25:07< shadowm> And toggle panels don't really ever have the keyboard focus AFAIK.
20161004 08:25:42< shadowm> Anyway, I guess I've triumphed over GUI2 once again.
20161004 08:26:24< shadowm> And I managed to keep all the changes in ttext_ so that this will purportedly automagically work with a multiline textbox or the password box too.
20161004 08:26:43< shadowm> \o/
20161004 08:26:58< vultraz> Yes, and it took a feature I added to do so (triumph, that is) :P
20161004 08:27:19< vultraz> Sometimes, into the gui2 internals, you must go.
20161004 08:30:28< vultraz> btw, could you perhaps remind me of the things I need to improve with the scrollbars
20161004 08:31:05-!- ancestral [~ancestral@75-168-189-115.mpls.qwest.net] has quit [Ping timeout: 248 seconds]
20161004 08:32:30-!- celticminstrel is now known as celmin|sleep
20161004 08:33:13-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20161004 08:37:51-!- irker380 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20161004 08:37:51< irker380> wesnoth: Charles Dang wesnoth:master a67809d24099 / data/gui/window/tooltip_large.cfg: Slightly bump initial tooltip width https://github.com/wesnoth/wesnoth/commit/a67809d240996fa297253a9313be8fe6fc2dc982
20161004 08:37:52< irker380> wesnoth: Charles Dang wesnoth:master e25aa4ac2ce0 / data/gui/window/mp_create_game.cfg src/gui/dialogs/multiplayer/mp_create_game.cpp: MP Create Game: use tooltips for mod descriptions https://github.com/wesnoth/wesnoth/commit/e25aa4ac2ce03639527a6029778c29783a428ad1
20161004 08:39:45< vultraz> i need to do something about the text in the tooltips
20161004 08:39:51< vultraz> it's slightly hard to look at for some reason
20161004 08:43:14-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 08:47:49-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 265 seconds]
20161004 08:49:07-!- benladin [d2564fc3@gateway/web/freenode/ip.210.86.79.195] has joined #wesnoth-dev
20161004 08:49:35< benladin> wow so many people here
20161004 08:56:00-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20161004 08:57:06< JyrkiVesterinen> benladin: The number of people is high, but only a handful discusses actively.
20161004 09:07:35-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20161004 09:10:02< benladin> lurker central
20161004 09:10:29< benladin> well ill leave it on for a few days and maybe somethign interesting comes up
20161004 09:11:44< Aginor> what counts as interesting?
20161004 09:13:57-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20161004 09:21:57< benladin> idk some ideas for wesnoth or discussion of development
20161004 09:24:55< JyrkiVesterinen> Such discussion does come up daily.
20161004 09:25:09< JyrkiVesterinen> It's quiet now, mainly because it's night in America.
20161004 09:25:55-!- Bonobo [~Bonobo@2001:44b8:254:3200:4d09:6c3e:552:bc78] has joined #wesnoth-dev
20161004 09:27:10-!- boucman_work [~boucman@fw-alt.idf.smile.fr] has joined #wesnoth-dev
20161004 09:28:25< benladin> so americans work on wesnoth but hardly play? interesting
20161004 09:29:51< Aginor> I think the majority of the active contributors are non-american actually
20161004 09:32:25-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 248 seconds]
20161004 09:36:32-!- Bonobo [~Bonobo@2001:44b8:254:3200:4d09:6c3e:552:bc78] has quit [Ping timeout: 252 seconds]
20161004 09:37:27-!- Bonobo [~Bonobo@2001:44b8:254:3200:4d09:6c3e:552:bc78] has joined #wesnoth-dev
20161004 09:39:27-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20161004 09:42:14< benladin> oh so its just quiet now because americans are the loudest :P
20161004 09:44:37< JyrkiVesterinen> It's quite correct. Two Americal developers (IIRC, one from USA, the other from Canada) are by far the most active developers in the project.
20161004 09:45:43< wedge009> I see conversation throughout the day. Most of it seems to happen when I'm at work or asleep...
20161004 09:45:46-!- stikonas [~stikonas@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 09:46:41< Aginor> that is very true
20161004 09:46:52< wedge009> You can have a look through the logs if you really want to see how animated it can get in here.
20161004 09:46:54< shadowm> Which one is from the USA?
20161004 09:47:10< JyrkiVesterinen> Vultraz?
20161004 09:47:14< wedge009> Isn't Vultraz from the States, even if he's in Micronesia?
20161004 09:47:26< shadowm> vultraz lives in Micronesia, yes.
20161004 09:47:37< shadowm> His timezone is in the future as well.
20161004 09:48:03< shadowm> So I wouldn't count him as American for the purpose of timezones.
20161004 09:48:10< wedge009> Your future maybe.
20161004 09:48:19< benladin> why is rail defense 0%
20161004 09:48:28< wedge009> Uh oh.
20161004 09:48:28< benladin> rails cant be that hard to defend from
20161004 09:48:37< shadowm> Well, Australians would consider his to be the recent past I think?
20161004 09:49:00< wedge009> benladin: https://gna.org/bugs/?22352
20161004 09:49:09< shadowm> It's UTC+a big number.
20161004 09:49:16< wedge009> shadowm: Actually, same as Australian Eastern Time.
20161004 09:49:29< wedge009> Which is where I am. Aginor is two hours ahead of me.
20161004 09:49:50< shadowm> vultraz: I'd tell you about the scrollbars but I doubt you'll read this without me polling you so... might as well wait.
20161004 09:50:19< Aginor> yes, you're all living in the past
20161004 09:50:58< wedge009> :p
20161004 09:53:04< benladin> well I'm in new zealand so actually you're all in my past
20161004 09:54:09< Aginor> impressive
20161004 09:54:28< shadowm> Every time I ask people from big timezones what the future is like they tell me it's actually pretty boring, so I'm not all that bothered about living in the past. :p
20161004 09:54:56< Aginor> shadowm: it will rain
20161004 09:55:08< shadowm> There, perhaps.
20161004 09:55:23< shadowm> I strongly doubt there'll be any substantial rain until the summer is over here.
20161004 09:55:29< Aginor> sooner or later, in your future, there will be rain
20161004 09:55:34 * Aginor has predicted
20161004 09:55:44< shadowm> i.e. at some point in June 2017.
20161004 09:56:03< Aginor> when the rain strikes, may you remember this day
20161004 09:56:09< benladin> really going out on a limb there aginor
20161004 09:56:19< Aginor> I'm feeling pretty limber
20161004 09:56:29< benladin> wedge009: i feel the rails issue is still unresolved completely
20161004 09:56:36< wedge009> I know.
20161004 09:56:43< wedge009> Nothing's changed since that report.
20161004 09:57:04< benladin> while the resolution is clearly just assign defence values to rails
20161004 09:57:37< wedge009> That's up to the people maintaining the units. Which no-one seems to want to do at this point.
20161004 09:58:32< wedge009> There's also some discussion in my PR: https://github.com/wesnoth/wesnoth/pull/536
20161004 10:01:06< zookeeper> we don't want them to have any defense value, the defense value of the underlying terrain should always be used.
20161004 10:02:03< zookeeper> and it's better that it shows up as 0% instead of, for example, the same as what the unit has for flat, because the latter would be more confusing since it'd imply that you actually receive that defense on rails, which you wouldn't.
20161004 10:03:47< zookeeper> the rails shouldn't really have even been added in the first place
20161004 10:04:11< shadowm> :(
20161004 10:04:56< wedge009> It does seem like a special case... which has resulted in this special case work-around.
20161004 10:05:06< zookeeper> shadowm, what, someone actually has a unit utilizing them?
20161004 10:05:31< zookeeper> ok, 2 add-ons in 1.12 seem to have a minecart unit
20161004 10:12:57-!- RatArmy [~RatArmy@om126237124204.9.openmobile.ne.jp] has joined #wesnoth-dev
20161004 10:16:12< benladin> mmm seems like they should either be removed as a terrain type or given a defense value
20161004 10:20:07-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 10:21:01-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161004 10:31:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 10:34:23-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20161004 10:35:53-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 248 seconds]
20161004 10:42:33-!- stikonas [~stikonas@wesnoth/translator/stikonas] has quit [Quit: AtomicIRC: The nuclear option.]
20161004 10:42:46-!- stikonas_ is now known as stikonas
20161004 10:43:33-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has joined #wesnoth-dev
20161004 10:46:01-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161004 10:47:01-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 10:47:16-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20161004 10:48:23-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 10:48:44-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20161004 10:49:26-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 11:01:40-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161004 11:03:44-!- RatArmy [~RatArmy@om126237124204.9.openmobile.ne.jp] has quit [Quit: Leaving]
20161004 11:09:01< benladin> having some issues there stikonas
20161004 11:10:31-!- esr [~esr@static-71-162-243-5.phlapa.fios.verizon.net] has quit [Quit: WeeChat 1.4]
20161004 11:10:41-!- esr [~esr@wesnoth/developer/esr] has joined #wesnoth-dev
20161004 11:20:28-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20161004 11:22:51-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 11:23:49-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20161004 11:24:42-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 11:29:35< zookeeper> wedge009, possible solution: add a [terrain_type] hide_if_impassable=yes|no(default no), with which rails archetype can be made to show up only if the unit actually makes use of it.
20161004 11:31:33< zookeeper> could also be used for the impassable archetype, because duh, impassable :p
20161004 11:38:06-!- irker380 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20161004 11:41:55-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161004 11:42:13-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 12:12:12-!- Bonobo [~Bonobo@2001:44b8:254:3200:4d09:6c3e:552:bc78] has quit [Read error: Connection reset by peer]
20161004 12:12:32-!- Bonobo [~Bonobo@2001:44b8:254:3200:4d09:6c3e:552:bc78] has joined #wesnoth-dev
20161004 12:14:05-!- Bonobo [~Bonobo@2001:44b8:254:3200:4d09:6c3e:552:bc78] has quit [Client Quit]
20161004 12:18:51-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20161004 12:19:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 12:24:30-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 264 seconds]
20161004 12:27:51-!- mkdroid [~null@unaffiliated/matthiaskrgr] has joined #wesnoth-dev
20161004 12:27:59-!- Nobun [~nobun@host219-23-dynamic.50-82-r.retail.telecomitalia.it] has joined #wesnoth-dev
20161004 12:42:43-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 272 seconds]
20161004 12:43:06-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 12:53:21-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20161004 12:58:47-!- gfgtdf [~chatzilla@x4e369d23.dyn.telefonica.de] has joined #wesnoth-dev
20161004 13:00:11-!- Nobun [~nobun@host219-23-dynamic.50-82-r.retail.telecomitalia.it] has quit [Quit: Salve a tutti]
20161004 13:02:33-!- atarocch [~atarocch@5.92.70.169] has quit [Ping timeout: 248 seconds]
20161004 13:02:59-!- atarocch [~atarocch@5.92.70.169] has joined #wesnoth-dev
20161004 13:03:48-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20161004 13:04:28-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20161004 13:06:09-!- mkdroid [~null@unaffiliated/matthiaskrgr] has quit [Ping timeout: 272 seconds]
20161004 13:10:29-!- Appleman1234 [~Appleman1@KD106154000111.au-net.ne.jp] has quit [Ping timeout: 252 seconds]
20161004 13:13:48-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20161004 13:23:04-!- mkdroid [~null@unaffiliated/matthiaskrgr] has joined #wesnoth-dev
20161004 13:44:00-!- fabi_ [~fabi@176.7.22.57] has joined #wesnoth-dev
20161004 13:46:42-!- JyrkiVesterinen [~JyrkiVest@nblzone-242-23.nblnetworks.fi] has quit [Quit: .]
20161004 13:47:18-!- fabi [~fabi@176.5.20.202] has quit [Ping timeout: 264 seconds]
20161004 13:50:33-!- celmin|sleep is now known as celticminstrel
20161004 14:02:38-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev
20161004 14:05:19-!- mkdroid [~null@unaffiliated/matthiaskrgr] has quit [Remote host closed the connection]
20161004 14:06:54-!- Appleman1234 [~Appleman1@KD106154011107.au-net.ne.jp] has joined #wesnoth-dev
20161004 14:14:57-!- tad_ [~Gregory@173.217.65.103] has joined #wesnoth-dev
20161004 14:15:39-!- tad_ [~Gregory@173.217.65.103] has quit [Client Quit]
20161004 14:18:43-!- tad_ [~Gregory@173.217.65.103] has joined #wesnoth-dev
20161004 14:20:24-!- tad_carlucci [~Gregory@173.217.65.103] has joined #wesnoth-dev
20161004 14:20:30-!- tad_carlucci [~Gregory@173.217.65.103] has quit [Client Quit]
20161004 14:20:30-!- tad_ [~Gregory@173.217.65.103] has quit [Client Quit]
20161004 14:21:59-!- tad_ [~Gregory@173.217.65.103] has joined #wesnoth-dev
20161004 14:30:37-!- tad_ [~Gregory@173.217.65.103] has quit [Quit: Leaving]
20161004 14:30:59-!- tad_ [~Gregory@173.217.65.103] has joined #wesnoth-dev
20161004 14:38:01< tad_> This should be "interesting" if I can avoid the rabbit holes. I'm trying to refactor my build network so I can build native Linux, native Windows, Windows cross from Linux, and Linux cross from Windows System for Linux; and share the repo across all four hosts.
20161004 14:41:42-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20161004 14:44:33< benladin> what does that even mean
20161004 14:48:03< tad_> I usually live on Linux and do all my work there. But every now any then we have to deal with Windows-isms and I want to be able to jump to Windows for that. But whether I'm on Linux or suffering under Windows, I want to be able to compile for the other system without having to jump back and forth.
20161004 14:49:42-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 264 seconds]
20161004 15:03:27-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20161004 15:06:44< benladin> but windows is so much better than linux
20161004 15:08:36< tad_> I would say that's only true if you drink the Kool-Aid, but, really, it's just a religious statement.
20161004 15:10:18< benladin> ha im jk
20161004 15:10:46< benladin> actually i prefer apple
20161004 15:10:51< tad_> Hey, at least I'm reigning myself in and not adding Android to the mix (yet).
20161004 15:13:32< benladin> why limit yourself
20161004 15:13:55< tad_> Patience. One rabbit-hole at a time.
20161004 15:15:00< tad_> Well, considering it's Windows .. perhaps it's presently a rat's nest ... :P
20161004 15:18:11< benladin> ha i actually have no opinion, i just take the opposite opinion of whoever im talking to for lols
20161004 15:24:07< tad_> I have this to say about Windows ... at least I can repaint the living room while I'm waiting on it to finish ...
20161004 15:27:42< tad_> Oh, drat, I forgot about that. Microsoft bought Xamarin and I forgot to uncheck the option. Looks like Android development joined the party, anyway.  **sigh**
20161004 15:30:04< benladin> painters across the world love windows
20161004 15:35:24< tad_> Yep, there it is ... "Applying Android Native Development Kit" ... ah well, I can just ignore it until I'm ready for that.
20161004 15:40:31-!- Nobun [~nobun@host219-23-dynamic.50-82-r.retail.telecomitalia.it] has joined #wesnoth-dev
20161004 15:54:50-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 15:54:56-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161004 15:55:02-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 16:05:48-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity]
20161004 16:08:16-!- DeFender1031 [~DeFender1@89-138-252-80.bb.netvision.net.il] has joined #wesnoth-dev
20161004 16:16:53-!- horrowind [~Icedove@2a02:810a:8380:10a8:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20161004 16:19:12-!- benladin [d2564fc3@gateway/web/freenode/ip.210.86.79.195] has quit [Ping timeout: 240 seconds]
20161004 16:35:51-!- mordante [~mordante@2001:984:5786:1:7a24:afff:fe8c:dea8] has joined #wesnoth-dev
20161004 16:35:51-!- mordante [~mordante@2001:984:5786:1:7a24:afff:fe8c:dea8] has quit [Changing host]
20161004 16:35:51-!- mordante [~mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20161004 16:36:05< mordante> servus
20161004 16:47:36-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20161004 16:47:36-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Read error: Connection reset by peer]
20161004 16:49:48< shadowm> celticminstrel, vultraz: Could you give PR #778 a last review before I rewrite the commit series and merge it?
20161004 16:51:20< celticminstrel> Sure...
20161004 17:00:29-!- atarocch [~atarocch@5.92.70.169] has quit [Ping timeout: 272 seconds]
20161004 17:18:33-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20161004 17:24:26-!- atarocch [~atarocch@ip-145-193.sn3.clouditalia.com] has joined #wesnoth-dev
20161004 17:25:02-!- tad_ [~Gregory@173.217.65.103] has quit [Quit: Leaving]
20161004 17:26:40-!- JyrkiVesterinen [~JyrkiVest@87-92-29-112.bb.dnainternet.fi] has joined #wesnoth-dev
20161004 17:39:49-!- ancestral [~ancestral@75-168-189-115.mpls.qwest.net] has joined #wesnoth-dev
20161004 17:40:20-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has joined #wesnoth-dev
20161004 17:50:00-!- ancestral [~ancestral@75-168-189-115.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20161004 17:54:33-!- irker193 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20161004 17:54:33< irker193> wesnoth: Jyrki Vesterinen wesnoth:master f3cdd5b2f06b / projectfiles/VC12/ (wesnoth.vcxproj wesnoth.vcxproj.filters): Update Visual Studio project https://github.com/wesnoth/wesnoth/commit/f3cdd5b2f06b7c9dfd98de8dc3a60407635a3947
20161004 18:01:07-!- mjs-de [~mjs-de@x4db554d7.dyn.telefonica.de] has joined #wesnoth-dev
20161004 18:20:55-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20161004 18:21:01-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20161004 18:42:27< Nobun> loonycyborg: as I wrote in my pull request. I noticed that my comments in some points need a little fix. I will fix code comments within tomorrow or the day after tomorrow.
20161004 18:42:52< loonycyborg> ok
20161004 18:43:41< Nobun> the actual code, however works as described in my first post in the pull request, but feel free to make additional tests in the meantime
20161004 18:47:51-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161004 18:53:23< irker193> wesnoth: Charles Dang wesnoth:master 9b6fc78690cc / src/ (4 files in 2 dirs): Revert "MP Create: basic implementation of saved games as their own category (de https://github.com/wesnoth/wesnoth/commit/9b6fc78690cceb292bf3fe8b5fbd0a3c39f25167
20161004 18:56:44< vultraz> mattsc ^
20161004 18:59:28-!- ancestral [~ancestral@75-168-189-115.mpls.qwest.net] has joined #wesnoth-dev
20161004 19:01:42-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20161004 19:04:01-!- gfgtdf [~chatzilla@x4e369d23.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 49.0.1/20160922113459]]
20161004 19:08:00-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 19:10:45-!- JyrkiVesterinen [~JyrkiVest@87-92-29-112.bb.dnainternet.fi] has quit [Quit: .]
20161004 19:11:08< mordante> shadowm, just did a fresh build and it compiles now
20161004 19:11:22< mordante> shadowm, I see you already removed the SDL_net stuff
20161004 19:14:20-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20161004 19:16:25-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 248 seconds]
20161004 19:16:25-!- wedge010 is now known as wedge009
20161004 19:22:34-!- aidanhs2 [~aidanhs@81.4.110.234] has quit [Quit: WeeChat 0.4.2]
20161004 19:25:21< shadowm> mordante: Okay, good to hear.
20161004 19:40:09-!- mjs-de [~mjs-de@x4db554d7.dyn.telefonica.de] has quit [Remote host closed the connection]
20161004 19:46:28-!- mordante [~mordante@wesnoth/developer/mordante] has quit [Quit: Leaving]
20161004 19:49:40-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 19:50:40-!- ChipmunkV [~vova@static-89-94-113-91.axione.abo.bbox.fr] has quit [Quit: ChipmunkV]
20161004 19:54:17-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 248 seconds]
20161004 19:55:06-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20161004 19:56:28-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20161004 19:57:43< vultraz> fuuuuuuuuuuu
20161004 19:58:18< vultraz> fuuuuuu
20161004 19:58:22< vultraz> why is this crashing
20161004 20:00:14< zookeeper> a wizard did it
20161004 20:00:21< vultraz> Program received signal SIGSEGV, Segmentation fault.
20161004 20:00:23< vultraz> In std::map<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, config::attribute_value, std::less<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::allocator<std::pair<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, config::attribute_value> > >::operator[](std::__cxx11::basic_string<char, std
20161004 20:00:24< vultraz> ::char_traits<char>, std::allocator<char> > const&) () ()
20161004 20:00:26< vultraz> hm
20161004 20:08:33-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 20:13:40< vultraz> ok, what the hell
20161004 20:13:57< vultraz> this crash happens if you click quickly and move your mouse away :|
20161004 20:15:22< vultraz> i don't get it..
20161004 20:16:37< vultraz> hm
20161004 20:17:35-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Read error: Connection reset by peer]
20161004 20:17:45< DeFender1031> Ah, Errors involving C++ templates. Such fun to try to track down.
20161004 20:17:56-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20161004 20:20:50< vultraz> it's a problem with the widget
20161004 20:21:02< vultraz> so not my fault
20161004 20:23:41< vultraz> I shall delegate this to celticminstrel 
20161004 20:24:01< celticminstrel> ...
20161004 20:24:46< vultraz> the bug appears to occur if you select a dropdown entry too quickly
20161004 20:24:51< vultraz> and move your mouse out of the list
20161004 20:57:29-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20161004 21:01:16-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20161004 21:13:16< Aginor> hmm
20161004 21:13:37< Aginor> so I started to fiddle around with pango text handling and SDL/wesnoth/pango alpha handling last night
20161004 21:14:07< Aginor> it didn't yield great results as a first go, I managed to make all text disappear (presumably throught transparency)
20161004 21:15:01< Aginor> I'm contemplating whether we should wrap up SDL_Color in a wrapper like we do with all other things
20161004 21:15:13< Aginor> and stop custom-packing uint32s
20161004 21:15:21-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 248 seconds]
20161004 21:15:25< Aginor> how would that potentially affect the WML?
20161004 21:18:15-!- ancestral [~ancestral@75-168-189-115.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20161004 21:31:01-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has quit [Ping timeout: 240 seconds]
20161004 21:35:19-!- Nobun [~nobun@host219-23-dynamic.50-82-r.retail.telecomitalia.it] has quit [Quit: Salve a tutti]
20161004 22:06:54-!- fabi [~fabi@176.5.137.156] has joined #wesnoth-dev
20161004 22:06:54-!- fabi_ [~fabi@176.7.22.57] has quit [Read error: Connection reset by peer]
20161004 22:16:08-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Read error: Connection reset by peer]
20161004 22:16:34-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20161004 22:17:01-!- ancestral [~ancestral@75-168-189-115.mpls.qwest.net] has joined #wesnoth-dev
20161004 22:17:48-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20161004 22:25:15-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20161004 22:27:09-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 22:29:48-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20161004 22:46:24-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20161004 22:56:13< mattsc> Huh, funny …
20161004 22:56:44-!- Duthlet [~Duthlet@dslb-188-104-253-155.188.104.pools.vodafone-ip.de] has quit [Quit: leaving]
20161004 22:57:01< mattsc> ai.check_stopunit() returns false for gamestate_changed on a unit without moves
20161004 22:57:15< mattsc> ai.stopunit_moves() returns true
20161004 23:01:10< vultraz> mattsc: did you see i reverted that problem commit?
20161004 23:01:19< vultraz> can you confirm the issue is gone for you
20161004 23:06:08-!- irker193 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20161004 23:24:07< mattsc> vultraz: I did see it, but I have not checked it out yet.  Will do so before the end of the day.
20161004 23:27:29< vultraz> ok
20161004 23:30:38-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20161004 23:33:05-!- irker506 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20161004 23:33:05< irker506> wesnoth: Charles Dang wesnoth:master 485e24ee2692 / src/gui/dialogs/multiplayer/mp_staging.cpp: MP Staging: fixed UB when assigning chat connection in local game (bug #25142) https://github.com/wesnoth/wesnoth/commit/485e24ee269261101626c1a6e249783461e80dcf
20161004 23:52:45< irker506> wesnoth: Charles Dang wesnoth:master 7ad4ba4083d5 / data/gui/widget/window_tooltip_large.cfg: Tooltips: some small tweaks to visuals https://github.com/wesnoth/wesnoth/commit/7ad4ba4083d5e4f4a63d763301880a266e1b5643
20161004 23:53:07< vultraz> feedback appreciated
20161004 23:56:35-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20161004 23:58:12-!- Greg-Bog_ [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
--- Log closed Wed Oct 05 00:00:57 2016