--- Log opened Wed Dec 16 00:00:21 2015
20151216 00:06:04-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20151216 00:06:17-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20151216 00:09:55-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20151216 00:11:06-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Client Quit]
20151216 00:43:37< shadowm> gfgtdf, celticminstrel, zookeeper, Elvish_Hunter, aquileia: Friendly reminder that I posted this thread and the clock is ticking: https://mail.gna.org/public/wesnoth-dev/2015-12/threads.html
20151216 00:44:11< shadowm> We got 1.13.2 sorted out already, but there's still a lot left to discuss and exactly two people (one via IRC query) have actually bothered to give me anything even remotely resembling an opinion.
20151216 00:45:09< shadowm> I'm sure I forgot other names on that list. Who else is active? loonycyborg and Ivanovic (who is on vacation and unable to see this).
20151216 00:46:03-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20151216 00:46:36< shadowm> vultraz: Read the logs.
20151216 00:47:30< shadowm> Also, if I'm sounding bossy I'd like to note that on this occasion it's completely deliberate. I already said I won't be here for a while so we need to get this stuff sorted out ASAP.
20151216 00:47:55< shadowm> If no-one says anything I'll just assume a) we won't participate in GSoC; b) that no-one's interested in 1.14 anymore.
20151216 00:48:12< celticminstrel> I really don't have anything to say about GSoC... well... I could be a mentor or something... but I don't have any ideas at the top of my head for projects, so unless someone else says something it's kind of irrelevant.
20151216 00:48:20< celticminstrel> Do I have to respond on the mailing list?
20151216 00:48:22< shadowm> It's as simple as that. It's been 14 days, plenty of time for you people to decide.
20151216 00:48:39< shadowm> Yes, you have to respond on the ML if you want your opinion to be counted.
20151216 00:50:41< ancestral> celticminstrel: I think people would appreciate your thoughts, whether they’re useful or not
20151216 00:50:54-!- Necrosporus_ [~Necrospor@unaffiliated/necrosporus] has quit [Ping timeout: 256 seconds]
20151216 00:51:58< vultraz> shadowm: I'll reply to the ML later - I have somewhere to be soon. But as for the release schedule, I support it. I don't know about GSoC. I'm not really mentor (or student) material, but I'm sure I could come up with project ideas should then need arise.
20151216 00:51:59< shadowm> IRC is an excellent medium for short-term communication, not so for discussing stuff that will matter two years from now and that will need to be quoted and linked every once in a while.
20151216 00:52:21< shadowm> Or for coordinating people on completely different time zones.
20151216 00:52:43< vultraz> shadowm: you can hand over twitter and gh whenever you're free
20151216 00:52:55< shadowm> vultraz: Assets. I need the assets.
20151216 00:53:05< celticminstrel> At least this mailing list properly has Reply-To set.
20151216 00:53:06< vultraz> assets?
20151216 00:54:24< shadowm> vultraz: 20:52:31  * shadowm ponders whether to wait for vultraz to hand him the GitHub and Twitter assets or just go ahead.
20151216 00:54:51< vultraz> oh, I read that wrong
20151216 00:54:53< vultraz> ok, ok
20151216 00:55:06< vultraz> what assets?
20151216 00:55:11< vultraz> sorry im kinda rushed here
20151216 00:56:31< shadowm> The avatar.
20151216 00:56:43< shadowm> GitHub apparently needs 200x200 px, not sure about Twitter.
20151216 00:57:30< shadowm> I'll tell you as soon as this behemoth of a profile picture loads up.
20151216 00:59:13< vultraz> you want my logo renders? https://www.dropbox.com/sh/lncz6sp106f3fyn/AADApxh8acnBZ29F6b4G2zzIa?dl=0
20151216 00:59:58< shadowm> It's always great how mobile broadband is like one whole decade behind in terms of user expectations.
20151216 01:00:29< shadowm> "lol who would ever need to transfer more than 1 GiB a month it's all you'll ever need in your life lol"
20151216 01:02:31< shadowm> Where do I see the profile photo size recommendations?
20151216 01:02:45< shadowm> Pretty sure they used to be in the profile edit view. Good job.
20151216 01:03:09< celticminstrel> They probably assume that phones don't watch movies or something weird like that.
20151216 01:03:21< shadowm> Or get software updates.
20151216 01:03:23< celticminstrel> Or that if they do they're downloaded on a computer and then transferred over.
20151216 01:03:26< celticminstrel> Ah, heh, yeah.
20151216 01:03:35< celticminstrel> Those can sometimes be large too.
20151216 01:03:35< shadowm> Also, s/phones/phones, laptops, and tablets (they actually market to all this stuff)/
20151216 01:04:25< shadowm> For example, updating to Windows 10 is a DVD-sized (4 GiBish) download I believe.
20151216 01:04:38< shadowm> People want to update to Windows 10 because it has a ten on it.
20151216 01:04:48< celticminstrel> Is that why they skipped 9?
20151216 01:05:05< shadowm> Wouldn't be surprised if that were the case.
20151216 01:05:17< shadowm> "Let's catch up with Apple, they have been on 10 for a decade already."
20151216 01:06:04< shadowm> vultraz: 400x400 px for Twitter.
20151216 01:06:28< vultraz> shadowm: can you just scale the 512 yourself?
20151216 01:10:05-!- irker570 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20151216 01:19:45< shadowm> Oh, turns out I do have the SVG version of the logo already!
20151216 01:29:12< shadowm> Aaaand my render looks completely wrong.
20151216 01:29:55< shadowm> I thought Groovy used Inkscape too.
20151216 01:31:51< shadowm> Won't upload because that'll take me forever, but for example, the shield's inner blue surface is darker and lacks the light reflection shiny.
20151216 01:32:16< shadowm> Perhaps I'm supposed to use a particular version of Inkscape... ¬_¬
20151216 01:38:36< vultraz> shadowm: i believe re recommended scaling the pngs
20151216 01:38:51< vultraz> which is what i did
20151216 01:39:01< vultraz> just scale the 512 to a 400
20151216 01:39:04< vultraz> ive done it, it looks fine
20151216 01:39:14< vultraz> dont waste time rendering a new version in inkscape
20151216 01:42:20< shadowm> Yeah I wasted exactly 30 seconds doing the renders.
20151216 01:42:58< shadowm> The most time-consuming part was inputting a 200 and a 400. :p
20151216 01:43:37< shadowm> (That's another "shadowm can't count past 2" joke in case you couldn't tell.)
20151216 01:43:55< celticminstrel> Rendering shouldn't take long when you consider that (approximately) it has to re-render it every time you change anything in the document...
20151216 01:44:20< celticminstrel> Or change the zoom level.
20151216 01:44:53< shadowm> The actual render takes about 1.7 seconds.
20151216 01:45:57< celticminstrel> I imagine it's about the same as scaling the already rendered one down.
20151216 01:46:19< celticminstrel> Considering the process as a whole (starting from opening the file).
20151216 01:47:34< shadowm> "Save to my Dropbox" -> "There was a problem completing this request."
20151216 01:48:19< shadowm> Silly Javascript applications that break with the smallest amount of network lag.
20151216 01:48:42< celticminstrel> Maybe that's why Youtube keeps crashing on me.
20151216 01:49:01< shadowm> YouTube is the worst in that regard, really.
20151216 01:49:27< shadowm> The HTML5 player is also infinitely more unreliable than Flash if you are running Firefox on Linux.
20151216 01:54:05< shadowm> Okay, just learned an important lesson: don't trust GitHub's downscaler.
20151216 01:54:41< shadowm> I have no idea what algorithm it uses but it's clearly bad: https://avatars2.githubusercontent.com/u/4024232?v=3&s=200
20151216 01:59:21< shadowm> Okay, replaced that with my own.
20151216 02:03:27< shadowm> Twitter decided I'm not cool enough to upload the full 400px.
20151216 02:04:16< shadowm> Probably because I'm unable to upload 193 KiB in under 2 seconds.
20151216 02:07:14-!- irker790 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20151216 02:07:14< irker790> wesnoth: gfgtdf wesnoth:master f41bbd835cce / data/campaigns/Legend_of_Wesmere/scenarios/chapter1/03_Kalian_under_Attack.cfg: Fix LoW scenario3 in debug mode http://git.io/v0KlX
20151216 02:07:22< shadowm> Okay! Managed to get it through after four painful attempts.
20151216 02:08:00< shadowm> vultraz: https://twitter.com/Wesnoth  https://github.com/wesnoth
20151216 02:09:49< celticminstrel> ...I don't think github actually scaled it down at all. It just set the width and height of the image so that the browser scales it down to fit.
20151216 02:10:08< celticminstrel> Unless there are places where it appears larger than in that link.
20151216 02:10:11< celticminstrel> I dunno.
20151216 02:10:28< celticminstrel> Not important.
20151216 02:11:45< shadowm> celticminstrel: 22:59:22 <shadowm> Okay, replaced that with my own.
20151216 02:11:56< celticminstrel> Yeah.
20151216 02:15:15-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 240 seconds]
20151216 02:35:13-!- gfgtdf_ [~chatzilla@x50abe5dd.dyn.telefonica.de] has joined #wesnoth-dev
20151216 02:36:52-!- gfgtdf_ [~chatzilla@x50abe5dd.dyn.telefonica.de] has quit [Client Quit]
20151216 02:36:53-!- gfgtdf [~chatzilla@f054135096.adsl.alicedsl.de] has quit [Ping timeout: 246 seconds]
20151216 02:41:47-!- ancestral [~ancestral@184-100-99-25.mpls.qwest.net] has quit [Quit: ancestral]
20151216 03:08:29-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20151216 03:09:29< shadowm> celticminstrel: Meaning that the jagged version isn't there anymore.
20151216 03:09:37< shadowm> Sorry, I was in the middle of typing that when power went out.
20151216 03:10:12< shadowm> Good guy Firefox brought up the four windows with 50 something tabs open back.
20151216 03:13:45< shadowm> It's pretty obvious at this point that I'll never be able to use a computer with less than 16 GiB of RAM ever again. The next logical step clearly is to upgrade to a 64 GiB box.
20151216 03:20:39< shadowm> Okay, it appears my desktop did not originate the power failure, so I'm going to resume whatever it was I was doing.
20151216 03:20:43< shadowm> Time for 1.13.2.
20151216 04:01:25< shadowm> neverEnough_: Is this you? http://forums.wesnoth.org/memberlist.php?mode=viewprofile&u=126961
20151216 04:01:57< shadowm> (I know the question sounds redundant, but you never know...)
20151216 04:02:58< celticminstrel> Could try looking at IP addresses, though I guess that's still not a sure thing.
20151216 04:05:03< shadowm> Too lazy and it's a very error-prone heuristic when NAT routing gets involved.
20151216 04:05:47< shadowm> (Mobile broadband providers around the world love that, especially with the advent of IPv4 address pool depletion.)
20151216 04:06:34< celticminstrel> Ah...
20151216 04:07:20< celticminstrel> I guess I'm behind a NAT router too, that's why my sister shows as having the same IP.
20151216 04:09:08< shadowm> Well, it's a typical configuration for a shared household. It just gets unimaginably messy for hosting providers and forum moderators when you have a whole country behind the same IP address.
20151216 04:09:42< shadowm> Or more commonly, a whole ISP.
20151216 04:10:03< celticminstrel> I can imagine...
20151216 04:10:19< shadowm> There are whole single-address countries too I'm sure, IIRC North Korea is one.
20151216 04:10:50< shadowm> But it also has its own private Internet that's inaccessible to the rest of the world (just like our Internet is inaccessible from within) so that kind of doesn't count.
20151216 04:11:04< celticminstrel> I thought that was China.
20151216 04:11:45< shadowm> China has a very aggressive Internet censorship policy and infrastructure in place but I don't think it's implemented the same way.
20151216 04:12:36< shadowm> IIRC it's just packet filtering.
20151216 04:13:59< shadowm> For example, I know for a fact Chinese users access our forums.
20151216 04:14:41< shadowm> Back in 2008 there was a period of time when they couldn't because we were hosted by ibiblio.org, which was apparently blacklisted for hosting politically-charged content or something.
20151216 04:23:21-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has joined #wesnoth-dev
20151216 04:24:28-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has left #wesnoth-dev []
20151216 04:33:51-!- shadowm changed the topic of #wesnoth-dev to: Wesnoth Developers Channel | >>> Want to help? Go here: http://r.wesnoth.org/t42911 (and thanks!) <<< | Logs: http://irclogs.wesnoth.org | Bug tracker: http://bugs.wesnoth.org
20151216 04:36:49< shadowm>     g++: internal compiler error: Killed (program cc1plus)
20151216 04:37:08< shadowm> I made gcc 4.7 sad.
20151216 04:44:05< shadowm> I can't build master on basilic OR gonzo, great.
20151216 04:44:42< shadowm> Not that anyone seems to use the dev server on baldras even.
20151216 04:46:10< shadowm>   Checking for Simple DirectMedia Layer library version >= 1.2.10... no
20151216 04:46:35-!- Appleman1234 [~Appleman1@KD119104008160.au-net.ne.jp] has quit [Ping timeout: 240 seconds]
20151216 04:50:55< shadowm> Ugh, why, sdldir is pointing to a dir that has SDL 1.2.15.
20151216 04:52:18< shadowm> And scons is ignoring sdldir. Great.
20151216 05:01:55-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 240 seconds]
20151216 05:07:10-!- irker790 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20151216 05:10:55-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20151216 05:36:17-!- Kwandulin [~Miranda@p200300760F250A838DBA335A814AC90F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20151216 05:43:50-!- Appleman1234 [~Appleman1@KD119104019080.au-net.ne.jp] has joined #wesnoth-dev
20151216 06:06:33< Aginor> shadowm: I had another look at http://wiki.wesnoth.org/DevelopmentRoadmap . I'd be interested to be in on the Display Engine Refactoring, I'm already poking a lot of those things
20151216 06:10:06-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 272 seconds]
20151216 06:10:37< shadowm> Aginor: Yeah, I imagine so.
20151216 06:11:10< shadowm> There's really not a lot that can be done with the engine moving forward without either rewriting the whole pipeline or adding more hacks on top of our current jenga tower.
20151216 06:11:47< Aginor> a spring clean might be useful
20151216 06:12:02< shadowm> And the most important part of all: documentation.
20151216 06:12:04< Aginor> I've been pondering the opengl transition
20151216 06:12:15< shadowm> Right now the worst part about the display code is that it has too many undocumented entry points.
20151216 06:12:28-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151216 06:12:33< Aginor> I've currently have a private branch with a bunch of changes to the resize handling I need to get in properly too
20151216 06:12:40< Aginor> I know
20151216 06:12:41< vultraz> Aginor: more than "might" :P
20151216 06:12:42< shadowm> So it shouldn't be any wonder that users (including GUI2) keep breaking it in subtle ways.
20151216 06:13:07< Aginor> I am still mapping out the tentacles it has
20151216 06:13:34< Aginor> but it would be good to break display into two parts. Display management and rendering, gamestate management.
20151216 06:14:07< Aginor> vultraz: was that you volunteering to start work? :D
20151216 06:14:29< vultraz> Aginor: I'm always willing to help
20151216 06:14:38< vultraz> as long as it doesn't involve copious Maths 
20151216 06:15:34< Aginor> vultraz: this won't involve maths yet. Although I should make a list of medium term wins for core game engine and rendering pipeline that I've thought of
20151216 06:15:43< shadowm> The display class currently has too many responsibilities, yes. Ideally it should just manage framebuffer access and event dispatching and let others implement the UI layer (theme UI) and map rendering in an aggregate fashion.
20151216 06:16:08< Aginor> shadowm: yes
20151216 06:16:28< shadowm> Or more accurately, viewport rendering.
20151216 06:16:39< Aginor> shadowm: display should manage display, not the other stuff.
20151216 06:16:44< shadowm> (Map + units + whatever else we might come up with in the future.)
20151216 06:17:15< Aginor> indeed
20151216 06:19:40-!- Appleman1234 [~Appleman1@KD119104019080.au-net.ne.jp] has quit [Ping timeout: 260 seconds]
20151216 06:21:02< Aginor> wedge009: ping?
20151216 06:25:35< Aginor> I'm planning to do the following on a resize: paint the window black. Not accept user input until there's been 100ms without resize events, at which point I will trigger a full redraw of the ui
20151216 06:25:51< shadowm> My hope is that at some point in the hopefully not too far future we'll be doing OpenGL with the ability to use shader effects in unit animations, terrain-based ambient light, animated fog, etc.
20151216 06:25:52< Aginor> that should fix the remaining redraw issues that are around
20151216 06:26:00< shadowm> As well as non-modal dialogs.
20151216 06:26:16< Aginor> shadowm: we need to kill the wesnoth blit-lookalike first
20151216 06:26:30< Aginor> and we need to kill sdl1 before we can do that
20151216 06:26:52< shadowm> You mean the local blit implementation?
20151216 06:27:08< Aginor> yes
20151216 06:27:26< Aginor> it's slow and needs to die.
20151216 06:27:33< Aginor> and it'll never work with opengl
20151216 06:27:54< shadowm> Obviously, the whole concept of blitting doesn't translate.
20151216 06:27:57< Aginor> (the last reason is the real reason it needs to die)
20151216 06:28:27< shadowm> The local blit implementation only exists to deal with shortcomings in the SDL 1.2 implementation revolving around alpha blending IIRC.
20151216 06:28:57< shadowm> If SDL 2.0 fixes that then there's obviously no point in using our custom blitter.
20151216 06:29:03< Aginor> Just replacing it with the sdl2 blitting leads to poor results. There's work to be done there.
20151216 06:29:17< Aginor> Potentially not an insignificant amount
20151216 06:34:58< shadowm> Not much I can say about that since I've never really understood the underlying theory behind that.
20151216 06:35:29< Aginor> poor results == different, broken/bad in this context
20151216 06:36:24< shadowm> I mean behind our custom blitter.
20151216 06:36:44< Aginor> ah
20151216 06:37:10< Aginor> it's to try to provide some alpha blending, I think
20151216 06:37:37< Aginor> I haven't looked at it in too great detail, it's just made it onto my mental map of blockers
20151216 06:37:50< Aginor> hmm, resizing in a story screen really breaks stuff
20151216 06:38:07< shadowm> Oh, that's a legacy bug.
20151216 06:38:25< Aginor> I wonder if I should keep bandaging stuff or do a complete revamp of the flow of window events
20151216 06:38:28< shadowm> The story screen simply doesn't opt into handling resize events.
20151216 06:38:44< Aginor> right
20151216 06:38:54< Aginor> I'm tempted to ignore that now then
20151216 06:39:08< shadowm> It's best to ignore pre-existing bugs for now, yes.
20151216 06:40:05< shadowm> I had a patch once adding the necessary code to opt into handling resize in story screens, but there were additional problems to deal with such as rescaling the background and re-laying out the contents, so I just kind of forgot about it.
20151216 06:40:47< Aginor> ok...
20151216 06:41:02< Aginor> I'll tidy up what I have then and put into the sdl2 branch
20151216 06:41:16< shadowm> The story screen is all hand-written code, that's the big issue with it.
20151216 06:41:21< Aginor> it fixes the biggest issues when resizing while playing the game, making it about as poor as sdl1
20151216 06:41:45< shadowm> It doesn't use GUI1 (which doesn't really handle screen resize events anyway) or GUI2.
20151216 06:43:10< shadowm> It does have a few GUI1 widgets lying around implementing the control buttons but they are just there, floating around in the infinite void of the heap, reacting to a local event loop as needed.
20151216 06:43:35< Aginor> shadowm: all of the local event loop stuff was a large contributor to crashes :)
20151216 06:44:03< Aginor> anyway, I have visitors coming around so I need to take off
20151216 06:44:32< Aginor> I'll get my current changes cleaned up, then I'll probably hit the merge button this weekend
20151216 06:44:49< Aginor> I'll hassle people for additional testing once I've done the next commit
20151216 06:45:07< shadowm> Well, you can merge any time now, really.
20151216 06:45:24< Aginor> shadowm: yes, but I want my next set of changes to go in at the same time
20151216 06:45:27< shadowm> Unless it won't compile/is known to not work for anyone at all.
20151216 06:45:41< Aginor> otherwise people will tell me that resizing turns the screen black while the game is running
20151216 06:46:16< Aginor> I have the fix sitting here, I just need to remove my debug statements and commit it
20151216 06:46:20< shadowm> In general, the only two places where resize works reliably with SDL 1.2 is GUI2 dialogs*, and the in-game theme UI (where you spend most of your time during gameplay), and the default MP lobby.
20151216 06:46:53< shadowm> (* except when they interact with the theme UI like the [message] box, then things get weird.)
20151216 06:47:19< Aginor> the in-game theme UI is looking broken in SDL2, it doesn't appear to get resize events fully. Invalidating the display on a resize fixes it
20151216 06:47:19< shadowm> (* so in a way that means the Title Screen and all GUI2 dialogs launched from it are the only ones immune to that clause)
20151216 06:47:49< Aginor> gui2 resize events are very cpu intensive for some reasons
20151216 06:47:51< shadowm> Also, the in-game theme UI glitches noticeably even when not blocked by a GUI2 event loop.
20151216 06:48:16< shadowm> I mean, it doesn't *glitch* per se. It works just fine, it's the UI buttons with icon overlays that glitch.
20151216 06:48:53< Aginor> forcing the redraw fixes that :)
20151216 06:48:55< shadowm> Because someone (haven't figured out who yet) apparently didn't think things through when they added support for icon overlays to buttons in the theme UI.
20151216 06:49:24< shadowm> And there are a lot of triggers besides window resize events for the same glitch.
20151216 06:49:37< Aginor> I really need to head off
20151216 06:49:39< shadowm> [redraw], disabling/enabling fog/shroud with debug mode, to name a few.
20151216 06:49:44< Aginor> I'll catch up later
20151216 06:49:52< shadowm> Okay, see you then.
20151216 06:49:56< Aginor> bye until then :)
20151216 06:50:37< shadowm> For the curious, the cause (if my memory serves) is that the theme UI buttons need to be recreated every time "something changes".
20151216 06:50:48< shadowm> Scare quotes because there are a lot of ways this can happen.
20151216 06:51:01< shadowm> So the UI buttons that have icon overlays get recreated.
20151216 06:51:05< shadowm> Without their overlays set.
20151216 06:51:12< shadowm> It's glorious.
20151216 06:51:47< shadowm> I wanted to get a fix in for 1.12.5 but didn't get the opportunity to work on it, unfortunately.
20151216 06:52:49< shadowm> Essentially, right now someone has to explicitly ask for the button states to be updated, at this someone is not all of the display redraw entry points.
20151216 06:54:35< shadowm> Or more accurately, the set of all display redraw entry points that cause theme UI buttons to be reinstantiated is not identical to the set of display redraw entry points that cause the theme UI buttons to have their overlays set (which IIRC is actually a void set).
20151216 06:55:12< shadowm> And this is why I hate the fact that there are so many different ways to update the display.
20151216 06:56:06< shadowm> All largely non-descript.
20151216 06:57:03< shadowm> I need to look into it again one of these days and see if I can finally land a fix for 1.12.6.
20151216 07:08:03-!- Kwandulin_2 [~Miranda@p200300760F250AFF8DBA335A814AC90F.dip0.t-ipconnect.de] has joined #wesnoth-dev
20151216 07:08:05-!- Kwandulin [~Miranda@p200300760F250A838DBA335A814AC90F.dip0.t-ipconnect.de] has quit [Ping timeout: 260 seconds]
20151216 07:19:12-!- Appleman1234 [~Appleman1@KD119104002187.au-net.ne.jp] has joined #wesnoth-dev
20151216 07:49:26-!- Appleman1234 [~Appleman1@KD119104002187.au-net.ne.jp] has quit [Ping timeout: 246 seconds]
20151216 08:10:29-!- celticminstrel is now known as celmn|sleep
20151216 08:10:51-!- celmn|sleep is now known as celmin|sleep
20151216 08:16:59-!- shadowm_desktop2 [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20151216 08:17:52-!- shadowm_desktop is now known as Guest26439
20151216 08:17:52-!- shadowm_desktop2 is now known as shadowm_desktop
20151216 08:18:23< Kwandulin_2> In 1.13.2a, not specifying a difficulty in the [campaign]-tag, leads to a crash.
20151216 08:19:11-!- Guest26439 [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 246 seconds]
20151216 08:20:07< vultraz> gahh
20151216 08:21:30-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20151216 08:22:53< Kwandulin_2> In 1.13.1, the game allowed to leave out the difficulty-tag (and i guess it just used the medium difficulty as a default); the 'choose a difficulty' window had just been skipped then. The current behaviour is a bit annoying for me, because in my campaign I just use one difficulty and I don't reallly want the 'choose a difficulty' window to pop up (because there is no choice anyway)
20151216 08:23:31< shadowm> Yes, it's most certainly not an intentional regression.
20151216 08:25:51< shadowm> (Rule of thumb: if WML or Lua can cause the game to crash or unexpectedly quit, it's a bug.)
20151216 08:42:12-!- boucman_work [~jrosen@lns-bzn-47f-62-147-140-162.adsl.proxad.net] has joined #wesnoth-dev
20151216 08:42:12-!- boucman_work [~jrosen@lns-bzn-47f-62-147-140-162.adsl.proxad.net] has quit [Changing host]
20151216 08:42:12-!- boucman_work [~jrosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20151216 08:47:15-!- Appleman1234 [~Appleman1@KD119104006222.au-net.ne.jp] has joined #wesnoth-dev
20151216 08:47:41-!- horrowind [~Icedove@2a02:810a:8b00:1c54:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20151216 08:49:41-!- irker542 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20151216 08:49:41< irker542> wesnoth: Charles Dang wesnoth:master 7a1b5c80ff2e / src/game_initialization/create_engine.cpp: Fix a crash when entering a campaign with no difficulties provided http://git.io/v061d
20151216 08:49:42< irker542> wesnoth: Charles Dang wesnoth:master 591b1aa6be70 / src/game_initialization/create_engine.cpp: Give new [difficulty] tag precedence over old difficulties= syntax http://git.io/v061F
20151216 08:49:46< vultraz> Kwandulin_2: ^
20151216 08:54:37< vultraz> ( shadowm also ^ so you know it's fixed)
20151216 09:06:50< zookeeper> anyone objects if i just un-announce http://forums.wesnoth.org/viewtopic.php?f=9&t=33460 ? that's a blanket ban on posting better sprites for example for campaign characters
20151216 09:07:45< zookeeper> i'd rather risk someone posting a bad attempt at improving the spearman once a year, than someone thinking we don't want a better haldric
20151216 09:07:56< zookeeper> with the latter being exactly what that thread claims
20151216 09:09:40< zookeeper> well, actually i already did it
20151216 09:11:39< zookeeper> i'll do the same for http://forums.wesnoth.org/viewtopic.php?f=18&t=28609 because all the core units are already covered and the campaign units should be in an "art needed" list
20151216 09:22:11-!- Appleman1234 [~Appleman1@KD119104006222.au-net.ne.jp] has quit [Ping timeout: 246 seconds]
20151216 09:52:52< zookeeper> and here's a replacement thread: http://forums.wesnoth.org/viewtopic.php?f=9&t=43345
20151216 09:53:07-!- Appleman1234 [~Appleman1@KD119104005179.au-net.ne.jp] has joined #wesnoth-dev
20151216 09:57:11-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 246 seconds]
20151216 09:57:39-!- Kwandulin_2 [~Miranda@p200300760F250AFF8DBA335A814AC90F.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20151216 10:00:17-!- Kwandulin [~Miranda@p200300760F250AFF38585A3DCD2EE6E0.dip0.t-ipconnect.de] has joined #wesnoth-dev
20151216 10:00:44< irker542> wesnoth: ln-zookeeper wesnoth:master e4089c02b41d / data/campaigns/The_South_Guard/units/Infantry_Commander.cfg: Use Infantry Lieutenant attack animation for Infantry Commander too http://git.io/v0imF
20151216 10:00:46< irker542> wesnoth: ln-zookeeper wesnoth:master 3c091dbe7d0a / / (18 files in 11 dirs): Merge branch 'master' of https://github.com/wesnoth/wesnoth http://git.io/v0imb
20151216 10:55:44-!- Appleman1234 [~Appleman1@KD119104005179.au-net.ne.jp] has quit [Ping timeout: 272 seconds]
20151216 11:38:50-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has joined #wesnoth-dev
20151216 11:52:00-!- Appleman1234 [~Appleman1@KD119104007122.au-net.ne.jp] has joined #wesnoth-dev
20151216 12:10:59< vultraz> zookeeper: are you still not ok with adding a deprecation warning for the old campaign difficulty syntax
20151216 12:13:24< zookeeper> i can live with it
20151216 12:14:11< zookeeper> but that's because i assume the internal conversion from old to new is not short and simple and easily encapsulated in one place
20151216 12:14:28< zookeeper> because if it is, then i see no need to deprecate because there's no maintenance cost involved
20151216 12:14:57< vultraz> I want to deprecate it because it's an incredibly confusing syntax
20151216 12:15:01< vultraz> especially to newcomers 
20151216 12:15:43< vultraz> plus conversion from old to new has to be handled wherever it's used
20151216 12:15:59< vultraz> (in this case both in the campaign difficulty dialog AND the [message] implementation)
20151216 12:16:36< zookeeper> what does [message] have to do with it? O.o
20151216 12:17:18< vultraz> MENU_IMG_TXT :P
20151216 12:17:39< vultraz> both that and the campaign difficulties used the old tlegacy_menu_item class
20151216 12:18:35< vultraz> I want to deprecate that syntax entirely
20151216 12:28:21< zookeeper> well, as said i can live with it
20151216 12:29:17< zookeeper> so what does [message] and MENU_IMG_TXT have to do with each other?
20151216 12:30:11< vultraz> latter was used with the former to give [options] three columns of content
20151216 12:30:18< vultraz> and image, label, and description
20151216 12:34:23-!- Kwandulin [~Miranda@p200300760F250AFF38585A3DCD2EE6E0.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20151216 12:34:53< irker542> wesnoth: Charles Dang wesnoth:master 6b82b0dace0b / src/gui/dialogs/campaign_difficulty.cpp: Added deprecation warning for [campaign] difficulties,difficulty_descriptions= http://git.io/v0ix4
20151216 12:34:56< irker542> wesnoth: Charles Dang wesnoth:master 2178ac49b14b / src/gui/dialogs/campaign_difficulty.cpp: tcampaign_difficulty: reduce a var scope http://git.io/v0ixB
20151216 12:55:47< irker542> wesnoth: Charles Dang wesnoth:master 3be43abb7407 / data/campaigns/Legend_of_Wesmere/lua/wml_tags.lua: LoW: Revert a small erroneous change carried over in be68ebc0a70 http://git.io/v0PUT
20151216 12:58:54< vultraz> I can't figure out why that map offset bug is back :(
20151216 12:59:15< vultraz> er
20151216 12:59:19< vultraz> map label offset bug in LoW
20151216 13:12:21< irker542> wesnoth: Charles Dang wesnoth:master bed6339a6599 / data/campaigns/tutorial/scenarios/02_Tutorial_part_2.cfg: T S2: used narrative messages to describe the Shaman and Archer on recruit (bug  http://git.io/v0POO
20151216 13:14:53-!- horrowind [~Icedove@2a02:810a:8b00:1c54:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20151216 13:56:01-!- mjs-de [~mjs-de@x4db5286e.dyn.telefonica.de] has joined #wesnoth-dev
20151216 14:05:56-!- Appleman1234 [~Appleman1@KD119104007122.au-net.ne.jp] has quit [Ping timeout: 256 seconds]
20151216 14:06:57-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20151216 14:40:49-!- prkc [~prkc@catv-80-98-216-55.catv.broadband.hu] has joined #wesnoth-dev
20151216 14:54:19-!- skyfaller [~skyfaller@73.188.0.22] has joined #wesnoth-dev
20151216 14:54:20-!- skyfaller [~skyfaller@73.188.0.22] has quit [Changing host]
20151216 14:54:20-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has joined #wesnoth-dev
20151216 14:54:22-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has quit [Remote host closed the connection]
20151216 14:59:39-!- skyfaller [~skyfaller@c-73-188-0-22.hsd1.pa.comcast.net] has joined #wesnoth-dev
20151216 14:59:39-!- skyfaller [~skyfaller@c-73-188-0-22.hsd1.pa.comcast.net] has quit [Changing host]
20151216 14:59:39-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has joined #wesnoth-dev
20151216 14:59:57< zookeeper> gah, i have no idea how to use lua
20151216 15:03:39-!- Appleman1234 [~Appleman1@KD119104004147.au-net.ne.jp] has joined #wesnoth-dev
20151216 15:11:00-!- horrowind [~Icedove@2a02:810a:8b00:1c54:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20151216 15:16:06-!- boucman_work [~jrosen@wesnoth/developer/boucman] has quit [Ping timeout: 240 seconds]
20151216 15:18:16-!- Kwandulin [~Miranda@p200300760F250AFF3DFA6D5512351E25.dip0.t-ipconnect.de] has joined #wesnoth-dev
20151216 15:37:00< vultraz> zookeeper: need help?
20151216 15:38:10< zookeeper> vultraz, well i could ask a few questions...
20151216 15:38:29< zookeeper> 1. is this the only way to add a global event, as in gets inserted everywhere? http://pastebin.com/6DQVMnVY
20151216 15:39:11< zookeeper> IIRC you can't have a top-level [event]s in pure WML
20151216 15:40:47< zookeeper> 2. what's the simplest way to use that code, but to also define the event body in WML instead of as super-awkward lua tables?
20151216 15:46:20< vultraz> You also have to load it globally in your campaign
20151216 15:46:47< vultraz> in toplevel
20151216 15:46:56< vultraz> and no you cannot in WML
20151216 15:47:05< vultraz> except in 1.13 we have events in [campaign]
20151216 15:47:05< zookeeper> what? i did just put that snippet at top-level and it worked
20151216 15:50:13-!- gfgtdf [~chatzilla@x50abe5dd.dyn.telefonica.de] has joined #wesnoth-dev
20151216 15:51:27< gfgtdf> vultraz: i added that code for the map header on purpose when parotign theose commits to master becasue that commit also changed some maps back to the old format.
20151216 15:51:34< gfgtdf> zookeeper: in 1.12 or 1.13 ?
20151216 15:51:49< zookeeper> 1.13
20151216 15:52:27< gfgtdf> zookeeper: in currentl 1.13.2 you can put [event] in [campaign] so that it gets added to every scenario of that campaign.
20151216 15:52:35-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 240 seconds]
20151216 15:52:50< gfgtdf> zookeeper: alternativeley you can have a [modification] with that event which you then ableble manually
20151216 15:53:02< gfgtdf> zookeeper: the second option is specialyl useful for debuggin event
20151216 15:53:38< zookeeper> at no point have i said anything about a campaign so why are you both assuming that :|
20151216 15:53:51< zookeeper> global means global
20151216 15:55:04< zookeeper> basically i just want to write a global event, in WML
20151216 15:55:16< gfgtdf> zookeeper: you can still try the modification, i dont know why exactly why you want this so i cannot helpy ou more
20151216 15:56:01< gfgtdf> zookeeper: if you wnat it to be present in all the scenarios you shoudl consider adding your event as c++ code to teh wesnoth engine.
20151216 16:02:31< gfgtdf> vultraz: map label offsetb bugs happens on 1.13 and on 1.12
20151216 16:02:39< gfgtdf> in LoW i mean
20151216 16:06:23-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20151216 16:12:20-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 260 seconds]
20151216 16:13:01-!- irker542 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20151216 16:14:24-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20151216 16:27:29-!- Kwandulin [~Miranda@p200300760F250AFF3DFA6D5512351E25.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20151216 16:31:34-!- Appleman1234 [~Appleman1@KD119104004147.au-net.ne.jp] has quit [Ping timeout: 256 seconds]
20151216 16:46:59< vultraz> zookeeper: yes, and?
20151216 16:52:40< zookeeper> vultraz, nothing? if there's no way then there's no way
20151216 16:53:27< vultraz> zookeeper: I said 1.13 had events in [campaign]
20151216 16:53:56< zookeeper> i couldn't have been responding to that since the time difference was 1 second
20151216 16:54:34< vultraz> oh
20151216 16:54:36< vultraz> uh
20151216 16:54:47< vultraz> i meant you can't do it in pure wml
20151216 16:55:07< zookeeper> what does "it" refer to?
20151216 16:55:12< vultraz> i think you might be able to get away with a >> {MACRO} << hack
20151216 16:55:19< vultraz> zookeeper: [02:39:10]	zookeeper	IIRC you can't have a top-level [event]s in pure WML
20151216 16:56:07< zookeeper> okay
20151216 16:56:16-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20151216 16:56:53-!- Kwandulin [~Miranda@p200300760F250AFF54817E69DD687679.dip0.t-ipconnect.de] has joined #wesnoth-dev
20151216 16:58:36< zookeeper> i can only presume that some kind of technical difficulty is the reason why there's no lua function for turning a piece of WML (as a string) into a table
20151216 16:59:02-!- zombah [~zombah@2a02:28:3:1:214:4fff:fe47:5920] has quit [Quit: Leaving]
20151216 16:59:25< zookeeper> or any other way to sort of embed some WML directly into lua code
20151216 17:05:40< vultraz> guess no one's done it yet
20151216 17:07:45< gfgtdf> actualyl its quiet easy to create wml tables in Lua, so the only usecase would be to convert existing wml tables to lua by people who want to rewrite their addons to lua.
20151216 17:10:28< zookeeper> a possible workaround: bundle a dummy unit type which has the event you want, then have the global event simply create+kill that unit in prestart. i guess that might work.
20151216 17:12:30< gfgtdf> zookeeper: yes there is a addon by dugi that does exactly that.
20151216 17:12:49< zookeeper> cool
20151216 17:12:50< vultraz> of course there is
20151216 17:14:52< zookeeper> yeah, found it
20151216 17:24:40-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Quit: shadowm_desktop]
20151216 17:28:44-!- Appleman1234 [~Appleman1@119.104.11.122] has joined #wesnoth-dev
20151216 17:45:17-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has quit [Remote host closed the connection]
20151216 17:47:36< celmin|sleep> zookeeper: You could try something like this for that event code: http://pastebin.com/akv7BM7B
20151216 17:49:21-!- celmin|sleep is now known as celticminstrel
20151216 17:50:07< gfgtdf> celticminstrel: you re sure that [args] i supported in non-actionswml lua ?
20151216 17:50:43< celticminstrel> If it's not, I would call that a bug.
20151216 17:51:03< zookeeper> at least that snippet doesn't work in this context
20151216 17:51:16< celticminstrel> I see.
20151216 17:53:01< zookeeper> but yeah i had completely missed that tag
20151216 17:54:58< gfgtdf> celticminstrel: y and i thought i already filed a bugerport for that but appearantly i didnt
20151216 17:55:57-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20151216 17:58:05-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has joined #wesnoth-dev
20151216 18:08:16-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has quit [Remote host closed the connection]
20151216 18:11:37-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151216 18:14:59-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 255 seconds]
20151216 18:15:00-!- wedge010 is now known as wedge009
20151216 18:15:43-!- irker809 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20151216 18:15:43< irker809> wesnoth: Charles Dang wesnoth:master aad6ce6380da / data/campaigns/Northern_Rebirth/scenarios/10a_Stolen_Gold.cfg: NR S10a: code cleanup http://git.io/v01p7
20151216 18:15:44< irker809> wesnoth: Charles Dang wesnoth:master 407cc365f235 / data/campaigns/Northern_Rebirth/scenarios/11a_The_Eastern_Flank.cfg: NR S11a: code cleanup http://git.io/v01p5
20151216 18:15:45-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20151216 18:18:47-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has joined #wesnoth-dev
20151216 18:20:00< shadowm> "Canned definition of the Leadership ability to be included in an [abilities] clause. Note: Works only on units of level 1-5 "
20151216 18:20:03< shadowm> What. Why?
20151216 18:21:56< shadowm> Using value=$($other_unit.level * 25) isn't an option I guess?
20151216 18:22:09< gfgtdf> shadowm: becasue it needs a [leadership] for each levle it supports
20151216 18:22:19< gfgtdf> shadowm: y that would be coodl but it doesnt work
20151216 18:22:25< gfgtdf> shadowm: (there is a fr about that)
20151216 18:22:36< shadowm> Okay, I see.
20151216 18:24:55< gfgtdf> shadowm: the previous impementation needed O(n^2) [leadership] tags to cover n levels
20151216 18:24:56< irker809> wesnoth: Ignacio R. Morelle wesnoth:master 3821b7c416e2 / data/tools/Makefile: Remove macro-reference.html on clean target http://git.io/v0MfO
20151216 18:25:01< vultraz> shadowm: persistent sides can be any sides as long as they have the same save_id right?
20151216 18:25:10< shadowm> Yes.
20151216 18:25:30< vultraz> well then
20151216 18:25:32< vultraz> NR is stupid
20151216 18:25:32< gfgtdf> vultraz: you should put persisent=yes in there
20151216 18:25:44< gfgtdf> vultraz: which default to yes only for human sides
20151216 18:25:47< shadowm> NR isn't stupid, it's just special.
20151216 18:26:07< vultraz> NR has 4 empty sides here in order to make sure these two sides are always 7 and 9
20151216 18:26:43< shadowm> Oh, I don't know if the side numbers need to match.
20151216 18:26:51< shadowm> Wouldn't be surprised if that were actually the case.
20151216 18:26:56< gfgtdf> they don't need to match
20151216 18:27:14< gfgtdf> at least i know about 1.13, not 100% sure about 1.12
20151216 18:27:27< shadowm> I rather doubt anyone bothered to check that back when NR was written anyway (1.1.x).
20151216 18:27:33< vultraz> the comment is "# Empty sides - to ensure proper recall lists carryover"
20151216 18:27:42< vultraz> recall lists aren't tied to sides :|
20151216 18:27:45< shadowm> It may even have worked differently.
20151216 18:27:46< vultraz> side numbers*
20151216 18:28:06< celticminstrel> $other_unit only works within the filters, so it's not available in the value key.
20151216 18:29:27< gfgtdf> shadowm: here te fr i was talking about: http://gna.org/bugs/?23063
20151216 18:29:41< celticminstrel> I imagine that the side numbers did need to match at some point in the past.
20151216 18:30:44< gfgtdf> vultraz: i think its best to remove those side
20151216 18:30:46< gfgtdf> s
20151216 18:30:53< vultraz> yes
20151216 18:30:55< vultraz> I am
20151216 18:37:18-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has quit []
20151216 18:40:01< irker809> wesnoth: Charles Dang wesnoth:master 167fe399c3e1 / data/campaigns/Northern_Rebirth/scenarios/12a_Get_the_Gold.cfg: NR S12a: code cleanup http://git.io/v0Mtx
20151216 18:41:09-!- Kwandulin [~Miranda@p200300760F250AFF54817E69DD687679.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20151216 18:43:29-!- iceiceice [~chris@24.250.23.28] has joined #wesnoth-dev
20151216 18:43:29-!- iceiceice [~chris@24.250.23.28] has quit [Changing host]
20151216 18:43:29-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20151216 18:47:20< shadowm> gfgtdf: Reminding you again to assign bugs to yourself when fixing them.
20151216 18:47:46< shadowm> (#24100 was fixed quite a while after my last reminder about this.)
20151216 18:47:51< gfgtdf> vultraz: dont forget to also update all other references to that sides
20151216 18:48:10< vultraz> oshit
20151216 18:48:22< gfgtdf> vultraz: i for exampel see a [store_starting_location] side=9 in that file
20151216 18:49:20< irker809> wesnoth: Charles Dang wesnoth:master 80ce7fb1ce9e / data/campaigns/Northern_Rebirth/scenarios/12a_Get_the_Gold.cfg: SR S12a: fixup sides references changed in 167fe399c3e1 http://git.io/v0MOA
20151216 18:49:50< gfgtdf> vultraz: maybe its even easier to make macros #define KRASH_SIDE 4 #enddef idk
20151216 18:50:04< vultraz> maybe
20151216 18:50:15< vultraz> I was just doing some superficial cleanup right now 
20151216 18:51:12< gfgtdf> vultraz: y that was just a thought in cast its too confusing othwewise. I think in that case the new dside number are distainct to the odl side number so it be that confising
20151216 18:51:25< gfgtdf> vultraz: also dont foget to update starting locations in the map
20151216 18:51:40< gfgtdf> although im not sure whether that applies to that scenario
20151216 18:53:19< irker809> wesnoth: Charles Dang wesnoth:master 4cfae8daef3d / data/campaigns/Northern_Rebirth/maps/12a_Get_the_Gold.map: NR S12a: updated starting locations http://git.io/v0Msr
20151216 18:55:09< shadowm> gfgtdf: Is Invalid really the correct status for https://gna.org/bugs/index.php?23686 ?
20151216 18:55:16< zookeeper> strange. if a unit is given a new standing animation which takes precedence over the old one, it won't be played until the unit is selected or something else happens to it.
20151216 18:55:40< shadowm> The code isn't requesting the current animation to be stopped probably.
20151216 18:55:54< zookeeper> yeah
20151216 18:56:12< shadowm> Is this with [effect]?
20151216 18:56:17< gfgtdf> shadowm: iirc this was an error in his wml code.
20151216 18:56:22< zookeeper> shadowm, yes
20151216 18:56:50< shadowm> Someone ought to make it a special case to check for modifications to animations that might be playing in the current context and restart them if it's the case.
20151216 18:57:05< gfgtdf> shadowm: i changes the default value of [set_global_variable] side= which was previously a require attribute
20151216 18:57:11< shadowm> (We wouldn't want to restart a unit's standing animation if we're just adding an attack animation, for example.)
20151216 18:57:11< gfgtdf> shadowm: to match the most common case
20151216 18:58:25< gfgtdf> shadowm: his eeror was trhat he wwrote side=1 instead of side=global
20151216 19:01:21< shadowm> Okay, closed all bugs handled in 1.12.5 or 1.13.1.
20151216 19:01:37< shadowm> That took unexpectedly long.
20151216 19:01:55< vultraz> And spammed my inbox
20151216 19:02:30< shadowm> Just as intended.
20151216 19:02:50< shadowm> *1.13.2
20151216 19:04:40-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20151216 19:24:24< gfgtdf> vultraz: http://gna.org/bugs/?24185
20151216 19:24:48< vultraz> yes I know
20151216 19:24:51< vultraz> I can't figure out why
20151216 19:25:02< vultraz> the filter code is the same as game load
20151216 19:25:04< vultraz> which doesn't crash
20151216 19:25:41< gfgtdf> vultraz: well usually the casue is that you at some place assume that there is at lest one entry maching the filter
20151216 19:25:49< gfgtdf> vultraz: you have a stacktrace of the crash?
20151216 19:31:15< gfgtdf> vultraz: i just loked at the code and its not the same as with game load
20151216 19:31:49< vultraz> huh?
20151216 19:33:30< gfgtdf> vultraz: https://github.com/wesnoth/wesnoth/blob/master/src/gui/dialogs/game_load.cpp#L266
20151216 19:37:28< vultraz> oh, hm...
20151216 19:56:46-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 240 seconds]
20151216 19:58:50< shadowm> wedge009: https://gna.org/bugs/?24186
20151216 19:59:03-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20151216 19:59:06-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Remote host closed the connection]
20151216 19:59:19< shadowm> Great timing there.
20151216 19:59:39< shadowm> loonycyborg: https://gna.org/bugs/index.php?24187
20151216 19:59:43-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev
20151216 20:00:04< shadowm> wedge009: https://gna.org/bugs/?24186
20151216 20:00:10< irker809> wesnoth: Charles Dang wesnoth:master 8903c454c082 / src/gui/dialogs/unit_create.cpp: tunit_create: fix crash with non-matching filter (bug #24185) http://git.io/v0MDR
20151216 20:00:13< irker809> wesnoth: Charles Dang wesnoth:master af7fb252173a / src/help/help.cpp: Further increase size of help browser http://git.io/v0MD0
20151216 20:00:16< irker809> wesnoth: Charles Dang wesnoth:master 75d125d0707c / src/help/help_topic_generators.cpp: Use full-size portraits in unit help pages (experimental) http://git.io/v0MDE
20151216 20:01:36< vultraz> gfgtdf: seems I didn't need that preview_pane bit
20151216 20:01:47< vultraz> celticminstrel: opinions on use of full-sized portraits?
20151216 20:02:44< celticminstrel> Seems like a dubious move.
20151216 20:03:00< vultraz> well, take a look
20151216 20:03:21< celticminstrel> ...I guess that means I need to compile Wesnoth.
20151216 20:03:41< celticminstrel> It'll probably take awhile.
20151216 20:03:59< vultraz> it looks best when the portraits don't have obvious borders
20151216 20:04:03< vultraz> but i can't really help that
20151216 20:04:33< vultraz> Troll Shaman, for example, looks great
20151216 20:04:37< vultraz> Dragoon, not so much :P
20151216 20:04:44 * shadowm facepalms.
20151216 20:05:44< celticminstrel> Based on what others have said in here, I think it was probably a bad move, but I guess I'll take a look.
20151216 20:06:23< shadowm> Bad doesn't really begin to describe it.
20151216 20:07:20< shadowm> Use -r 800x600 to see what I mean.
20151216 20:07:45-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20151216 20:07:54< celticminstrel> I've never changed the resolution from the default, so I think it's already set to that.
20151216 20:08:07< shadowm> Also make sure to use non-high DPI hardware.
20151216 20:08:21< celticminstrel> On the dev copy at least. The stable copy I've made a lot bigger.
20151216 20:08:27< shadowm> Otherwise 800x600 means something else entirely IIRC or something. Stupid Macs.
20151216 20:08:29< celticminstrel> If you mean retina monitors, I don't have one.
20151216 20:08:41< vultraz> I see
20151216 20:08:55< shadowm> vultraz seems to be naïvely assuming all big portraits are the same size as well.
20151216 20:09:03< vultraz> I know they aren't
20151216 20:09:22< shadowm> Oh, also, it can crash if it runs out of space for the text.
20151216 20:09:23< shadowm>   Dev message: Condition 'max_width > 0' failed at src/marked-up_text.cpp:439 in function 'word_wrap_text'.
20151216 20:09:33< celticminstrel> Is that <img> tag Pango markup?
20151216 20:09:41< shadowm> Not really sure which (mainline) unit it was, I think it was an elf.
20151216 20:09:43< shadowm> No.
20151216 20:09:52< shadowm> The help system does not use Pango markup.
20151216 20:09:59< vultraz> it's custom gui1 crapmarkup
20151216 20:10:07< shadowm> It's not GUI1 markup either.
20151216 20:10:29< celticminstrel> The <img> tag is almost HTML but not quite.
20151216 20:10:38< shadowm> Ah, it's the Elvish Shyde.
20151216 20:26:24< vultraz> I guess I have to do something about that
20151216 20:26:32< celticminstrel> Eh? Why do I have an error about missing fontconfig.h...
20151216 20:51:15< irker809> wesnoth: Ignacio R. Morelle website:master c9b34ddc4b00 / start/1.12/styles/westin.css: Make page translations list wrap instead of overflowing the footer (bug #23756) http://git.io/v0DUk
20151216 21:01:05-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 260 seconds]
20151216 21:03:57< shadowm> aquileia, gfgtdf: Re PR #558, I insist that a WML file is a better idea.
20151216 21:04:50< shadowm> The file would just contain information like the version numbers, download URLs, checksums, and whatever else the updater needs to know which is liable to change over time.
20151216 21:04:57< gfgtdf> shadowm: hm iirc you said the opposite when i asked you last time but maybe i miszunderstood.
20151216 21:05:28< shadowm> Pretty sure I suggested WML because (paraphrasing) "we use WML like others would use JSON".
20151216 21:06:01< shadowm> It'd be maintained by the release team, obviously, like everything else in the website regarding releases.
20151216 21:07:05< gfgtdf> shadowm: i thought only javascuipt applications use json?
20151216 21:07:42< gfgtdf> shadowm: well for the cient side, i am not 100% sure which woudl be easier
20151216 21:07:58< shadowm> JSON was born from a Javascript idiom but it's by no means unique to Javascript applications.
20151216 21:08:12< gfgtdf> shadowm: maybe it can just use teh netowrk_asio socket class
20151216 21:08:16< gfgtdf> shadowm: then it woudl be easy
20151216 21:08:49< shadowm> Um, well, no, because that expects to talk a WML server, doesn't it?
20151216 21:09:20< gfgtdf> shadowm: hm i thought thats wat you wanted?
20151216 21:09:21< shadowm> I'd rather use piggyback on HTTP than have yet another WML server to maintain.
20151216 21:09:50< shadowm> Okay, let me clarify. The release information would be provided by the server in a WML file served via HTTP.
20151216 21:10:02< shadowm> The HTTP server.
20151216 21:10:47< shadowm> Then Wesnoth downloads and parses that like it would any other plain WML file (i.e. without invoking the full preprocessor pipeline) into a config.
20151216 21:12:38< gfgtdf> shadowm: i nevr messed much with the wml parser but it its just calling read(config&, std::istream&) then it shoudl be easy
20151216 21:16:20< shadowm>     filesystem::scoped_istream stream = filesystem::istream_file(pbl_path);
20151216 21:16:24< shadowm>     read(cfg, *stream);
20151216 21:16:51< shadowm> From get_addon_pbl_info() (src/addon/manager.cpp:92). Throws config::error on errors.
20151216 21:19:59< shadowm> I was going to propose a syntax for the file but then realized that I don't really know what the updater's needs are.
20151216 21:20:21< gfgtdf> shadowm: y, pr #558 just handles the wesnoth.exe client side part (downloading the http) to make the updater it ofc needs also teh server and the updater.exe na di actually think that part that i wrote so far is the easiert part.
20151216 21:20:30< shadowm> A good starting point is being able to specify the base URL for files, since it's not guaranteed we'll still be using SF.net three years from now.
20151216 21:21:25< gfgtdf> shadowm: also note that the download currently doesnt use https or similar
20151216 21:21:37< gfgtdf> shadowm: i personally this that shoudl done after the basib updater works
20151216 21:21:46< shadowm> Well, yeah, that's rather unfortunate but our official links don't use encryption either.
20151216 21:22:23< gfgtdf> shadowm: you mena wesnoth.org homepage or the installer download?
20151216 21:22:58< shadowm> The download links in the front page and wiki an release announcements.
20151216 21:23:24< shadowm> In fact, even if they did use encryption, it seems SF.net's ad page will redirect you to a non-encrypted download anyway.
20151216 21:24:18< shadowm> So yeah, https is pointless with SF.net downloads beause they are served from hosts that don't support https. :|
20151216 21:24:25< gfgtdf> shadowm: that code in #558 isnt teh one that shoudl download the file if i uzndersttod corrently
20151216 21:24:44< gfgtdf> shadowm: its the onyl that doanlods the link to the updater package and the chacksum
20151216 21:24:57< gfgtdf> shadowm: (the wml docuemnt)
20151216 21:25:18< gfgtdf> shadowm: and if that is hosted on our server i see no reason why that shodnt be encrypted
20151216 21:25:20< shadowm> Yes, I noticed it's a micro HTTP client.
20151216 21:25:38< shadowm> https is pointless for wesnoth.org too, because we only support https on the main vhost.
20151216 21:26:22< shadowm> i.e. https://wesnoth.org/ works as expected, but https://<ANYTHING>.wesnoth.org/ will give you https://wesnoth.org/ and break.
20151216 21:26:42< shadowm> (Not my responsibility or area of competence, I'm afraid.)
20151216 21:27:09< gfgtdf> shadowm: whose else coudl it be?
20151216 21:27:36< shadowm> files.wesnoth.org is one of the few vhosts for which this can be worked around by using the actual doc root path relative to the main vhost, in this case, https://wesnoth.org/files/.
20151216 21:28:28< shadowm> But wiki.wesnoth.org and forums.wesnoth.org can't be used like that and will force you to use the vhosts.
20151216 21:30:25< shadowm> gfgtdf: Well, it's complicated. I'm in charge of the high level aspect to this issue (vhosts configuration), but I can't deal with the https set-up; that's purportedly Rhonda's task and last time we discussed this we agreed that there was a viable solution. Only problem is that the solution in question would leave Windows XP MSIE users (as well as a few other non-Windows legacy operating systems) out in the cold.
20151216 21:30:37< gfgtdf> shadowm: maybe we coudl encrypt the update packages manually with some private key where the publibk ke is harcoded in wesnot to make sure the update pagages won't be maniulated but i dont really know mch about this stuff so i cannot say.
20151216 21:30:49< shadowm> But she hasn't had time to look into the solution, so we still have the broken https set-up.
20151216 21:32:37< shadowm> gfgtdf: I think validating downloads against a list of file hashes downloaded from https://wesnoth.org/files/ (instead of http://files.wesnoth.org/) would be good enough.
20151216 21:32:56< shadowm> The file hashes would probably be in the same file with the URLs and versions.
20151216 21:33:47< shadowm> As long as wesnoth.org's authenticity can be verified everything else can be assumed safe.
20151216 21:35:02< shadowm> (Of course, wesnoth.org's security can be compromised without invalidating its https certificate and then there's absolutely nothing to stop a malicious party from pushing malware to users, but let's not get too paranoid here.)
20151216 21:36:37< Rhonda> shadowm, gfgtdf: I'll be looking into SNI and implement it over the christmas break.
20151216 21:36:47< Rhonda> This looks like a good howto for it: https://www.digicert.com/ssl-support/apache-multiple-ssl-certificates-using-sni.htm
20151216 21:37:00< Rhonda> And: I plan to use letsencrypt for the ssl cert if that's fine with everyone?
20151216 21:38:56< shadowm> Christmas break? I'd rather have you people spending that time with family and friends than doing unpaid labor. :p
20151216 21:39:12< shadowm> (Especially considering I won't be around.)
20151216 21:40:18< Rhonda> One of my partners does work, the other potential partner is spending their time with their own primary partner, and my kid is with their grand parents.
20151216 21:40:37< Rhonda> … and my company does close for the last week of the year.
20151216 21:41:33< Rhonda> And I have to work on the website for our co-housing project anyway.
20151216 21:42:03< shadowm> Oh well.
20151216 21:42:22 * Rhonda . o O ( too much information?  you get that with me if you ask :) )
20151216 21:43:25< shadowm> I'll try to be around in case you need help or ask me questions about our absurdly convoluted set-up, but I won't be able to guarantee immediate responses I'm afriad.
20151216 21:43:31< shadowm> Afraid even.
20151216 21:45:19< Rhonda> At some point I would like to upgrade the host.  Not sure if the virtualization tools are up to pace though?
20151216 21:46:13< shadowm> Upgrading the host would be great, considering that systemd requires a newer kernel and figuring out how to not pull systemd is the only thing that's kept me from upgrading our containers to jessie.
20151216 21:47:00< shadowm> The thing is I think PVE published a new feature release recently but IIRC it removes support for OpenVZ containers, which would be a problem for us.
20151216 21:48:40< shadowm> Proxmox VE 4.0, drops OpenVZ in favor of LXC (OpenVZ is not available for Kernels above 2.6.32).
20151216 21:48:46< Rhonda> There are conversion tools for all these container things I thought?
20151216 21:48:59< shadowm> Yes, they have a guide for it on their site: http://pve.proxmox.com/wiki/Convert_OpenVZ_to_LXC
20151216 21:49:31< shadowm> It's still going to be a bit risky considering we don't have off-site backups atm.
20151216 21:49:42< Rhonda> Oh, openvz to lxc is not that big of an issue.
20151216 21:50:06< Rhonda> Maybe we should do that first then.  How much space requirement would that backup need?
20151216 21:52:32< shadowm> I don't remember where to check. There should be a specific dir on the host with the compressed OpenVZ ct backups.
20151216 21:54:33< shadowm> Oh, backup-weekly should be a good measurement since it encompasses all containers sans 199.
20151216 21:55:28< shadowm> Rhonda: About 110 GiB for one weekly backup.
20151216 21:56:10< Rhonda> Ok, I guess I can find temporary space for that.
20151216 21:56:13< shadowm> Worth noting that's a set of LZO-compressed tarballs.
20151216 21:57:16< shadowm> It also excludes some files, most (but not all) of which that we can recreate easily.
20151216 22:01:45< shadowm> The excluded files make up 2.5 GiB (daily database dumps I keep just in case because restoring from container backups isn't the fastest thing to do) + 11 GiB (private add-ons server archives).
20151216 22:02:41< gfgtdf> vultaz: i think the new loadgame dialog might have broken replays
20151216 22:03:31< shadowm> Rhonda: But more ideally we should have a system for doing incremental off-site backups periodically to a safe, permanent host. If someone could work on setting up that it'd just be a matter of asking Dave to finance the hosting.
20151216 22:04:27< shadowm> (Incremental as opposed to what we do right now locally, dumping the entire filesystem of every container into archives and wasting like 12 hours a day on this.)
20151216 22:05:15< shadowm> (Especially considering that in most cases, 80% of the filesystem remains unchanged every time because it's the OS.)
20151216 22:05:46< shadowm> Anyway, I must leave now to acquire water and nutrients.
20151216 22:07:58< loonycyborg> shadowm: I'll likely need help from actual artists to update that installer graphic
20151216 22:08:30< shadowm> Then go ask them in Art Contributions.
20151216 22:09:36< shadowm> (Oh also, my figures don't take into account any miscellaneous non-recreatable stuff on the PVE host such as the DNS config, which hopefully is all under 5 MiB.)
20151216 22:13:01-!- gfgtdf [~chatzilla@x50abe5dd.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 42.0/20151029151421]]
20151216 22:14:05< shadowm> loonycyborg: Do the files actually have to be in the Windows bitmap format as opposed to e.g. PNG?
20151216 22:15:07< loonycyborg> need to check docs
20151216 22:17:09< shadowm> loonycyborg: https://dl.dropboxusercontent.com/u/21371130/wesnoth-logo-55-o.png  Replaces windows_installer_w_881.bmp, leaving any required conversions up to you.
20151216 22:17:21< shadowm> I don't even know where that's actually used.
20151216 22:18:01< shadowm> I think I could actually update the side banner too unless you want something different.
20151216 22:18:21< loonycyborg> I don't have particular preference
20151216 22:18:34< loonycyborg> I reused banner from previous installer
20151216 22:18:41< loonycyborg> and it's fine for me as it is
20151216 22:18:43< shadowm> I'll ask vultraz if he wants something different then.
20151216 22:18:47< loonycyborg> ok
20151216 22:22:00< shadowm> Rhonda: Oh, and the most important part, we need to have a downtime window set in advance for a host upgrade so no-one panics and assumes the project finally died or something. I imagine it'll take quite long to upgrade the base system even in the best case.
20151216 22:22:07< shadowm> (Okay, off now for real.)
20151216 23:11:29-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Remote host closed the connection]
20151216 23:11:56-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20151216 23:17:17-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 265 seconds]
20151216 23:19:16-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20151216 23:25:56-!- horrowind [~Icedove@2a02:810a:8b00:1c54:21b:fcff:fee3:c3ff] has quit [Ping timeout: 255 seconds]
20151216 23:32:33-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20151216 23:44:22-!- Askarii [~Askarii@206-248-175-248.dsl.teksavvy.com] has joined #wesnoth-dev
20151216 23:47:10< Askarii> might get burned for saying this... is there any way a JavaScript dev. could lend a hand to the project?
20151216 23:51:39-!- irker809 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20151216 23:59:06< zookeeper> Askarii, maybe for some web-related things? i'm not sure if we currently use any anywhere.
20151216 23:59:23< Askarii> hi zookeeper 
20151216 23:59:38< Askarii> My first thought was about any web assets related to the project
--- Log closed Thu Dec 17 00:00:15 2015