--- Log opened Tue Feb 23 00:00:08 2016
--- Day changed Tue Feb 23 2016
20160223 00:00:07< vultraz> ok, I need to use name= then
20160223 00:00:16< zookeeper> id
20160223 00:01:22< zookeeper> gfgtdf, i don't know whether there's some new glitchy editor dialogs which do that, but in-game it doesn't happen
20160223 00:01:59< gfgtdf> zookeeper: hmm maybe it onlyhappen to gui2 dialog in 1.12 afaik the loadgame and createunit dialogs are gui1 in 1.12
20160223 00:02:00< zookeeper> and i hope this new behavior is not supposed to stay, because it's awful
20160223 00:02:21< irker806> wesnoth: Charles Dang wesnoth:master 2d9996093010 / data/campaigns/Under_the_Burning_Suns/scenarios/02_Across_the_Harsh_Sands.cfg: UtBS S2: fixed Go’hag not being included in undead group death event https://github.com/wesnoth/wesnoth/commit/2d99960930106464b3cd12517fe7fcd57d10d4df
20160223 00:03:31< zookeeper> vultraz, you forgot the translation marker
20160223 00:04:14< zookeeper> when using a translation, it'd now try to match the translated name with the untranslated one and possibly fail
20160223 00:04:27< zookeeper> as far as i know anyway
20160223 00:04:52< irker806> wesnoth: Charles Dang wesnoth:master 7f02961ddaf4 / data/campaigns/Under_the_Burning_Suns/scenarios/02_Across_the_Harsh_Sands.cfg: Fixup 2d9996093010 https://github.com/wesnoth/wesnoth/commit/7f02961ddaf467a29882a4acd32aa55201188ad5
20160223 00:05:04< zookeeper> perfect
20160223 00:06:24< vultraz> Note that I didn't go into the scenario to check
20160223 00:06:27< vultraz> but it should work fine
20160223 00:06:57< zookeeper> yeah. it'll go through testing and fiddling anyway before 1.14...
20160223 00:07:21 * zookeeper goes through sleeping
20160223 00:08:13< celticminstrel> Huh, I didn't think name= was a valid filter key, but the source says that it is.
20160223 00:11:47-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 248 seconds]
20160223 00:15:52-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160223 00:25:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160223 00:27:45-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160223 00:29:38< gfgtdf> celticminstrel: for units ?
20160223 00:34:37-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 00:34:37< celticminstrel> Yeah.
20160223 00:36:10< celticminstrel> Translations can completely replace the Markov namelists, right?
20160223 00:36:58< celticminstrel> So really, what's the use of filtering on name=?
20160223 00:38:13< celticminstrel> It's undocumented, too.
20160223 00:39:10< celticminstrel> Any name= filter can equally well be done as an id= filter.
20160223 00:50:01< gfgtdf> celticminstrel: y i thinkwe shoul remove it
20160223 00:50:12< gfgtdf> celticminstrel: let me chack the commit that introduces it before we do that
20160223 00:50:31-!- EliDupree [~quassel@idupree.com] has quit [Remote host closed the connection]
20160223 00:50:49-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20160223 00:51:30< gfgtdf> celticminstrel: y i thinkwe can remove it
20160223 00:54:23< irker806> wesnoth: gfgtdf wesnoth:master a5d03e951995 / src/unit_filter.cpp: remove support for name= in standart unit filters https://github.com/wesnoth/wesnoth/commit/a5d03e9519954bf09ca17f52fe51438146bafdb7
20160223 00:54:25< gfgtdf> celticminstrel: ^
20160223 00:54:48< celticminstrel> gfgtdf: You forgot to remove the use of it that vultraz just added.
20160223 00:58:29< irker806> wesnoth: gfgtdf wesnoth:master 89e8341b1aa1 / data/campaigns/Under_the_Burning_Suns/scenarios/02_Across_the_Harsh_Sands.cfg: fix use of invalid name= attribute in standard unit filtes in UtbS scenario2 https://github.com/wesnoth/wesnoth/commit/89e8341b1aa1ff72a05c014c1136b481df9479cd
20160223 00:58:37< gfgtdf> celticminstrel, vultraz: ^
20160223 00:59:14-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160223 01:04:51-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160223 01:05:13-!- louis94 [~~louis94@91.178.242.51] has quit [Quit: Konversation terminated!]
20160223 01:06:31< celticminstrel> Might want to double-check that no other campaigns/scenarios use it, but I know no good way to do that, since name= is a valid attribute in many contexts.
20160223 01:19:02-!- mjs-de [~mjs-de@f049131231.adsl.alicedsl.de] has quit [Remote host closed the connection]
20160223 01:47:35-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160223 02:11:21-!- un214 [~un214@2602:306:cccf:b439:56a0:50ff:fe57:101d] has joined #wesnoth-dev
20160223 02:20:44< mattsc> Ugh — writing documentation is always so much work.
20160223 02:20:59< mattsc> Especially when you have to undo 5 years of patchwork.
20160223 02:22:44< celticminstrel> HI.
20160223 02:27:35-!- prkc [~prkc@46.166.188.220] has quit [Remote host closed the connection]
20160223 02:28:17< mattsc> Hello
20160223 02:43:55< aeth> In case anyone is interested, assuming a maximum of 10 year desktop hardware support, the earliest point to switch to OpenGL 3.3 exclusively would be the very end of 2016 because the earliest cards that max out on OpenGL are late 2006. For Vulkan, it would be March to April 2022 depending on if you want to count discrete or discrete and Intel integrated.
20160223 02:44:21< aeth> s/OpenGL are/OpenGL 3 are/
20160223 02:45:21< aeth> You could round to January 2017 because the 3.3 cards were very late in 2006. (Integrated graphics don't matter for OpenGL 3.3 because there's software emulation for 3.3 in mesa iirc, based on an earlier conversation here.)
20160223 02:46:44< aeth> So, basically, OpenGL 3.3 isn't really an issue because by the time a port is finished, it'd probably be at least 2017.
20160223 02:47:16< aeth> My assumptions were also moving over completely, but if someone were to put in the work you could have e.g. OpenGL 3.3 and Vulkan in parallel.
20160223 02:50:39< aeth> There's only one generation on OpenGL 4.x and not on Vulkan, so targeting 4.x exclusively wouldn't make any sense except to save work, but by 2020 (going with the 10 years assumption) there are probably going to be enough libraries to make working on Vulkan easier than OpenGL 4.x
20160223 02:51:17< celticminstrel> I haven't the slightest clue what Vulkan is.
20160223 02:52:32< aeth> celticminstrel: Vulkan is basically a much lower level graphics API than OpenGL, by the same people as OpenGL. The main advantages are that it has lower driver overhead, better CPU performance, and the same API on mobile as on desktop. The main disadvantage (which will probably only be until the right libraries appear) is that it is considerably harder to use. Another disadvantage is that it only supports hardware from 2012 onward.
20160223 02:53:15< celticminstrel> ...sounds terrible.
20160223 02:53:38< aeth> There's basically no need for Wesnoth to adopt Vulkan, except perhaps on the Android port if it exists. It's not really in serious consideration, I just put it there for comparison.
20160223 02:53:42-!- gfgtdf_ [~chatzilla@f054060000.adsl.alicedsl.de] has joined #wesnoth-dev
20160223 02:53:52< celticminstrel> OpenGL is already almost incomprehensible. Why would you want to go even more low-level?
20160223 02:54:22< aeth> celticminstrel: OpenGL is incomprehensible because there's a lot of driver magic going on. More low-level means that the equivalent of most of OpenGL would be done in open source libraries whose code you could inspect. Less magic.
20160223 02:54:30< celticminstrel> But that's just my opinion and unrelated to Wesnoth specifically.
20160223 02:54:41< aeth> Also, the performance is considerably better. Considerably. Not like it matters for a 2D game.
20160223 02:55:01< celticminstrel> I'm not sure that's the real reason it's incomprehensible, but whatever.
20160223 02:55:31-!- gfgtdf [~chatzilla@f050179066.adsl.alicedsl.de] has quit [Ping timeout: 252 seconds]
20160223 02:55:33< aeth> Well, the other reason is that graphics cards have changed a lot. So the OpenGL model no longer makes any sense on modern GPUs, so OpenGL best practices seem to make no sense.
20160223 02:55:37-!- gfgtdf_ is now known as gfgtdf
20160223 02:56:21< celticminstrel> Well, I wouldn't have anything against using Vulkan if it was wrapped in some high-level library, like SDL or whatever.
20160223 02:56:30< aeth> What will probably happen is that a combination of Vulkan libraries and useful Vulkan features ported to OpenGL extensions (e.g. SPIR-V) will make a transition to Vulkan in 4-8 years a good thing.
20160223 02:56:43< aeth> And easier.
20160223 02:56:47< aeth> Right now, no need.
20160223 02:57:19-!- ancestral [~ancestral@63.92.240.233] has joined #wesnoth-dev
20160223 02:57:29 * celticminstrel forgot to test that thing for ancestral 
20160223 02:57:52< irker806> wesnoth: Celtic Minstrel wesnoth:master 43e3f8e63164 / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: XCode: Split up copy files phase a little https://github.com/wesnoth/wesnoth/commit/43e3f8e6316462fe1f3d837ad63d073ac0ca62a9
20160223 02:57:56< ancestral> Well, I think it fails for gfgtdf anyway
20160223 02:58:11< gfgtdf> ancestral: ?
20160223 02:58:14< celticminstrel> I'm assuming you mean a replay save.
20160223 02:58:21< ancestral> celticminstrel said he didn’t test the bug
20160223 02:58:22< ancestral> Yes
20160223 02:58:22< ancestral> http://gna.org/bugs/?24439
20160223 02:58:26< ancestral> So no worries
20160223 02:58:28< celticminstrel> Choosing "Show Replay" for a normal save causes no problems.
20160223 02:58:36< ancestral> Normal save?
20160223 02:58:42< ancestral> Single player, you mean?
20160223 02:58:47< celticminstrel> (Apart from missing [item] like I mentioned awhile back.)
20160223 02:59:00< aeth> celticminstrel: What I would personally do, just using rough numbers, is on a new game engine only support back 5 years (which means Vulkan exclusives make sense in April next year) and in a newer game engine aim for at least 10 (which is where I get 2022 for Wesnoth)
20160223 02:59:05< ancestral> I’m afraid I have to go, but I’ll likely be on again later
20160223 02:59:05< celticminstrel> The one I tried on was an MP scenario.
20160223 02:59:46< ancestral> aeth: “Wesnoth 2022” sounds like a modern X-COM-like spin-off
20160223 02:59:50-!- ancestral [~ancestral@63.92.240.233] has quit [Client Quit]
20160223 03:00:40< aeth> s/a newer game engine/an older game engine/
20160223 03:02:10< aeth> Wesnoth's OpenGL port should be using 3.3 as the baseline for the desktop. Mobile might be an interesting area to introduce experimental Vulkan support in a few years because the hardware lifespan is shorter.
20160223 03:03:28< aeth> The most interesting things, if I can get it to OpenGL 3.3 to work in Wesnoth at all, will be lava and water shaders, as well as a 2D lighting engine.
20160223 03:04:45< gfgtdf> celticminstrel: i thought i fixes missing item ?
20160223 03:05:01< celticminstrel> Did you? I haven't checked.
20160223 03:05:29< gfgtdf> celticminstrel: you need to mkae  a replay from latest master, it wont work with older savefiles becasue the bug is 'in the savefiles'
20160223 03:05:49-!- un214 [~un214@2602:306:cccf:b439:56a0:50ff:fe57:101d] has quit [Remote host closed the connection]
20160223 03:08:15< celticminstrel> Okay, maybe I'll try that at some point.
20160223 03:18:31< vultraz> aeth: why 10 years?
20160223 03:23:51< vultraz> hmm
20160223 03:23:57 * vultraz saw your arguments
20160223 03:24:01< vultraz> I'd rather we only go back 5
20160223 03:24:05< aeth> vultraz: 10 years is the maximum level of reasonable support in computers, usually.
20160223 03:24:19< aeth> Going back 5 is obviously way easier
20160223 03:24:29< vultraz> Easier is better
20160223 03:24:33< celticminstrel> My computer is ten years old.
20160223 03:24:44< aeth> celticminstrel: then it can probably support OpenGL 3.3
20160223 03:24:59-!- shadowm [~ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 244 seconds]
20160223 03:25:19< aeth> and if it's not a laptop and doesn't support opengl 3.3, you can probably buy an old card that can support it at a very low price
20160223 03:25:30< celticminstrel> Nope.
20160223 03:25:41< vultraz> 3.3 sounds reasonable
20160223 03:25:51< vultraz> I just want to make sure we're not missing things by not going for 4.x
20160223 03:25:52< celticminstrel> It's virtually impossibly to buy a new card for my computer.
20160223 03:25:57< aeth> celticminstrel: what graphics?
20160223 03:26:10< celticminstrel> Though I wouldn't be surprised if it already supports it.
20160223 03:26:32< aeth> vultraz: In OpenGL, you can backport new featuers as extensions, so even if you use 3.3 you can take new stuff as extensions where supported to speed things up
20160223 03:26:39-!- shadowm [~ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20160223 03:26:45< celticminstrel> NVidia GeForce 7500 GT.
20160223 03:27:11< aeth> celticminstrel: nope, your card maxes out at 2.1
20160223 03:27:15< aeth> Officially, at least
20160223 03:27:20< aeth> https://en.wikipedia.org/wiki/GeForce_7_series
20160223 03:27:39< vultraz> well, that sucks for celmin :/
20160223 03:28:27< aeth> Afaik, modern OpenGL is impossible in OpenGL 2.1
20160223 03:28:49< celticminstrel> Pretty sure it does support shaders and stuff though.
20160223 03:29:27< aeth> Forward-compatible shaders are impossible in OpenGL 2.1, too. It's version 1.2 of GLSL according to Wikipedia, while Vulkan supports a minimum of 1.4
20160223 03:30:05< vultraz> aeth: generally, I think we should support as far back is feasible and simple. So if we happen to have 10 years (which is a long time), good. But if it's less, then oh well.
20160223 03:30:33< vultraz> if 5 is easier, we should go for that
20160223 03:30:37< aeth> vultraz: OpenGL 3.3 happens to be as far back as feasible if you want to write it in modern (AZDO) style and with shaders that will also compile to SPIR-V (future-proof even in OpenGL, not just Vulkan)
20160223 03:31:04< aeth> So just under 10 years, probably a few months newer than celticminstrel's computer, happens to be the earliest feasible point by coincidence
20160223 03:31:05< vultraz> modern is good
20160223 03:31:13< celticminstrel> Perhaps by the end of the year I'll have a new Mac, anyway.
20160223 03:31:32< celticminstrel> ...probably not, but I can hope. >_>
20160223 03:31:58< aeth> Macs might never support Vulkan because Apple is pushing Metal, which is lower level than OpenGL and faster than OpenGL, but higher level than Vulkan and slower than Vulkan.
20160223 03:32:23< aeth> So if game engines start favoring Vulkan, you're going to have a hard time playing modern games from 2017 onward until Apple gives in, if ever.
20160223 03:32:28< celticminstrel> Which means that switching exclusively to Vulkan is probably not an option.
20160223 03:32:34< celticminstrel> For Wesnoth I mean.
20160223 03:33:33< aeth> Probably not until 2020-2022, which is sufficiently far enough away to make planning impossible except for perhaps a rule that GLSL shaders must be able to compile to Vulkan SPIR-V (future-proofing the shaders, at least)
20160223 03:34:11< celticminstrel> That's only if Apple supports Vulkan,
20160223 03:34:21< aeth> What could happen is Metal for OS X and iOS, Vulkan for newer desktops and Android, and OpenGL for legacy desktops. But in order to have legacy OpenGL, you'd have to first have OpenGL.
20160223 03:34:31< aeth> So OpenGL 3.3 is a minimum baseline for just about any future graphics planning.
20160223 03:34:39< vultraz> Alright
20160223 03:34:47< celticminstrel> gfgtdf: What is the purpose of commit 5deea2e and subsequent ones?
20160223 03:35:05< aeth> Metal for OS X would (1) double the work and (2) be slower than Vulkan on Windows and Linux, though. So it would be an inferior experience on OS X.
20160223 03:35:21< gfgtdf> celticminstrel: to remove the cfg_ member from unti class
20160223 03:35:30< celticminstrel> Hmm...
20160223 03:35:34< vultraz> aeth: what about Metal for OS X and DX12 for windows
20160223 03:36:10< aeth> vultraz: DX12 isn't worth it. It's almost identical to Vulkan in role and performance, but it isn't portable. There's literally no reason to use it. It would add work for no gain.
20160223 03:36:16< vultraz> (unless I'm misunderstanding, I thought Metal was basically Apple's DX)
20160223 03:36:50< aeth> Metal is considerably worse than either, but it is the only option for OS X and iOS so people will use it anyway.
20160223 03:37:19< aeth> For the games I write, I will simply just not support Apple platforms until they add Vulkan support. I doubt I'll be the only one.
20160223 03:37:31< vultraz> Metal is worse?
20160223 03:37:33< aeth> Yes.
20160223 03:37:37< vultraz> But they made it sound so good in the keynote :|
20160223 03:37:39< aeth> It's not as low level.
20160223 03:37:43< aeth> Ironically, given the name.
20160223 03:38:13< aeth> What will probably happen is a few generations of Android destroying iOS in gaming performance, and then Apple quietly giving in to the industry standard.
20160223 03:38:38< aeth> What could happen, though, is Apple just holding back Android 3D performance, though, since iOS is big enough that people might just code to the lowest common denominator in features.
20160223 03:39:03< aeth> It's also possible that Apple could give in on desktop, but not on iOS.
20160223 03:39:05< vultraz> I'd rather target iOS than Android if made to pick
20160223 03:40:24< vultraz> it's a better market anyway
20160223 03:40:53< aeth> And that's why Apple might just hold back 3D mobile gaming for years instead of being forced to give in. Not like it matters. Shovelware and poor app discovery is holding back mobile gaming more than anything else.
20160223 03:41:02< aeth> I have given up trying to find a mobile game that's playable.
20160223 03:41:15< aeth> I'm sure they exist, but it's not worth the effort, and if they exist they're probably pay to win crap anyway.
20160223 03:41:24< vultraz> Fallout Shelter?
20160223 03:41:38< vultraz> The Room?
20160223 03:41:59< vultraz> Fruit Ninja? 2048?
20160223 03:41:59< aeth> Never heard of either. (see: app discovery problem)
20160223 03:42:16< aeth> I have heard of quite a few pay to win crap games. It seems like all the shady games have moved from Facebook to mobile
20160223 03:42:27-!- gfgtdf [~chatzilla@f054060000.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 44.0.2/20160210153822]]
20160223 03:42:29< vultraz> well app discovery is hard with the amount of absolute shit there is 
20160223 03:42:38< vultraz> 95% of apps are shit
20160223 03:42:41< vultraz> and it's worse in Android 
20160223 03:42:44< aeth> Steam seems to be decent at it
20160223 03:42:54< aeth> Steam has quite a lot of shovelware too these days, but I rarely notice it.
20160223 03:43:10< vultraz> Android apps are very shit because Google Play doesn't have Apple's strict upload policies 
20160223 03:43:28< celticminstrel> git reset vs git checkout is super-annoying.
20160223 03:43:36< aeth> And Apple apps tend to be shit because of those strict upload policies holding back upgrades iirc
20160223 03:43:46< aeth> So mobile overall is shit, there's no happy middle ground like Steam.
20160223 03:44:16< vultraz> Mobil is a lot like 'here's a device to use this small selection of apps everyone uses' :P
20160223 03:44:29< vultraz> then again, show me a platform where that's not the case
20160223 03:45:09< aeth> 90% of my mobile usage (99%?) is Duolingo and Firefox with ublock origin. Add in an ssh app, and those 3 apps are pretty much the only apps I use.
20160223 03:45:25< aeth> The demise of desktop software was greatly overrated imo.
20160223 03:45:47< celticminstrel> I wonder if we should just deprecate lua_function and instead allow the use of [lua] tags in filters...
20160223 03:45:59< aeth> I only use Firefox on Android for ublock origin. It makes quite a performance difference on mobile.
20160223 03:46:31< vultraz> I guess it's not a secret that I greatly prefer iOS to Android
20160223 03:46:34< aeth> News sites are basically killing themselves by making adblocking the only way to load their sites on low end devices.
20160223 03:46:57< celticminstrel> Anyone?
20160223 03:47:00< vultraz> Ads are also less shitty on iOS
20160223 03:47:06< vultraz> I fucking hate ads on adroid 
20160223 03:47:09< vultraz> android*
20160223 03:47:14< vultraz> all that shovelware 
20160223 03:47:16< vultraz> celticminstrel: yes
20160223 03:47:17< vultraz> do that
20160223 03:47:34< aeth> All ads take up too much screen real estate. If your app has adds, I will probably uninstall it the same day that I installed it.
20160223 03:47:54< vultraz> well they're a necessary evil
20160223 03:48:06< aeth> To me, they're more of an evil evil.
20160223 03:49:22< vultraz> Also
20160223 03:49:29< vultraz> pay-to-win microtransaction shit
20160223 03:49:37< aeth> The way ads are implemented, there is no expectation of privacy anywhere where there is a digital ad.
20160223 03:50:09< aeth> Everyone would just cheat the ad system if ad networks didn't also spy on you.
20160223 03:50:12< vultraz> I don't care if I get targeted ads. I care if I get shitty ads.
20160223 03:50:47< vultraz> Like, when I play Relic Run (Tomb Raider mobile game by Square Enix), there's a nice ad for another Tomb Raider mobile game, Tomb Raider Go.
20160223 03:50:51< vultraz> That's a nice ad 
20160223 03:51:00< vultraz> (this is on iOS)
20160223 03:51:02< aeth> "Targeted" is a joke. Unless Facebook has changed a lot in the last few years (and they had far more info on me than Google), the extent of Facebook targeting was "male? single? dating site!"
20160223 03:51:11< vultraz> but then I got to android and play a random game I downloaded off the store
20160223 03:51:26< vultraz> and I get  "FREE WIFI EVERYWHERE"
20160223 03:52:48< aeth> vultraz: When people say they are against ads, they usually mean ad networks, not ads themselves. An "ad" for a related game or DLC while in the game itself doesn't really count imo.
20160223 03:52:58< aeth> Because it's probably built in, not using an ad network.
20160223 03:53:17< aeth> And it's quite often useful that way. If you like the game, why not try a sequel?
20160223 03:54:00< aeth> Even something like movie trailers at the start of a movie in the theater aren't that bad.
20160223 03:55:05< aeth> anyway, this is very off-topic probably.
20160223 03:55:48< aeth> Wesnoth is already on mobile, and probably ad-free there afaik, so there isn't really any productive conclusions that can be drawn from this discussion afaik.
20160223 03:56:25< aeth> And since Wesnoth isn't a mobile exclusive, and since Vulkan isn't exclusive to Android, there would also be no need to not support Vulkan on Android assuming such support could be implemented.
20160223 03:58:08< vultraz> Wesnoth is on mobile, yes
20160223 03:58:21< vultraz> But it's behind desktop (1.10, IIRC)
20160223 03:58:25< vultraz> and not done very well
20160223 03:58:42< vultraz> AFAIK it's basically like 'we managed to make the desktop version work on this small screen'
20160223 03:58:51< aeth> A Vulkan Wesnoth could mean an easier to manage codebase, at least between Android and desktop
20160223 03:58:56< vultraz> I'd like to make an official port for iOS and potentially android
20160223 03:59:14< vultraz> With dedicated mobile support for new features such as Force Touch in the iPhone 6s+
20160223 03:59:31< aeth> It would be considerably easier to do an official port for Android than for iOS because any work on Vulkan support in Android could be used in desktop too.
20160223 04:00:01< vultraz> But again, I think iOS is a better market to focus on :P
20160223 04:00:49< aeth> Which is exactly why iOS can pull shit like making a worse graphics API required over the better, vendor neutral one.
20160223 04:01:57< aeth> Maybe Windows Mobile is like that too with DirectX but no one cares about Windows Mobile in 2016
20160223 04:03:07< aeth> Although Apple did its best to hold OpenGL back, so it doesn't surprise me that Apple is now doing its best to hold Vulkan back.
20160223 04:05:24-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 04:10:11-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 248 seconds]
20160223 04:14:18< aeth> Apple should just give in to Vulkan because then there's only one graphics API all applications on all platforms will have to deal with, at least in the future as Vulkan gets easier with libraries etc. Until then, Apple just doubles the work, for no gain (unlike DirectX which was considerably better than OpenGL in the later days of OpenGL) and will only be able to get iOS exclusives out of it, not OS X exclusives.
20160223 04:14:55< aeth> If Apple never gives in, Wesnoth will never be able to retire OpenGL support in 2022 :-P
20160223 04:44:10-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160223 04:53:22-!- mikey_ [1812f9d8@gateway/web/freenode/ip.24.18.249.216] has joined #wesnoth-dev
20160223 04:56:30-!- mikey_ [1812f9d8@gateway/web/freenode/ip.24.18.249.216] has quit [Client Quit]
20160223 05:19:35-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 250 seconds]
20160223 05:30:59-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160223 05:41:10< ancestral> aeth: FYI any Mac that is capable of running the latest version of OS X supports at least OpenGL 3.3
20160223 05:41:22< ancestral> Basically, any Mac in the last 7 years, and some going as far back as 2007
20160223 05:42:06< celticminstrel> Maybe even 2006.
20160223 05:42:06< aeth> ancestral: which is good... because now with Vulkan we'll probably see 3.3 as the main target for OpenGL... only one generation of nvidias (GeForce 400s) can support OpenGL 4.x but not Vulkan
20160223 05:42:19< celticminstrel> But not mine, apparently.
20160223 05:42:26< aeth> And OpenGL 3.3 AZDO is fairly competitive with Vulkan right now afaik
20160223 05:43:30< aeth> Even Wesnoth one day (if it's still alive) will probably have two renderers, one for OpenGL 3.3 (legacy hardware) and one for Vulkan
20160223 05:44:10< aeth> Hardware acceleration will hopefully lead to some really nice terrain and spell effects
20160223 05:46:56< ancestral> (It wouldn’t surprise me if OS X 10.8 and newer uses some stuff from OpenGL 3.3 for things)
20160223 05:47:01< ancestral> 3.x
20160223 05:56:00< celticminstrel> Does [lua] support multiple [args] subtags or only a single one? I'd guess the latter, but not sure.
20160223 05:56:17< vultraz> I thought I saw an example with multiple somewhere
20160223 06:00:39< vultraz> tfw there are 18 people in the 1.12 lobby :|
20160223 06:01:22< vultraz> ah, and 21 in game
20160223 06:01:23< vultraz> ok
20160223 06:01:50< vultraz> so ~ 35 people
20160223 06:02:04< vultraz> (to adjust for changing numbers)
20160223 06:02:35< vultraz> Not gonna lie, this is kinda discouraging 
20160223 06:02:47 * celticminstrel has no idea what "tfw" means.
20160223 06:04:14< vultraz> the feeling when
20160223 06:04:23< vultraz> mrw is 'my reaction when'
20160223 06:06:56< vultraz> we should add mandatory account creation when launching wesnoth to track active player statistics 
20160223 06:07:36< celticminstrel> I use a different name on the forum and on the MP server (on those rare occasions when I actually visit the server).
20160223 06:08:14< vultraz> (only being semi-factious)
20160223 06:08:25< vultraz> we'll need account creation in the future
20160223 06:08:49< celticminstrel> I disagree that it's necessary, but don't have any problems with adding it.
20160223 06:09:23< vultraz> I want to bind addons to profiles
20160223 06:10:09 * celticminstrel can't remember how to load Lua code from a string.
20160223 06:12:23< celticminstrel> The server might be more active at other times of day, right?
20160223 06:12:44< celticminstrel> Try stats.wesnoth.org or whatever.
20160223 06:13:35< vultraz> I'd like to know how many people are playing wesnoth
20160223 06:13:37< vultraz> sp OR mp
20160223 06:13:57< celticminstrel> It's kinda impossible to track SP.
20160223 06:14:17< celticminstrel> Unless you add tracking features, but that would probably annoy some people.
20160223 06:14:23< aeth> just distribute Wesnoth on Steam and you can track people on Steam in SP
20160223 06:14:25< celticminstrel> Like me.
20160223 06:14:27< aeth> which will be most
20160223 06:14:42< celticminstrel> Sorta, sure.
20160223 06:14:54< aeth> celticminstrel: no just make the tracking feature part of nsa.c and people will understand it's a joke and not care
20160223 06:14:59< celticminstrel> But you can't really distinguish MP and SP players in that count.
20160223 06:15:10< aeth> celticminstrel: sure, SP = total - MP
20160223 06:15:14< aeth> and you can already track MP right?
20160223 06:15:21< celticminstrel> KInda, I guess.
20160223 06:15:27-!- Kwandulin [~Miranda@p200300760F0BC52A2D1B033D03745A0C.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160223 06:15:30< aeth> You'd miss people who are MP on their own servers, but that's probably close to 0%
20160223 06:15:47< aeth> and probably LAN more than true MP
20160223 06:16:28< celticminstrel> LAN is slightly different from running their own servers.
20160223 06:16:36< celticminstrel> Even though internally it's the same.
20160223 06:19:06< aeth> I've connected to the other servers in the client out of curiosity and they're always empty, maybe once there was one game in one
20160223 06:20:27< celticminstrel> I think they're mainly intended as backup servers in case the main one goes down for some reason.
20160223 06:21:04< aeth> Yes.
20160223 06:21:36< vultraz> maybe we should implement auto rerouting 
20160223 06:22:18< aeth> Does the lobby have matchmaking yet? So instead of seeing a list of games, you pick the map and then wait
20160223 06:22:25< aeth> Or some other parameters like player count
20160223 06:22:38< aeth> That'd probably be a nice to have feature before going on Steam if Wesnoth does go on Steam
20160223 06:22:58< celticminstrel> How is that different from starting a game and waiting for people to join?
20160223 06:24:08< celticminstrel> vultraz: That shouldn't be hard to do at the client side.
20160223 06:25:32< vultraz> aeth: well, yes
20160223 06:25:35< aeth> celticminstrel: It's very similar, but it only makes a game if there isn't currently a game of the same type, so in theory games would be formed faster given enough players.
20160223 06:25:38< vultraz> aeth: it always has.
20160223 06:25:47< vultraz> you create a game, it shows up in the lobby as open
20160223 06:25:58-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20160223 06:26:05< celticminstrel> So if two people start a game on the same map, they could get merged into a single game.
20160223 06:26:13< aeth> yes
20160223 06:26:30< vultraz> hm
20160223 06:26:34< vultraz> that is an idea, yes
20160223 06:26:35< aeth> instead of sitting in the lobby forever, wondering why no one is joining
20160223 06:26:49< aeth> Only a big deal if they coincidentally make it around the same time
20160223 06:30:13< vultraz> aeth: people will argue "but what if the two games don't have exactly the same settings"
20160223 06:30:33< aeth> vultraz: It would probably only work on default settings
20160223 06:30:42< vultraz> eh, true
20160223 06:30:46< vultraz> but what about mods?
20160223 06:30:47< aeth> vultraz: So it would be "Quick Start" or something rather than "Start" and not allow settings tweaks and thus not be for everyone
20160223 06:30:55< vultraz> ahhhh
20160223 06:30:57< vultraz> yes
20160223 06:30:59< vultraz> good idea
20160223 06:31:01< vultraz> i like this
20160223 06:31:01< aeth> It would work with mods and add-on maps as long as it's defaults
20160223 06:31:25< vultraz> you should file a FR with loonycyborg 
20160223 06:31:25< aeth> But if you have 5 mods and a custom map, it'll only match you with someone with the same 5 mods and a custom map, so it probably won't be as helpful
20160223 06:31:34< aeth> You'd have to select which mods to use, I guess
20160223 06:31:41< aeth> defaulting to 0 of course
20160223 06:31:43< aeth> same with era
20160223 06:31:56< vultraz> I think we should take inspiration from dota2's custom games
20160223 06:32:14-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 06:32:25< aeth> I've never played Dota 2.
20160223 06:32:31< vultraz> (I'm taking inspiration from dota2 for a lot of my UI design ideas)
20160223 06:32:37< aeth> I wanted to try it once but it had some kind of glitched perpetually-updating state so I just uninstalled it
20160223 06:33:13< aeth> vultraz: good because my strategy influences in the past few years are mostly paradox strategy games so I'm glad our influences are different
20160223 06:33:17< aeth> it'll give us different ideas
20160223 06:33:23< aeth> although idk if Paradox has a matchmaking system... probably not
20160223 06:37:15-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 276 seconds]
20160223 07:03:37-!- celticminstrel is now known as celmin|sleep
20160223 07:06:54-!- mjs-de [~mjs-de@f048098199.adsl.alicedsl.de] has joined #wesnoth-dev
20160223 07:23:34< Aginor> 'lo
20160223 07:38:42< vultraz> lo
20160223 07:49:54< Aginor> 'sup?
20160223 07:52:08< vultraz> pondering gui2
20160223 07:52:20-!- Alduin_ [~Alduin@p2003008644549600E29467FFFE0EE8C4.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160223 07:52:40< vultraz> if I can find a way to add size keys to the individual widgets...
20160223 07:53:06-!- boucman_work [~jrosen@193.56.60.161] has joined #wesnoth-dev
20160223 07:53:06-!- boucman_work [~jrosen@193.56.60.161] has quit [Changing host]
20160223 07:53:06-!- boucman_work [~jrosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160223 08:10:20-!- vn971 [~vasya@91.247.233.240] has quit [Quit: Leaving.]
20160223 08:18:39< ancestral> Why is it that #enddef must end immediately after the string instead of say, a return onto the next line?
20160223 08:18:58< ancestral> for example, data/themes/default.cfg
20160223 08:19:42< ancestral> If I put in a return in sone of the #defines, it seems it breaks something
20160223 08:19:48< ancestral> *some
20160223 08:20:04< ancestral> (Not specifically this file, but other files I’ve tried it)
20160223 08:22:52-!- Kwandulin [~Miranda@p200300760F0BC52A2D1B033D03745A0C.dip0.t-ipconnect.de] has quit [Ping timeout: 250 seconds]
20160223 08:23:28-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160223 08:26:00< zookeeper> gfgtdf, what the heck, why did you remove it?
20160223 08:27:17< zookeeper> there's no reason to remove it and we don't know how many uses of that exist even in mainline, so please revert that
20160223 08:31:22< vultraz> he said it causes oos
20160223 08:31:25< vultraz> so it's fine
20160223 08:33:27< zookeeper> not on this channel
20160223 08:33:44< zookeeper> and i can't see how it could cause OOS
20160223 08:34:10< vultraz> in the commit message
20160223 08:35:13< zookeeper> oh i see. still, i can't see how.
20160223 08:35:35< zookeeper> language doesn't cause different results in comparison, because both strings being compared get translated.
20160223 08:35:48< iceiceice> the string has a different value when you load the game in a different language
20160223 08:35:57< zookeeper> it only causes OOS if you would do the same mistake as you originally, as in not translation-marking one of the strings being compared
20160223 08:37:05< iceiceice> yeah but its a really bad idea to filter on name anyways
20160223 08:37:08< iceiceice> why not filter on id
20160223 08:37:14< iceiceice> this is like literally the reason to have id
20160223 08:37:20< zookeeper> iceiceice, so if you save a game using language X, change language to Y, then load it, you'll see a mix of both languages?
20160223 08:37:41< iceiceice> i dont remember all the nitty gritty of how wesnoth handles tstrings
20160223 08:37:49< iceiceice> i would imagine that it is wildly inconsistent
20160223 08:38:00< iceiceice> regarding like, saving the game in language X and reloading in Y,
20160223 08:38:17< iceiceice> saving and reloadinga replay,
20160223 08:38:23< iceiceice> playing in lang X with someone in lang Y in multiplayer...
20160223 08:38:28< iceiceice> but i really have no idea
20160223 08:39:00< zookeeper> well, testing is easy. let's see...
20160223 08:41:23< irker806> wesnoth: Charles Dang wesnoth:master 1a6bfca90356 / data/gui/default/window/preferences/ (01_general.cfg 02_display.cfg 03_sound.cfg 05_advanced.cfg): tpreferences: cleaned up some unnecessary grid nesting https://github.com/wesnoth/wesnoth/commit/1a6bfca90356f90a182ae4267f16ad11eaa9d4c7
20160223 08:42:54< irker806> wesnoth: Charles Dang wesnoth:master 05411b6d34ad / data/gui/default/window/preferences/01_general.cfg: tpreferences: made sure Scroll Speed slider fills all available space https://github.com/wesnoth/wesnoth/commit/05411b6d34ad19440565feea8a4bf6d5271bb0a9
20160223 08:43:36< zookeeper> well, a simple name= _ "Konrad" event filter doesn't work in 1.12.5, translations or not. translation marker makes no difference either.
20160223 08:43:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 08:44:36< zookeeper> it works as soon as i remove the translation marker from the unit's name though
20160223 08:44:47< vultraz> so see, it's broken
20160223 08:45:02< zookeeper> it's a bug!
20160223 08:45:25< vultraz> but iceiceice is right
20160223 08:45:29< vultraz> filtering on name isn't good
20160223 08:45:34< shadowm> gfgtdf: Perhaps you might want to consider displaying an error message when name is found in SUFs instead of ignoring it silently.
20160223 08:46:10< iceiceice> idk i mean
20160223 08:46:27< zookeeper> all i see is people saying "filtering on name isn't good" without being able to explain why exactly that's the case
20160223 08:46:31< iceiceice> to my mind the reason to have id is exactly to allow easy filtering on stuff
20160223 08:46:46< iceiceice> b/c testing translatable strings for equality is actually pretty complex i guess
20160223 08:47:03< iceiceice> i dont even remember allt he data that wesnoth stores about translatable strings
20160223 08:47:13< iceiceice> probably it stores like, the text? the text domain?
20160223 08:47:23< iceiceice> what happens when a translatable string is concatenated with a regular string
20160223 08:47:29< iceiceice> should it test equal to some form of the concatenation?
20160223 08:47:35< iceiceice> how is that all going to work under the hood
20160223 08:47:37< zookeeper> iceiceice, you can't have the same id for multiple units, but you can have the same name. let's say you have 10 guard units with the same "stone guardian" name, and naturally different id's. filtering by name is quite convenient in that case, because otherwise you have to give them roles or something.
20160223 08:47:56< iceiceice> i see
20160223 08:47:59-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Ping timeout: 240 seconds]
20160223 08:48:03< iceiceice> can you filter on like, a regular expression of the name?
20160223 08:48:06< shadowm> Use [filter_wml] and unit variables.
20160223 08:48:09< iceiceice> or a prefix of the name?
20160223 08:48:10< zookeeper> iceiceice, nope
20160223 08:48:28< iceiceice> so i mean, if filtering on name is a sanctioned technique,
20160223 08:48:30< iceiceice> then why have id at all
20160223 08:48:42< iceiceice> why not always filter on name when you would filter on id
20160223 08:48:49< shadowm> Either that or add a tag attribute to units (but that'd be duplicating role's functionality? not sure).
20160223 08:49:03< zookeeper> because names would then require translation markers everywhere, and might contain characters that are awkward or break saves or whatever
20160223 08:49:11< zookeeper> well, scratch the latter, but still
20160223 08:51:10< iceiceice> i thought names always require translation markers anyways
20160223 08:51:16< zookeeper> nothing requires them
20160223 08:51:24< shadowm> No translatable attributes are ever required to be translatable.
20160223 08:51:28< iceiceice> hmm yeah what i said cant be right
20160223 08:52:12< iceiceice> well i mean
20160223 08:52:17< iceiceice> i guess what you could do is reenable it
20160223 08:52:24< iceiceice> but make it emit warnigns whenever anything squirrelly is happening
20160223 08:52:35< iceiceice> and make some tests to see exactly what squirrely things can happen
20160223 08:52:39< iceiceice> idk
20160223 08:52:48< iceiceice> i guess its not unreasonable to filter on name if you really want to
20160223 08:53:12< zookeeper> it's of course possible to cause OOS by using a translated attribute wrong
20160223 08:53:15< iceiceice> but it does seem tricky
20160223 08:53:30< zookeeper> but name= is hardly the only possibility for that. you can just wrap it in a [filter_wml] if you want, after all.
20160223 08:54:38< vultraz> shadowm: are you going to elaborate on why you reverted that scrollbar commit?
20160223 08:54:59< vultraz> shadowm: I made sure to adhere to both of your stipulations 
20160223 08:55:44< shadowm> I don't take kindly to people rules lawyering matters like this.
20160223 08:56:28< shadowm> If I said it wasn't a reasonable solution for an issue and that it shouldn't be done, then it still shouldn't be done even when not targeting the original issue.
20160223 09:02:14< zookeeper> tested: when wrapped in a [filter_wml], a name= event filter works correctly when changing language, loading a save, etc etc. so quite possibly the SUF version fails because it's misguidedly trying to account for translations somehow.
20160223 09:02:43< zookeeper> which the commit diff didn't seem to indicate though, so dunno
20160223 09:03:18< zookeeper> well, maybe comparing with u.name() is the cause
20160223 09:04:29< zookeeper> in any case, i don't see reason to suspect that there's fundamentally anything wrong with filtering by name if the same thing works as it should in a [filter_wml]
20160223 09:04:57< iceiceice> i will say, i think it probably shoudl be that
20160223 09:05:11< iceiceice> if you filter on a name whcih is translatable
20160223 09:05:18< iceiceice> and use a untranslatable key in the filter,
20160223 09:05:24< iceiceice> that should probably just be an error or something
20160223 09:05:29< vultraz> zookeeper: would you mind if I decrease the frame length of ANIMATION_18_70?
20160223 09:05:34< vultraz> (Windmills)
20160223 09:05:38< iceiceice> rather than just a "see if this name happens to match in this langauge"
20160223 09:05:44< zookeeper> vultraz, as long as it doesn't start looking bad, then no
20160223 09:05:54< iceiceice> b/c that would create undo testing burden, imo
20160223 09:06:00< iceiceice> *undue
20160223 09:06:36< iceiceice> idk what wesnoth actually does
20160223 09:07:47< shadowm> vultraz: What's the problem with the current length? The animation looks just fine to me.
20160223 09:08:23< shadowm> Any faster and it'd look more like one of those spinny toys.
20160223 09:08:24< zookeeper> iceiceice, eh, i don't really see the point. you'd still only catch cases where that's done in SUF's, not when it's done in [if] conditionals etc.
20160223 09:08:27< vultraz> shadowm: I propose reducing delay by 40 for a smooth, more 'windmill-y' experience
20160223 09:08:45< shadowm> Any faster and it'd look more like one of those spinny toys.
20160223 09:09:09< zookeeper> ohh. decrease. i interpreted that as increase.
20160223 09:09:13< vultraz> I just gave you my proposed reduction so you could look at it yourself before copypasting the same line
20160223 09:10:12< zookeeper> you should make it slightly slower, because slower animations are better for performance :>
20160223 09:10:25< vultraz> . . .
20160223 09:11:00< shadowm> You broke the GUI1 scrollbar groove and block.
20160223 09:11:08< vultraz> I did not
20160223 09:11:40< shadowm> Then why do I only see the arrow buttons?
20160223 09:12:46< vultraz> Because you don't have latest master
20160223 09:12:48< vultraz> I fixed that
20160223 09:13:22< shadowm> Okay, then you did break the GUI1 scrollbar groove and block in the past.
20160223 09:13:48< vultraz> I broke them in 5aa86c42f004  and fixed them in 499cadc37b99 
20160223 09:14:43< shadowm> I don't see what's not "windmilly" about the current speed (or what "windmilly" is supposed to be), and the increase in speed makes it look much more conspicuous than before.
20160223 09:15:12< shadowm> Which I suspect is the opposite of what you want.
20160223 09:16:59< vultraz> I want it smoother 
20160223 09:17:07< vultraz> I dislike that our animated terrains are not smooth
20160223 09:17:11< vultraz> ly animated
20160223 09:17:12< zookeeper> it's the smoothest animation we have
20160223 09:17:14< shadowm> Faster is not the solution.
20160223 09:17:49< shadowm> I also honestly think you are looking for an excuse to make a line edit.
20160223 09:19:59< shadowm> There are 215 unassigned bugs with status None on the bug tracker.
20160223 09:21:17< shadowm> And 302 feature requests with the same status.
20160223 09:21:49< zookeeper> oh, it's actually :130 currently
20160223 09:23:15< zookeeper> if you really had to, i guess it could go down to :100
20160223 09:29:59-!- vn971 [~vasya@91.247.233.240] has joined #wesnoth-dev
20160223 09:33:10-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160223 09:46:14-!- vn971 [~vasya@91.247.233.240] has quit [Quit: Leaving.]
20160223 10:28:19-!- atarocch [~atarocch@31.159.138.0] has joined #wesnoth-dev
20160223 10:29:01-!- vn971 [~vasya@91.247.233.240] has joined #wesnoth-dev
20160223 10:31:22< vultraz> shadowm: do you still want me to do this? https://gna.org/bugs/index.php?24164
20160223 10:37:07< irker806> wesnoth: Charles Dang wesnoth:master f23bc20e8217 / src/gui/dialogs/preferences_dialog.cpp: tpreferences: fixed incorrect use of doxygen comments https://github.com/wesnoth/wesnoth/commit/f23bc20e8217d558cd11f450d9a1633555a6bd11
20160223 10:37:12-!- Alduin_ [~Alduin@p2003008644549600E29467FFFE0EE8C4.dip0.t-ipconnect.de] has quit [Ping timeout: 250 seconds]
20160223 10:37:28-!- atarocch [~atarocch@31.159.138.0] has quit [Ping timeout: 255 seconds]
20160223 10:38:16-!- atarocch [~atarocch@31.159.138.0] has joined #wesnoth-dev
20160223 10:54:59-!- atarocch [~atarocch@31.159.138.0] has quit [Ping timeout: 240 seconds]
20160223 11:00:03< irker806> wesnoth: Charles Dang wesnoth:master 168fce195844 / data/campaigns/Legend_of_Wesmere/scenarios/chapter1/03_Kalian_under_Attack.cfg: LoW S3: fixed incorrectly placed labels (bugs #22931, #23776) https://github.com/wesnoth/wesnoth/commit/168fce195844c0b6b840373da240fdee90b1590e
20160223 11:01:23< vultraz> I don't understand why that fixes it, but it does
20160223 11:06:35< vultraz> Probably something to do with toplevel wml not using the lua wml actions or something
20160223 11:09:50-!- atarocch [~atarocch@31.159.138.0] has joined #wesnoth-dev
20160223 11:21:43< vultraz> gfgtdf: i just noticed there're warnings in stderr about unknown unit "is_local" attribute
20160223 11:21:53< vultraz> gfgtdf: do you know anything about this
20160223 11:25:40< vultraz> gfgtdf: should it just be added to the list of known attributes?
20160223 11:42:10-!- exciton [chuck-the-@89.208.170.132] has joined #wesnoth-dev
20160223 11:49:36-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has joined #wesnoth-dev
20160223 11:49:58< vultraz> gfgtdf: also I discovered that all commands like ;cl and ;turn are allowed in replays which causes asserts
20160223 11:50:07< vultraz> gfgtdf: i think we should disable those in replays 
20160223 11:57:01-!- Netsplit *.net <-> *.split quits: exciton
20160223 12:03:41< vultraz> (the assert is at synced_context.cpp:94)
20160223 12:09:07-!- vn971 [~vasya@91.247.233.240] has quit [Ping timeout: 248 seconds]
20160223 12:09:40-!- vn971 [~vasya@91.247.233.240] has joined #wesnoth-dev
20160223 12:21:52-!- prkc [~prkc@gateway/vpn/privateinternetaccess/prkc] has quit [Ping timeout: 255 seconds]
20160223 12:36:53-!- prkc [~prkc@46.166.190.188] has joined #wesnoth-dev
20160223 12:40:27-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160223 12:57:40< shadowm> I will not be available for the remainder of this week. Hopefully I'll see you later.
20160223 12:57:42-!- shadowm [~ignacio@wesnoth/developer/shadowm] has quit []
20160223 13:07:58< zookeeper> roger
20160223 13:11:13-!- nkr [Elite14718@gateway/shell/elitebnc/x-vpxpgyvzbfxngwpy] has left #wesnoth-dev ["Leaving..."]
20160223 13:16:32-!- gfgtdf [~chatzilla@f054060000.adsl.alicedsl.de] has joined #wesnoth-dev
20160223 13:16:33< gfgtdf> 20160223 08:26:00< zookeeper> gfgtdf, what the heck, why did you remove it?
20160223 13:16:55< zookeeper> p.s. also see the following discussion :P
20160223 13:18:12< gfgtdf> zookeeper: becasue it can cause OOS. There are mutiple reasons this can happen,even if both the filter name and the unitnames are translatable.
20160223 13:19:02< zookeeper> can you specify any such case? because so far i've got nothing
20160223 13:19:07< gfgtdf> zookeeper: first itmight depend on the tranlation ofthe unit names whichmight be different in different langugages, for example2 units miht have teh same name in one langugage or  diffeent names in another
20160223 13:19:54< gfgtdf> zookeeper: also tranlsations areknown to be orken in mp. so i wouldnt rely on this working correctlyeven if both attributearesothis is also a problem
20160223 13:20:03< gfgtdf> broken
20160223 13:21:05< gfgtdf> zookeeper: on the other hand i really dont see a reason to ever use namein filters when you could lso use idsor role
20160223 13:21:27< gfgtdf> 20160223 11:21:43< vultraz> gfgtdf: i just noticed there're warnings in stderr about unknown unit "is_local" attribute
20160223 13:21:43-!- Appleman1234 [~Appleman1@KD119104007069.au-net.ne.jp] has quit [Ping timeout: 255 seconds]
20160223 13:22:16< zookeeper> even if there is some specific unverified case where one might be able to cause OOS with a name filter, the point still stands that we don't know what the removal might break in mainline. and removing it is certainly not going to unbreak any UMC that's currently OOS-prone because of it.
20160223 13:22:58< gfgtdf> vultraz: yyes thisis a known attribute of [side] (internally, you cannot change it in [side] in [scenario]) it should be addedtotheknown attrbuts list in team.cpp
20160223 13:23:43< iceiceice> do you guys know if filtering on name is even a part of 1.12?
20160223 13:23:54< iceiceice> as i recall i refactored unit filters at some point,
20160223 13:24:02< iceiceice> and then gfgtdf later refactored again
20160223 13:24:06< iceiceice> or possibly just reverted my refactors
20160223 13:24:07 * zookeeper doesn't know
20160223 13:24:12< gfgtdf> zookeeper: i think we can assume that this feature not wideley uses, becasue it was codumented in the wiki.
20160223 13:24:16< iceiceice> it may be that i added it by mistake or something
20160223 13:24:48< iceiceice> i'm trying to figure it out
20160223 13:24:50< gfgtdf> iceiceice: no 'name'fltreing evenexists before your refactors, i looked, at the commit that introduces it before i remved it
20160223 13:24:57< iceiceice> oh, ok
20160223 13:25:06< zookeeper> i think we can safely assume that it has existed in the past, because of how it's been used in mainline. so if someone has removed it since, mainline usecases have not been updated and that's probably the fault of whoever removed it :p
20160223 13:26:33< zookeeper> however, as i noticed, it's currently broken even in 1.12, although it can be circumvented by use of [filter_wml]. as for why it's broken, i don't know.
20160223 13:26:55< zookeeper> i cannot test it on master because translations don't work there for me
20160223 13:28:10< gfgtdf> zookeeper: well if its borken even in 1.12 thni sureleydidnt breaak anything by removing it. imight add a message about namenot supported when 'name' is given thought (along with a comment in the code why t not supported)
20160223 13:28:11< zookeeper> well, i mean i can test whether it works without translations (before gfgtdf's removal), and... it doesn't
20160223 13:28:50< gfgtdf> 20160223 11:49:58< vultraz> gfgtdf: also I discovered that all commands like ;cl and ;turn are allowed in replays which causes asserts
20160223 13:29:14< irker806> wesnoth: Charles Dang wesnoth:master cca804e78d3b / src/team.cpp: Add 'is_local' to recognized key list https://github.com/wesnoth/wesnoth/commit/cca804e78d3b08e6c940bb30ef43ccb5ae41142f
20160223 13:29:38< gfgtdf> vultraz: well yes that would be good. but i thing that commands like 'inspect' are still useful in repplays
20160223 13:29:39< zookeeper> gfgtdf, sure. but that doesn't mean that keeping an unknown number of usecases broken is better than fixing them by fixing the filter.
20160223 13:29:53< vultraz> gfgtdf: yes those are useful
20160223 13:30:15< vultraz> gfgtdf: inspect shouldn't be a synched action tho
20160223 13:30:52< gfgtdf> zookeeper: no i relly dont think so, i mean mainline campaigns were tested n 1.12 by the user so there are sureley arnt any important uecasesof name=
20160223 13:30:58< vultraz> gfgtdf: so I'm not sure if a checked needs to be added to synced_context.cpp (I actually tried but it didn't really seem to work) or in each individual command callback in menu_events.cpp
20160223 13:31:03< gfgtdf> vultraz: yes it isnt
20160223 13:31:17< zookeeper> gfgtdf, i'm sure there's no game-breaking usecases, yes
20160223 13:32:18< gfgtdf> zookeeper: even missing dialogswould mostlikeley be noticed since dialogs often dont makes sense if random messagesaremissing in them.
20160223 13:32:36< gfgtdf> vultraz: well you cancheckfor pc_->is_repaly()
20160223 13:32:57< vultraz> not resources::controller?
20160223 13:33:08< gfgtdf>  vultraz im menu_events.cpp i mean
20160223 13:33:21< vultraz> oh ok
20160223 13:33:39< vultraz> so each individual command that shouldn't work in replays should get that check?
20160223 13:33:52< gfgtdf> vultraz: well you can also use resurces::controller . but if you already have a local play cntrollerobject then why not use that
20160223 13:34:15< gfgtdf> vultraz: y maybe there is a nother solution but imnot sure
20160223 13:34:44< vultraz> well synced_context::run_in_synced_context_if_not_already does an anit-replay check
20160223 13:36:11< gfgtdf> vultraz: hmm te mainpoint of synced_context::run_in_synced_context_if_not_already is thatit also works from insid synced context which is not what we want
20160223 13:36:34< vultraz> yes but I was thinking to add a similar check to run_and_throw
20160223 13:37:54< vultraz> tho that doesn't work
20160223 13:37:55< vultraz> hm
20160223 13:38:07< gfgtdf> vultraz: hmm imsuprosed thatdoesnt work,whats the error ?
20160223 13:39:37< vultraz> gfgtdf: resources::controller->is_replay() is false for some reason
20160223 13:40:04< vultraz> even if you call the synced action from a replay
20160223 13:40:04< gfgtdf> vultraz: durng a normal replay?
20160223 13:41:20< vultraz> yes
20160223 13:42:06< gfgtdf> vultraz: hmm the code is 'virtual bool is_replay() { return false; }'
20160223 13:42:18< vultraz> what
20160223 13:42:19< gfgtdf> vultraz: in play_controller.hpp
20160223 13:42:42< gfgtdf> vultraz: i thought this would be overiddn in playsingle_controller.hpp but this seems not to be the case
20160223 13:42:54< vultraz> but it's not reimplemented anywhere :/
20160223 13:42:57< vultraz> er
20160223 13:42:59< vultraz> yeah
20160223 13:43:10< vultraz> is this a bug?
20160223 13:43:20< gfgtdf> vultraz: mostlikeley yes
20160223 13:43:40< vultraz> can you fix it?
20160223 13:43:54< gfgtdf> vultraz: y but not fromthis pc
20160223 13:44:00< vultraz> ok 
20160223 13:44:32< gfgtdf> vultraz: playsingle_controller should have is_replayoverrideen which should return smething like 'mp_replay_.get() != NULL'
20160223 13:44:59-!- boucman_work [~jrosen@wesnoth/developer/boucman] has quit [Ping timeout: 240 seconds]
20160223 13:45:00< gfgtdf> and we shoudl also rename the mp_eplay_ object since its not really related to mp anymore
20160223 13:46:38< vultraz> hm it has a get_replay_controller function
20160223 13:46:40< vultraz> so
20160223 13:47:08< vultraz> bool is_replay() const { get_replay_controller() != NULL; } ?
20160223 13:48:09< gfgtdf> vultraz: yes but in playsingle_controller.hpp
20160223 13:54:02-!- gfgtdf [~chatzilla@f054060000.adsl.alicedsl.de] has quit [Quit: ChatZilla 0.9.92 [Firefox 44.0.2/20160210153822]]
20160223 13:59:32-!- boucman_work [~jrosen@bob75-2-81-56-46-209.fbx.proxad.net] has joined #wesnoth-dev
20160223 13:59:32-!- boucman_work [~jrosen@bob75-2-81-56-46-209.fbx.proxad.net] has quit [Changing host]
20160223 13:59:32-!- boucman_work [~jrosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160223 14:10:39-!- Alduin_ [~Alduin@p2003008644549600E29467FFFE0EE8C4.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160223 14:11:42< irker806> wesnoth: Charles Dang wesnoth:master 1c0085616a8a / src/ (hotkey_handler_sp.cpp playsingle_controller.hpp): Implement is_replay() in playsingle_controller https://github.com/wesnoth/wesnoth/commit/1c0085616a8a07da937d4bc636b5f496c64f8a44
20160223 14:11:45< irker806> wesnoth: Charles Dang wesnoth:master 800a590525be / src/synced_context.cpp: Moved synced_context replay check deeper https://github.com/wesnoth/wesnoth/commit/800a590525bea30305fe68afe152ffa1eb571ad4
20160223 14:11:48< irker806> wesnoth: Charles Dang wesnoth:master 53190dfb4c55 / src/replay.cpp: Fixed a typo in an error https://github.com/wesnoth/wesnoth/commit/53190dfb4c550da77266f6d673e4af14eb5f6322
20160223 14:11:51< vultraz> gfgtdf ^
20160223 14:11:56< vultraz> I hope I did that right
20160223 14:12:01< vultraz> especially 800a590525be 
20160223 14:12:14< vultraz> since i just moved the check deeper instead of duplicating it
20160223 14:12:16-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160223 14:13:29-!- boucman_work [~jrosen@wesnoth/developer/boucman] has quit [Ping timeout: 240 seconds]
20160223 14:13:51< irker806> wesnoth: gfgtdf wesnoth:master 3683490ac688 / src/unit_filter.cpp: add wanring for name= in standard unit filters. https://github.com/wesnoth/wesnoth/commit/3683490ac6885e58c7577a3080e747975dd1f9dc
20160223 14:14:34-!- boucman_work [~jrosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160223 14:15:10-!- Alduin_ [~Alduin@p2003008644549600E29467FFFE0EE8C4.dip0.t-ipconnect.de] has quit [Ping timeout: 250 seconds]
20160223 14:16:11-!- gfgtdf [~chatzilla@f054060000.adsl.alicedsl.de] has joined #wesnoth-dev
20160223 14:16:22< gfgtdf> hm i see no reason why this shouldn't work.
20160223 14:27:13-!- Alduin_ [~Alduin@2001:4ca0:4fff:4::84] has joined #wesnoth-dev
20160223 14:36:51< zookeeper> well, i'm still not buying it. i tested the [filter_wml] workaround in 1.12 multiplayer, host client in spanish and another client in english, and a filter on a translated name passed correctly on both clients.
20160223 14:37:06< zookeeper> then i started a new game by loading that save on the spanish client, and again it worked.
20160223 14:37:29< zookeeper> well, in between i also tested it by hosting on the english client
20160223 14:37:42< gfgtdf> zookeeper: what exactly worked ?
20160223 14:37:44< zookeeper> and that's the game i loaded from
20160223 14:37:45< zookeeper> the filter
20160223 14:39:57< gfgtdf> zookeeper: y the conditions on how to trigger the mp bug ina  wayas that is casues OOS when used with name filters are not that easy, what id# try is creating a mp campaign and then fultering by name by lua code in the second scenario.
20160223 14:40:54< zookeeper> i wouldn't even know how to do that
20160223 14:41:33< gfgtdf> zookeeper: what exactly is unclear about that ?
20160223 14:42:02< zookeeper> both the MP campaign and lua part
20160223 14:45:07< gfgtdf> zookeeper: well you add something like [lua] code = << wesnoth.get_units name = { wesnoth.textdomain("<low textdomain>")("name of th unit") }[1].hitpints = 1000 >> in LoW scenario 2 
20160223 14:51:29< zookeeper> takes forever to test that
20160223 14:53:30< zookeeper> but fine
20160223 14:54:34< zookeeper> so is this correct? [lua] code = << wesnoth.get_units name = { wesnoth.textdomain("wesnoth-low")(_"the unit name") }[1].hitpoints = 1000 >> [/lua]
20160223 14:56:52< gfgtdf> i'd say yes and ofc you need to test it on a language whereteh name is actualyl different form th iroginal (untraslated) name
20160223 14:57:00< zookeeper> of course
20160223 14:59:02< zookeeper> wow, this stuff is really broken in 1.12. both velon and the orcish leader i killed were called zookeeper.
20160223 14:59:20-!- celmin|sleep is now known as celticminstrel
20160223 14:59:38< gfgtdf> zookeeper: y i comketeleey forgot about that bug
20160223 14:59:39< zookeeper> except on the second spanish client where they have the correct names
20160223 15:00:57< zookeeper> oh i can't recall in the second scenario. how convenient, my testcase went down the drain.
20160223 15:02:55< zookeeper> another try, then...
20160223 15:03:44< celticminstrel> Does ERR_CF go to chat?
20160223 15:05:41< zookeeper> gfgtdf, i cannot test that. i create a unit in the second scenario with a translated name, and it shows up in english on both clients.
20160223 15:06:00< zookeeper> in the first scenario the translated name showed up
20160223 15:06:14 * vultraz just took a look at the gamestate inspector code
20160223 15:06:21< vultraz> the hell is all this o_O
20160223 15:06:28-!- boucman_work [~jrosen@wesnoth/developer/boucman] has quit [Ping timeout: 252 seconds]
20160223 15:07:10< zookeeper> and, like, "syntax error near 'name'"
20160223 15:08:33< celticminstrel> vultraz: What's what?
20160223 15:09:12< vultraz> celticminstrel: everything in gamestate_inspector.cpp. It's unlike any gui2 dialog I've seen O_O
20160223 15:09:18< vultraz> special classes
20160223 15:09:30< vultraz> "model" implementations 
20160223 15:09:33< vultraz> what is all this o_o
20160223 15:12:57< zookeeper> gfgtdf, looks like you had the opening bracket in the wrong place. third try...
20160223 15:13:31< iceiceice> vultraz, there's this thing called mvc
20160223 15:13:38< iceiceice> you should probably read about it
20160223 15:13:41< vultraz> ??
20160223 15:13:49< iceiceice> model view controller is a design pattern
20160223 15:14:23< iceiceice> i think its not just that dialog, several of the more complicated dialogs in wesnoth hve "model" classes and such
20160223 15:15:33< vultraz> ...huh
20160223 15:17:06< vultraz> ok...
20160223 15:17:08< gfgtdf> celticminstrel: no it doesnt go to chat it just goes to stderr
20160223 15:17:12< vultraz> but why was this used
20160223 15:17:15< vultraz> i mean
20160223 15:17:17< vultraz> why mvc
20160223 15:17:19< celticminstrel> Oh well.
20160223 15:17:21< vultraz> why here
20160223 15:17:30< zookeeper> gfgtdf, still the same thing. unit created in the second scenario with a translated name shows up in english. and even fixing the bracket, the lua code produces an error.
20160223 15:17:30< vultraz> why not everywhere in gui2
20160223 15:17:33< vultraz> confused, I am
20160223 15:17:36< celticminstrel> Still probably better if it doesn't spam the console.
20160223 15:17:54< zookeeper> so pardon me but i'm not going to further pursue this in an attempt to prove myself wrong
20160223 15:18:11 * celticminstrel has already made the commit, anyway.
20160223 15:18:51< gfgtdf> zookeeper: well you already saw that at lest the leadernames are wrong in 1.12 . Anther other code about 2 units can have the same name ina translatation is also unrelated ot that test
20160223 15:19:56< gfgtdf> zookeeper: on the other side you offered no usecases that might show that auch an attribute in unuit fitkler migth make sense
20160223 15:20:02< gfgtdf> such*
20160223 15:23:04-!- boucman_work [~jrosen@193.56.60.161] has joined #wesnoth-dev
20160223 15:23:04-!- boucman_work [~jrosen@193.56.60.161] has quit [Changing host]
20160223 15:23:04-!- boucman_work [~jrosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160223 15:27:02< irker806> wesnoth: Celtic Minstrel wesnoth:master fd939cee76c4 / src/unit_filter.cpp: Unit filters: move name warning to constructor and suggest alternatives https://github.com/wesnoth/wesnoth/commit/fd939cee76c47207db73f71a8194a3d2611f7252
20160223 15:27:04< irker806> wesnoth: Celtic Minstrel wesnoth:master 18bcd519e0dd / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: XCode: Suppress warnings about unknown warnings https://github.com/wesnoth/wesnoth/commit/18bcd519e0dd9de9f411e0bb1aecd77f555629eb
20160223 15:29:29-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Quit: Leaving]
20160223 15:30:02< celticminstrel> Hmm. Would copying a unit on every iteration of a filter be too expensive?
20160223 15:30:45< gfgtdf> celticminstrel: i'd say yes
20160223 15:30:56< zookeeper> yeah, as i thought the comparison of vcfg["name"] with u.name() is broken, it won't match even when the strings are identical
20160223 15:30:57< celticminstrel> I guess I can't use luaW_pushlocalunit then.
20160223 15:30:58< gfgtdf> celticminstrel: why woudl you wan tot do that ?
20160223 15:31:05< zookeeper> as per my debug output, "name Konrad not blank but not equal to Konrad" :p
20160223 15:31:50< gfgtdf> celticminstrel: luaW_pushlocalunit shoudlnt copy units
20160223 15:32:00< celticminstrel> It requires a non-const unit.
20160223 15:32:13< celticminstrel> Copying the unit would be one way of working around that.
20160223 15:32:45< gfgtdf> celticminstrel: hmm what do you want to do in the first place ?
20160223 15:33:00< celticminstrel> I'm implementing [lua] in unit filters.
20160223 15:33:44< zookeeper> lua_function= isn't enough?
20160223 15:38:24-!- travis-ci [~travis-ci@ec2-54-146-170-187.compute-1.amazonaws.com] has joined #wesnoth-dev
20160223 15:38:26< travis-ci> wesnoth/wesnoth#8567 (master - 3683490 : gfgtdf): The build has errored.
20160223 15:38:26< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/111215481
20160223 15:38:26-!- travis-ci [~travis-ci@ec2-54-146-170-187.compute-1.amazonaws.com] has left #wesnoth-dev []
20160223 15:38:36-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc]
20160223 15:39:02< zookeeper> as said i can't easily test this, but it's possible that this is all that's needed to fix the name filter:
20160223 15:39:03< zookeeper> -if (!vcfg["name"].blank() && vcfg["name"].str() != u.name()) {
20160223 15:39:03< zookeeper> +if (!vcfg["name"].blank() && vcfg["name"].t_str() != u.name()) {
20160223 15:41:02-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160223 15:42:41< gfgtdf> celticminstrel: i also wonder why we need more than lua_function= and how you alternative looks like ?
20160223 15:48:02< celticminstrel> Well, lua_function pretty much works fine (though it has no way to know about the filter's second unit, if present), but I like having consistent syntax, and lua_function requires the function to be a toplevel global.
20160223 15:50:21< gfgtdf> celticminstrel: so what is your suggested syntax ?
20160223 15:50:26-!- stikonas_ [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160223 15:50:27-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 248 seconds]
20160223 15:50:39< celticminstrel> Same as everywhere else - [lua]code=[args]
20160223 15:50:59< gfgtdf> celticminstrel: yes but how should the content of code lok like
20160223 15:51:06< zookeeper> ...i mean, i can confirm it fixes the bug when not using a translation, but i cannot easily test whether it fixes the issue when the name is actually translated.
20160223 15:51:21< gfgtdf> celticminstrel: also note that queriing a toplebel lua variable is faster then compliting loa code
20160223 15:51:36< gfgtdf> compiling
20160223 15:51:36< zookeeper> celticminstrel, you want to make the lua code be executed every time a unit is evaluated for a match?
20160223 15:51:58< celticminstrel> The equivalent of lua_function="blah" would be code="return blah(...)", I think.
20160223 15:52:23< celticminstrel> zookeeper: Huh? That's not something that can be avoided.
20160223 15:52:35< celticminstrel> If you meant compiled rather than executed, the answer is no.
20160223 15:52:37< zookeeper> yeah
20160223 15:52:50< gfgtdf> celticminstrel: yes and im quite sure that thats slower than  the lua_function solution.
20160223 15:53:59< zookeeper> using lua in a WML filter is such a niche case that consistent syntax doesn't strike me as important enough to introduce any sort of complications anywhere for
20160223 15:54:22< zookeeper> and we're using "consistent syntax" in a very loose way here anyway :P
20160223 15:55:21< gfgtdf> celticminstrel: when exactly is the function compiled ?
20160223 15:55:37< celticminstrel> I intend to do that in the filter's constructor.
20160223 15:56:07< Ravana_> being able to define the function exactly where it is used would be comfortable, but not vitally needed
20160223 15:57:46< celticminstrel> I haven't quite worked out how to keep track of it though.
20160223 15:58:28< zookeeper> if you can do it for no cost then it sounds good, but if you have to complicate the code noticeably or otherwise make some trade-offs, then probably not worth it
20160223 15:58:50< celticminstrel> I guess I'll give it a try and see how well it works, at least.
20160223 15:59:35< zookeeper> because as said it'd mainly be a minor convenience feature for someone who wouldn't have any problem working around the lack of it in the first place
20160223 16:00:41-!- louis94 [~~louis94@91.178.241.57] has joined #wesnoth-dev
20160223 16:02:08-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160223 16:12:27-!- horrowind [~Icedove@2a02:810a:8380:834:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20160223 16:19:46-!- Kwandulin [~Miranda@p200300760F0BC5BCD536C763B6EF4124.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160223 16:20:44< vultraz> celticminstrel: do you think there's a way to unify all the quit confirmation messages
20160223 16:21:40< celticminstrel> I thought gfgtdf already did that ages ago.
20160223 16:21:48< vultraz> well, no, actually
20160223 16:22:00< vultraz> kinda
20160223 16:22:08< vultraz> but there's still much duplicate functionality 
20160223 16:22:13< vultraz> also inconsistent behavior in the editor
20160223 16:22:15< vultraz> ie
20160223 16:22:38< vultraz> if you click the window X, it doesn't warn you about modified maps
20160223 16:22:54< gfgtdf> celticminstrel: no i only added a quit configmation mesage in case teh game was quitted by pressing th rd cross in the corrner
20160223 16:23:09< gfgtdf> vultraz: hm are you use it doesnt ?
20160223 16:23:23< vultraz> gfgtdf: no, it doesn't
20160223 16:24:29< vultraz> I'm thinking the quit_confirmation class could be made use of by all 'quit' message cases 
20160223 16:24:35< vultraz> Ima take a shot at it
20160223 16:24:58< gfgtdf> vultraz: hmm it doesnt show a quit conformation to me from the editor
20160223 16:25:50< vultraz> weird
20160223 16:27:10< gfgtdf> vultraz: well it unrelated to whether the map was changes though
20160223 16:28:10< vultraz> if the map is changed and you use the menu -> quit [] button it does warn you
20160223 16:28:12< vultraz> the red x does not
20160223 16:28:34< gfgtdf> vultraz: but ti shows te nromal quit confirmation ?
20160223 16:28:47< vultraz> yes
20160223 16:29:46< gfgtdf> vultraz: hmm ok
20160223 16:29:59< vultraz> I have an idea how to fix this
20160223 16:30:25< gfgtdf> vultraz: but the normal menu quit also show s aconfirmation if the map was not changes, i wonder whether it is needed
20160223 16:34:36< vultraz> gfgtdf: why are all the quit_confirmation members static?
20160223 16:35:16-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 16:35:28< vultraz> so they can be called without an object?
20160223 16:35:44< zookeeper> on an almost unrelated note, if you open a map with the editor without modifying the default map, it should not keep the latter open.
20160223 16:40:44< celticminstrel> I agree.
20160223 17:02:23-!- louis94 [~~louis94@91.178.241.57] has quit [Ping timeout: 244 seconds]
20160223 17:03:33< vultraz> well this is new
20160223 17:03:44< vultraz> "rror: 'class quit_confirmation quit_confirmation::quit_confirmation' is inaccessible|"
20160223 17:04:30< celticminstrel> I thought it was a static class.
20160223 17:04:39< celticminstrel> Why are you even calling the constructor?
20160223 17:05:27< vultraz> I'm... not
20160223 17:05:29< vultraz> I don't think so
20160223 17:06:00< celticminstrel> What are you doing, then?
20160223 17:06:57< vultraz> https://github.com/Vultraz/wesnoth/commit/9bce1276fb72f1be41f91791a33a9d4460b74b78
20160223 17:08:56< celticminstrel> vultraz: Don't forget that we don't want any confirmation in certain places, like at the title screen.
20160223 17:10:15< celticminstrel> What line is the error on?
20160223 17:11:17< vultraz> celticminstrel: first file 219
20160223 17:12:14-!- Kwandulin [~Miranda@p200300760F0BC5BCD536C763B6EF4124.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20160223 17:12:20< gfgtdf> celticminstrel: no its not a pureley static class. Its basiabally liske you can create 'quit_confirmation' object which will prefevent the game from quitting during their lifetime
20160223 17:12:35< celticminstrel> But not at any other place where you called show_prompt?
20160223 17:12:48< gfgtdf> celticminstrel: 'prevent' as in you it will show the 'are you sure..' primt
20160223 17:13:08< vultraz> celticminstrel: I don't know, since that's the first one it gets to
20160223 17:13:13< vultraz> s/it/the compiler
20160223 17:14:09-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160223 17:14:22-!- boucman_work [~jrosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20160223 17:18:26< Soliton> "if(quit_confirmation::show_prompt(get_video()) {" is a syntax error.
20160223 17:19:37< vultraz> ah, yes. thanks
20160223 17:20:05< celticminstrel> Missing close parenthesis, true.
20160223 17:20:37< gfgtdf> vultraz: did you test whether http://gna.org/bugs/?22263 is fixed ?
20160223 17:20:51< vultraz> gfgtdf: yes
20160223 17:26:23-!- SeamusFD [~SeamusFD@72-161-184-67.dyn.centurytel.net] has joined #wesnoth-dev
20160223 17:28:24< gfgtdf> github now allows to change files in the githup webpage via drag and drop, lets test this.
20160223 17:30:17< gfgtdf> "Something went really wrong, and we can’t process that file. "
20160223 17:34:42< vultraz> ughh
20160223 17:34:51< gfgtdf> vultraz: ?
20160223 17:34:59< vultraz> still getting that error
20160223 17:35:06< vultraz> only that case
20160223 17:37:25-!- atarocch [~atarocch@31.159.138.0] has quit [Remote host closed the connection]
20160223 17:39:46< vultraz> so it must be something specific to where it's being called
20160223 17:44:08< vultraz> hm
20160223 17:44:27< vultraz> if I add it as a public inheritor to editor_controller it builds but gives a warnings
20160223 17:44:44< vultraz> still, not what we want..
20160223 17:47:21< vultraz> hmmm
20160223 17:47:30-!- EdB [~edb@89.193.129.77.rev.sfr.net] has joined #wesnoth-dev
20160223 17:47:48< vultraz> why was controller_base inheriting 
20160223 17:48:10-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160223 17:48:39< vultraz> ok, remove that...
20160223 17:48:50< vultraz> "incomplete type used in nested name specifier"
20160223 17:50:11< vultraz> ok...
20160223 17:50:14< vultraz> think I got it
20160223 17:52:03< vultraz> YES
20160223 17:52:31< Aginor> morning
20160223 17:52:33< vultraz> unfortunately, now wesnoth doesn't launch at all :|
20160223 17:52:53< vultraz> I screwed up :|
20160223 17:53:36< celticminstrel> What does the console output say?
20160223 17:54:11< vultraz> I don't even get that far :P
20160223 17:54:18-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160223 17:54:35-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 17:54:35< vultraz> http://pastebin.com/h6T33pZY
20160223 17:54:37< vultraz> stacktrace
20160223 17:55:11< celticminstrel> That doesn't help.
20160223 17:55:50< celticminstrel> ...wait, #4.
20160223 17:56:39< celticminstrel> What is that function? Looks like something like quit_confirmation::count()? Is there such a function?
20160223 17:57:01< vultraz> count_ is a static int member
20160223 17:57:10< celticminstrel> Ah, that'd be it.
20160223 17:57:18< celticminstrel> Hmm.
20160223 17:57:29< celticminstrel> So, what's it initialized as?
20160223 17:57:39 * celticminstrel would assume 0.
20160223 17:57:46< vultraz> well there's no constructor but it's set to 0
20160223 17:57:56< celticminstrel> Of course there's no constructor.
20160223 17:58:07< celticminstrel> Hmm.
20160223 17:58:21< vultraz> maybe it has to do with my default_message_ variable
20160223 17:58:26< celticminstrel> Ohh.
20160223 17:58:46< vultraz> (I made it non-const locally) 
20160223 17:59:09< vultraz> but maybe using it as a default value isn't allowed?
20160223 17:59:45< celticminstrel> No.
20160223 17:59:47< celticminstrel> It's not that.
20160223 17:59:52< celticminstrel> I think I know what it is.
20160223 18:00:09< celticminstrel> You've asked the game to resolve a translated string before translations have been initialized.
20160223 18:00:18< celticminstrel> Try declaring it a t_string instead of a std::string.
20160223 18:00:53< celticminstrel> Also, does the default message need to be obtainable by anything else?
20160223 18:01:07< celticminstrel> If not, you could even just declare it as a constant in the source file.
20160223 18:01:37< vultraz> no, the default message is only for the default value of the second argument of show_prompt
20160223 18:02:01< vultraz> hm...t_string not changing anything
20160223 18:02:07< celticminstrel> Ah, so the header needs to know about it, unless you have two overloads of show_prompt.
20160223 18:02:27< celticminstrel> Actually, why not do that?
20160223 18:02:40< celticminstrel> Instead of having a show_prompt with a default value, overload it.
20160223 18:03:00< celticminstrel> Then you can inline the translated string nicely.
20160223 18:04:18-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Ping timeout: 276 seconds]
20160223 18:04:26< vultraz> that works
20160223 18:04:54< celticminstrel> Okay, so basically: never make a translatable string static.
20160223 18:05:16< celticminstrel> (Function-static is fine, but not class or global static.)
20160223 18:05:22< vultraz> good to know
20160223 18:05:33< vultraz> dammit...
20160223 18:05:46< vultraz> now the prompt won't popup at all :(
20160223 18:06:07< celticminstrel> By which you mean, the game just closes without warning?
20160223 18:06:23< vultraz> yes
20160223 18:07:06< celticminstrel> The open_ guard should probably be moved into show_prompt.
20160223 18:07:23< celticminstrel> Since its purpose is to make sure a quit confirmation is not put up when a quit confirmation is onscreen.
20160223 18:07:42< celticminstrel> I doubt that's causing the issue you're seeing, though.
20160223 18:08:52< vultraz> (OT, but I need to create another function that throws quit_game_exception())
20160223 18:09:37< vultraz> maybe there was a reason controller base was inheriting from quit_confirmation...
20160223 18:10:09< celticminstrel> Yes, there was.
20160223 18:10:31< celticminstrel> By inheriting from quit confirmation, it increases the "refcount", thus enabling the dialog.
20160223 18:10:36< celticminstrel> Or something like that.
20160223 18:10:48< celticminstrel> Basically, when no quit confirmation object exists, the dialog is skipped.
20160223 18:11:58< celticminstrel> This is why there's no confirmation from the title screen.
20160223 18:13:13-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160223 18:19:55-!- stikonas_ is now known as stikonas
20160223 18:27:08-!- irker806 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20160223 18:42:04< gfgtdf> vultraz: i thinkk teh best wy is to have a boost::function<bool()> callback in the quit_confirmation object and istrad of having a statis int cont_ have a static stack of quit_confirmation*
20160223 18:42:50< gfgtdf> so that the editor can construct its quit_confirmation  with somethign like boost::bind(editor_controllrer::quit_configmation, this)
20160223 18:43:04-!- SeamusFD is now known as SeamusFD[Off]
20160223 18:43:14-!- SeamusFD[Off] [~SeamusFD@72-161-184-67.dyn.centurytel.net] has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
20160223 18:48:24-!- irker437 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160223 18:48:24< irker437> wesnoth: Celtic Minstrel wesnoth:master 708af5640953 / src/scripting/ (game_lua_kernel.cpp lua_api.cpp lua_api.hpp): Fix function name typo https://github.com/wesnoth/wesnoth/commit/708af564095390490048d04fbd40709cbcda9185
20160223 18:52:02< celticminstrel> gfgtdf, zookeeper: This is roughly what it would look like (though this version doesn't compile):
20160223 18:52:03< celticminstrel> https://github.com/CelticMinstrel/wesnoth/commit/d3d515b1f52689e31af17fd463b8046e72c93542
20160223 18:52:15< celticminstrel> Should I keep working on it, or do you think it's not worth it?
20160223 18:52:36< celticminstrel> If not, I'd like to at least allow lua_function="some_table.some_func" to work.
20160223 18:54:30< gfgtdf> celticminstrel: i think a same ple wml code how o use it woudl help
20160223 18:54:51-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20160223 18:55:34< celticminstrel> Maybe something like [lua]code=<<local u = ...; return u.health < u.max_health / 2;>>[/lua]
20160223 18:55:42< gfgtdf> vultraz: do you plan to keep the confirmation message in case no changes were made? Because i dont realyl think this is useful
20160223 18:56:29< gfgtdf> celticminstrel: i assume variable substtution (&this_unit)wont work inside lua filters?
20160223 18:57:20< celticminstrel> Variable substitution will work in [args] but not in code=
20160223 18:57:35< celticminstrel> this_unit is passed as the first argument, hence that "local" declaration.
20160223 18:57:36< gfgtdf> celticminstrel: right this makes sense
20160223 19:00:53< gfgtdf> celticminstrel: well yes i know that the lua_function annoyed me whne i started when i writing my addon. But im also quite worring about performance issues, for example i know that we removed some formula=-.. filters from the ai code becasue of performance issues. So if you know that your changes don't have a big performane impact im fine with it.
20160223 19:01:05< gfgtdf> celticminstrel: also did you solve the const unit issues ?
20160223 19:02:22< celticminstrel> Yes.
20160223 19:02:35< celticminstrel> I just didn't use luaW_pushlocalunit.
20160223 19:03:18< celticminstrel> gfgtdf: So, do you think it's simple enough code-wise to go ahead and get it working to see if it's fine performance-wise?
20160223 19:04:15< gfgtdf> celticminstrel: hmm the point of luaW_pucklocalunit is that it (unlike usualy lua units) also work with unit that are currently not on the map.
20160223 19:04:41< celticminstrel> Honestly, the lua_unit class is really confusing now.
20160223 19:04:50< gfgtdf> celticminstrel: y i know
20160223 19:09:24-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has quit [Ping timeout: 250 seconds]
20160223 19:11:21-!- vn971 [~vasya@91.247.233.240] has quit [Quit: Leaving.]
20160223 19:15:02< zookeeper> great, now even building 1.12 fails...
20160223 19:15:35< celticminstrel> gfgtdf, zookeeper: Given rebasing, is it okay to merge PR 596?
20160223 19:15:54< gfgtdf> celticminstrel: invalid uncode characters again ...
20160223 19:16:00< celticminstrel> ?
20160223 19:16:19< gfgtdf> celticminstrel: you last message convtained in valid character
20160223 19:16:23< gfgtdf> contained*
20160223 19:16:27< celticminstrel> oh.
20160223 19:16:28< gfgtdf> an*
20160223 19:16:35< celticminstrel> Do you have an answer to the question though?
20160223 19:16:40< gfgtdf> "merge[0x1d] PR 596?"
20160223 19:16:47< celticminstrel> Sorry.
20160223 19:17:01< gfgtdf> celticminstrel: well yes i if it rebased into 1 or 2 commit it can be merged i think
20160223 19:17:05-!- vn971 [~vasya@91.247.233.240] has joined #wesnoth-dev
20160223 19:18:03< gfgtdf> vultraz: still wokring on the quit confirmation thing?
20160223 19:18:07< gfgtdf> vultraz: i can also do that
20160223 19:18:32< zookeeper> celticminstrel, if it works as advertised, then sure. could use some indentation-fixing too, though.
20160223 19:18:36< irker437> wesnoth: Robert Wallace wesnoth:master afd53cf6a1b2 / data/multiplayer/scenarios/ANL_utils/ (4 files): Fixed deprecated code in ANL https://github.com/wesnoth/wesnoth/commit/afd53cf6a1b20b9f5efc6143274faa888a308a3f
20160223 19:18:38< irker437> wesnoth: Robert Wallace wesnoth:master df45dd9aab03 / data/core/about.cfg: Update about.cfg https://github.com/wesnoth/wesnoth/commit/df45dd9aab034ff08c0af27ac087bf986dddc603
20160223 19:18:54< gfgtdf> celticminstrel: did you replace it for him ?
20160223 19:21:24< gfgtdf> celticminstrel: rebase*
20160223 19:24:15< gfgtdf> celticminstrel: dont forget to close the pr after meging manually (and _dont_ press the merge button)
20160223 19:24:34< celticminstrel> Yeah, I was in the process of writing a comment before closing.
20160223 19:25:06< celticminstrel> And yeah, I rebased it.
20160223 19:25:13< celticminstrel> That's why there's no merge commit.
20160223 19:26:22< celticminstrel> zookeeper: Anything to say about PR604?
20160223 19:26:47< celticminstrel> Should we close PR585?
20160223 19:30:39< zookeeper> i'd say yes.
20160223 19:30:54< zookeeper> because clearly we're not going to take it
20160223 19:36:53< irker437> wesnoth: gfgtdf wesnoth:master 0484fdd4149f / src/ (9 files in 4 dirs): fix quit confirmation promt in editor https://github.com/wesnoth/wesnoth/commit/0484fdd4149fc794d59517f7a19873bbb5fe6ad6
20160223 19:37:33< gfgtdf> vultraz, celticminstrel: i made a commit abotu the quit conformation thing ^
20160223 19:47:05-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160223 19:52:20-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160223 20:11:12-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160223 20:14:59-!- Alduin_ [~Alduin@2001:4ca0:4fff:4::84] has quit [Ping timeout: 240 seconds]
20160223 20:25:33-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20160223 20:25:34-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20160223 20:41:49< irker437> wesnoth: Celtic Minstrel wesnoth:master baed863df073 / src/ (27 files in 10 dirs): Don't include config.hpp when possible https://github.com/wesnoth/wesnoth/commit/baed863df07360f8753c81d160c3ea202da681c5
20160223 20:41:51< irker437> wesnoth: Celtic Minstrel wesnoth:master c08ea95c7256 / src/config.hpp: config: make static string constants private https://github.com/wesnoth/wesnoth/commit/c08ea95c72566614a5d44a347e2517f4491d0e70
20160223 21:02:59-!- gfgtdf [~chatzilla@f054060000.adsl.alicedsl.de] has quit [Read error: Connection reset by peer]
20160223 21:07:36-!- EdB [~edb@89.193.129.77.rev.sfr.net] has quit [Quit: Konversation terminated!]
20160223 21:10:27-!- aidanhs [~aidanhs@81.4.110.234] has quit [Ping timeout: 248 seconds]
20160223 21:29:38-!- aidanhs [~aidanhs@81.4.110.234] has joined #wesnoth-dev
20160223 21:40:56< vultraz> celticminstrel: ahhhh reducing config includes :D 
20160223 21:41:07< celticminstrel> ?
20160223 21:41:18< vultraz> baed863df073 
20160223 21:41:26< celticminstrel> What about it?
20160223 21:41:46< vultraz> it's a good commit
20160223 21:42:20< vultraz> config.hpp is like display.hpp - a huge file better off included as little as possible 
20160223 21:43:00< celticminstrel> display.hpp, huh? Maybe I should do the same with that later. Unless someone already did.
20160223 21:44:52< vultraz> Not sure
20160223 21:45:08< vultraz> Either way, I see gfgtdf did the quit confirmation thing
20160223 21:45:12< vultraz> Nice of him, I guess.
20160223 21:47:09-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160223 21:47:44< vultraz> still want to do some cleanup
20160223 21:54:20< vultraz> celticminstrel: dammit. getting errors about strcoll not being declarded in help/help-impl.hpp
20160223 21:54:23< vultraz> what is strcoll
20160223 21:57:37< vultraz> seems putting back the config include fixes that
20160223 22:00:28< irker437> wesnoth: Charles Dang wesnoth:master c0df87a47049 / src/help/help_impl.hpp: Restore a config.hpp include https://github.com/wesnoth/wesnoth/commit/c0df87a4704924ba8331858e595090a74da1a638
20160223 22:01:50< celticminstrel> Uhh, why?
20160223 22:02:03< celticminstrel> You don't need config.h for that.
20160223 22:02:43< vultraz> well it fixed it
20160223 22:02:46< vultraz> don't ask me
20160223 22:02:51< celticminstrel> vultraz: strcoll is in <cstring>
20160223 22:03:01< celticminstrel> So include that instead.
20160223 22:03:01< vultraz> oh
20160223 22:03:50< celticminstrel> Not sure where config.hpp gets it. It doesn't include it directly.
20160223 22:04:13< celticminstrel> Maybe through <string>. That would make sense.
20160223 22:04:51-!- fabi [~quassel@176.2.68.235] has joined #wesnoth-dev
20160223 22:04:51-!- fabi [~quassel@176.2.68.235] has quit [Changing host]
20160223 22:04:51-!- fabi [~quassel@wesnoth/developer/fendrin] has joined #wesnoth-dev
20160223 22:05:19-!- gfgtdf [~chatzilla@f054060000.adsl.alicedsl.de] has joined #wesnoth-dev
20160223 22:05:22< irker437> wesnoth: Charles Dang wesnoth:master 56f0c9a99a6d / src/help/help_impl.hpp: Redid c0df87a47049 correctly https://github.com/wesnoth/wesnoth/commit/56f0c9a99a6d42359db2407acc367f6e5f5863ee
20160223 22:05:46< gfgtdf> celticminstrel: i really wonder whether strcoll is te right way to do it
20160223 22:06:03< celticminstrel> I dunno.
20160223 22:06:17< gfgtdf> celticminstrel: usually i'd think that there is also a c++ way to do its that used std::locale/boost::locale
20160223 22:06:29< celticminstrel> Probably!
20160223 22:06:47< gfgtdf> vultraz: do you ever compare strings in lua ?
20160223 22:06:49< irker437> wesnoth: Charles Dang wesnoth:master c40c2cd9af52 / data/multiplayer/scenarios/ANL_utils/ (ANL_research_options.cfg ANL_worker_options.cfg): ANL: strip unnecessary wmltools guards and ran wmlindent https://github.com/wesnoth/wesnoth/commit/c40c2cd9af52d1dbfecbad38684b92ce3dd5ba8b
20160223 22:06:57< celticminstrel> I think that would be in std::locale somewhere, probably.
20160223 22:07:02< vultraz> celticminstrel: ^ would have saved Robert a lot of trouble :|
20160223 22:07:20< vultraz> gfgtdf: uh... yes?
20160223 22:07:25< vultraz> don't you usually?
20160223 22:07:27< celticminstrel> wmlindent caused him triuble.
20160223 22:07:30< celticminstrel> ^o
20160223 22:07:50< vultraz> celticminstrel: no, I mean the 'start ignoring' bits
20160223 22:07:55< gfgtdf> vultraz: could you give me a code example ?
20160223 22:08:17< celticminstrel> Yeah, doesn't seem like those have a good reason for them.
20160223 22:08:37< celticminstrel> gfgtdf: Did you try "str1 < str2"?
20160223 22:08:52< vultraz> oh that's what you mean
20160223 22:08:57< gfgtdf> celticminstrel: well yes i know how to do it
20160223 22:09:03< gfgtdf> celticminstrel: but i want usecase examples
20160223 22:09:14< celticminstrel> Eh?
20160223 22:09:22< gfgtdf> becasue seems like lua internally uses strcoll to compare strings which could cause OOS in some situations and i wonder whethe ri should change it to strcmp
20160223 22:09:31< celticminstrel> Ohhh.
20160223 22:09:35< gfgtdf> so i want to know how it is usually used
20160223 22:09:43< celticminstrel> I see.
20160223 22:10:05< vultraz> gfgtdf: well I didn't write this bit, 8680 did but: "table.sort(inv_list_data, function(a,b) return tostring(a.name) < tostring(b.name) end)"
20160223 22:10:07< vultraz> is that what you mean
20160223 22:11:01< celticminstrel> Since we have translatable strings in Lua, it might be okay to use strcmp instead of strcoll, but I wonder if it might confuse people used to proper collation...
20160223 22:11:27-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160223 22:11:38< celticminstrel> Actually, are the comparison operators overloaded for translatable strings in Lua?
20160223 22:12:46< gfgtdf> celticminstrel: hmm y making it only use locale dpeended orders for tstring sounds like a good idea
20160223 22:12:57-!- horrowind [~Icedove@2a02:810a:8380:834:21b:fcff:fee3:c3ff] has quit [Quit: horrowind]
20160223 22:14:43< gfgtdf> "attempt to compare 2 userdata values"
20160223 22:15:07< celticminstrel> I think that implies that the required metafunctions aren't implemented.
20160223 22:15:20< celticminstrel> So, you'd need to implement them.
20160223 22:15:29< celticminstrel> In C++, of course.
20160223 22:16:53< celticminstrel> Yeah, it looks like they only implemented __contact, __gc, and __tostring.
20160223 22:16:58< celticminstrel> ^__concat
20160223 22:17:17< celticminstrel> lua_common.cpp, if you were wondering.
20160223 22:20:20< vultraz> hm
20160223 22:21:22< vultraz> I wonder if it'd be possible to give quit_confirmation::quit() a different exception to fire 
20160223 22:21:50< celticminstrel> Why?
20160223 22:22:10< vultraz> like if I wanted it to throw quit_game_exception
20160223 22:22:16< vultraz> instead of CVideo::quit
20160223 22:22:18< vultraz> sometimes
20160223 22:23:33< celticminstrel> Do you have a particular reason for wanting to use it for "quit to title"?
20160223 22:23:52-!- travis-ci [~travis-ci@ec2-50-19-40-148.compute-1.amazonaws.com] has joined #wesnoth-dev
20160223 22:23:53< travis-ci> wesnoth/wesnoth#8575 (master - c08ea95 : Celtic Minstrel): The build has errored.
20160223 22:23:53< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/111312415
20160223 22:23:53-!- travis-ci [~travis-ci@ec2-50-19-40-148.compute-1.amazonaws.com] has left #wesnoth-dev []
20160223 22:24:53< vultraz> keeping code in one place and such. 
20160223 22:25:21 * vultraz is still trying to understand gfgtdf's changes
20160223 22:29:25< celticminstrel> If I read that right, it would be possible to get two (or more) prompts in a row.
20160223 22:29:49< celticminstrel> Is that intentional, gfgtdf?
20160223 22:31:23< gfgtdf> celticminstrel: y currently there is afaik always at most 1 promt so i dint think about it that much.
20160223 22:31:37< vultraz> gfgtdf: I'm thinking command_executor::quit_to_desktop() could be replaced with a call to guit_confirmation::quit()
20160223 22:32:48< gfgtdf> vultraz: y i see no reason against that
20160223 22:33:11< celticminstrel> Can it be replaced?
20160223 22:33:42< celticminstrel> If not, at least forwarding to quit_confirmation::quit() seems perfectly reasonable.
20160223 22:34:28< vultraz> celticminstrel: I was just going to put it in 'case HOTKEY_QUIT_TO_DESKTOP'
20160223 22:34:35< vultraz> line 315
20160223 22:35:29< celticminstrel> So it is replaceable then.
20160223 22:35:51< vultraz> and then remove the two local functions
20160223 22:36:17< celticminstrel> ""local" functions?
20160223 22:36:23< celticminstrel> You mean member functions, right?
20160223 22:36:35< vultraz> er
20160223 22:36:37< vultraz> yes
20160223 22:36:38< vultraz> sorry
20160223 22:36:40< vultraz> just woke up
20160223 22:39:24< mattsc> Aginor, celticminstrel, vultraz (I forgot if somebody else was involved in this yesterday): I had a brief exchange with Alarantalara about the Experimental AI.  He really wrote most of that AI, I usually just play with Fred.
20160223 22:40:06< mattsc> He reminded me that the ExpAI has problems on certain maps, which makes it actually play worse on those than the default.
20160223 22:40:24< celticminstrel> I guess that's why it's not yet the default.
20160223 22:40:28< celticminstrel> Which maps?
20160223 22:40:44< mattsc> E.g. Aethermaw and Sablestone Delta; and he suspects that it won’t do well on survival maps either.
20160223 22:41:06< celticminstrel> What about them causes problems?
20160223 22:41:25< mattsc> Those could likely be fixed, but only with more intensive testing that we are getting at the moment.
20160223 22:41:41< mattsc> “Aethermaw is a problem because the correct hex to recruit is not one closest to the enemy and so it often has units get killed piecemeal, and Sablestone Delta encourages it to isolate its leader with the large number of keeps.”
20160223 22:42:32< mattsc> So yes, that’s one reason why it’s not the default; the other is that it is simply not been tested extensively enough to know whether there are other problems we have not found yet.
20160223 22:43:29< mattsc> Alarantalara suggested that one way to deal with this is to add a key that can be used tp the default AI for a MP map and use this for some of the “standard” maps.
20160223 22:43:44< celticminstrel> Huh?
20160223 22:44:10< celticminstrel> Basically, something like "ai_algorithm=ai_default"? (Hypothetical syntax)
20160223 22:44:17< mattsc> yes
20160223 22:44:21< celticminstrel> (Did my commit add that? I can't remember.)
20160223 22:44:22< mattsc> I guess we have a key already!
20160223 22:44:33< celticminstrel> We have a key, but I don't think it does that at the moment.
20160223 22:44:35< mattsc> No, that’s been around for a long time, but it’s buggy.
20160223 22:44:44< celticminstrel> Maybe it does. I dunno.
20160223 22:44:51< irker437> wesnoth: Charles Dang wesnoth:master 2346b3508f8d / src/hotkey/ (command_executor.cpp command_executor.hpp): Simplify handling of HOTKEY_QUIT_TO_DESKTOP https://github.com/wesnoth/wesnoth/commit/2346b3508f8d58b0581ad0485b63014094faea67
20160223 22:44:53< mattsc> Yes, and in it’s current implementation I think it does not.
20160223 22:44:54< irker437> wesnoth: Charles Dang wesnoth:master b73cc5eb08f4 / src/ (quit_confirmation.cpp quit_confirmation.hpp): Small formatting cleanup https://github.com/wesnoth/wesnoth/commit/b73cc5eb08f4a26f4e7284c1236fe4dc73bb2d22
20160223 22:45:00< celticminstrel> Why is it buggy again? I feel like you said this yesterday.
20160223 22:45:05< celticminstrel> Aww, Vultraz beat me.
20160223 22:45:11< mattsc> I looked into the C++ implementation of it at some point, and it’s essentially har-coded to only work for the idel_ai
20160223 22:45:15< vultraz> celticminstrel: beat you to what?
20160223 22:45:17< celticminstrel> Now I have to deal with merge conflicts.
20160223 22:45:20< mattsc> or something like that
20160223 22:46:03< mattsc> It’s buggy in that you can only use it once in a scenario (or maybe only at the start of the scenario), but than cannot change it back to normal.
20160223 22:46:13< mattsc> Or something like that, I don’t remember the details either.
20160223 22:46:16< celticminstrel> I'm not sure if I'd call that buggy.
20160223 22:47:26< mattsc> celticminstrel: https://gna.org/bugs/?func=detailitem&item_id=16107
20160223 22:47:42< mattsc> It’s only been open for 6 years!
20160223 22:47:57-!- gfgtdf [~chatzilla@f054060000.adsl.alicedsl.de] has quit [Read error: Connection reset by peer]
20160223 22:50:25< celticminstrel> Was controller=null recently removed?
20160223 22:50:30< mattsc> also this: https://gna.org/bugs/?func=detailitem&item_id=21294
20160223 22:50:46< mattsc> Umm, it better had not, it’s used all over the place.
20160223 22:51:04< mattsc> But controller=null and idle_ai are not equivalent anyway.
20160223 22:51:18< celticminstrel> What's the difference?
20160223 22:51:20< mattsc> The null controller also disables all side events etc.
20160223 22:51:48< mattsc> plus income and upkeep etc., I think, but I am not 100% sure about that.
20160223 22:53:15< mattsc> Anyways, back to the Exp AI: before we decide how to do it, I think the question is whether we want to make it the default for some or all maps
20160223 22:53:59< vultraz> celticminstrel: you're not working on adding game_quit_exception to g_c::quit are you?
20160223 22:54:21< celticminstrel> vultraz: No
20160223 22:54:34< celticminstrel> But I was doing something in the same file.
20160223 22:54:35< vultraz> ok
20160223 22:57:13< irker437> wesnoth: Celtic Minstrel wesnoth:master f067332d0ff5 / src/ (4 files in 3 dirs): fix spelling https://github.com/wesnoth/wesnoth/commit/f067332d0ff5c78541f5cb31c4cdff47f83e61cb
20160223 22:57:17< celticminstrel> Hence, merge conflicts.
20160223 22:57:30< celticminstrel> mattsc: Do you think the version key is still needed?
20160223 22:57:49< vultraz> You undid my format fixup D:
20160223 22:58:01< celticminstrel> Did I? Sorry about that.
20160223 22:58:27-!- Netsplit *.net <-> *.split quits: irker437, Ivanovic, celticminstrel, enchi, aidanhs, vultraz, zookeeper, _laco
20160223 22:58:40-!- Netsplit *.net <-> *.split quits: ancestral, vn971, prkc, kodomazer, EliDupree, bumbadadabum, Soliton
20160223 22:59:02-!- Netsplit *.net <-> *.split quits: Gambit, stikonas, Elsi, aeth, shikadibot, AI0867, knotwork__, kidneb
20160223 22:59:22-!- Netsplit *.net <-> *.split quits: Rhonda, higgins, pydsigner, mattsc, fabi, Crendgrim, DDR, Jetrel, timotei_, Samual
20160223 22:59:40-!- Netsplit *.net <-> *.split quits: vincent_c, iwaim__, janebot, oldlaptop, quentinp, minzbonbon, TC01, TheJJ, Greg-Boggs, mjs-de,  (+1 more, use /NETSPLIT to show all of them)
20160223 22:59:47-!- Netsplit *.net <-> *.split quits: tomreyn
20160223 22:59:51-!- Netsplit *.net <-> *.split quits: legoktm, Greywhind, {V}
20160223 23:00:21-!- Netsplit *.net <-> *.split quits: Jetrel_bot, loonycyborg, Aginor, matthiaskrgr, Lohengramm, new_one, molgrum, Ravana_
20160223 23:00:34-!- celticminstrel [~celmin@bas1-kitchener09-1176182734.dsl.bell.ca] has joined #wesnoth-dev
20160223 23:02:27-!- celticminstrel [~celmin@bas1-kitchener09-1176182734.dsl.bell.ca] has quit [Client Quit]
20160223 23:04:32-!- _laco [~laco@static.95.25.4.46.clients.your-server.de] has joined #wesnoth-dev
20160223 23:04:32-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20160223 23:04:32-!- vincent_c [~bip@107.191.117.101] has joined #wesnoth-dev
20160223 23:04:32-!- midzer [~quassel@pD9FA2D72.dip0.t-ipconnect.de] has joined #wesnoth-dev
20160223 23:04:32-!- higgins [~higgins@105.ip-167-114-152.net] has joined #wesnoth-dev
20160223 23:04:32-!- TheJJ [~rofl@ipbcc36ea9.dynamic.kabel-deutschland.de] has joined #wesnoth-dev
20160223 23:04:32-!- quentinp [~quentin@ns363174.ip-91-121-196.eu] has joined #wesnoth-dev
20160223 23:04:32-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20160223 23:04:32-!- pydsigner [~pydsigner@unaffiliated/pydsigner] has joined #wesnoth-dev
20160223 23:04:32-!- iwaim__ [~iwaim@rasteenie.alib.jp] has joined #wesnoth-dev
20160223 23:04:32-!- Crendgrim [~crend@wesnoth/forum-moderator/crendgrim] has joined #wesnoth-dev
20160223 23:04:32-!- minzbonbon [~min@meta23.net] has joined #wesnoth-dev
20160223 23:04:32-!- oldlaptop [~quassel@192.183.14.123] has joined #wesnoth-dev
20160223 23:04:32-!- TC01 [~quassel@london.acm.jhu.edu] has joined #wesnoth-dev
20160223 23:04:32-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20160223 23:04:32-!- Jetrel [~Jetrel@c-73-228-139-39.hsd1.mn.comcast.net] has joined #wesnoth-dev
20160223 23:04:32-!- DDR [~david@ec2.happyspork.com] has joined #wesnoth-dev
20160223 23:04:32-!- fabi [~quassel@wesnoth/developer/fendrin] has joined #wesnoth-dev
20160223 23:04:32-!- timotei_ [~timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20160223 23:04:32-!- Rhonda [~rhonda@wesnoth/developer/rhonda] has joined #wesnoth-dev
20160223 23:04:32-!- knotwork__ [~markm@99.192.83.208] has joined #wesnoth-dev
20160223 23:04:32-!- aeth [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev
20160223 23:04:32-!- Gambit [~derek@wesnoth/developer/grickit] has joined #wesnoth-dev
20160223 23:04:32-!- Elsi [~Elsi@luwin.ulrar.net] has joined #wesnoth-dev
20160223 23:04:32-!- kidneb [kidneb@not.allthetime.xyz] has joined #wesnoth-dev
20160223 23:04:32-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20160223 23:04:32-!- mattsc [~mattsc@fw.hia.nrc.ca] has joined #wesnoth-dev
20160223 23:04:32-!- matthiaskrgr [matthiaskr@gateway/shell/panicbnc/x-qooiscroglnzlumb] has joined #wesnoth-dev
20160223 23:04:32-!- Lohengramm [sid1929@gateway/web/irccloud.com/x-adtjhodchgggyaux] has joined #wesnoth-dev
20160223 23:04:32-!- Aginor [~andreas@unaffiliated/aginor] has joined #wesnoth-dev
20160223 23:04:32-!- new_one [~new_one@2604:a880:1:20::22e:d001] has joined #wesnoth-dev
20160223 23:04:32-!- enchi [enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20160223 23:04:32-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20160223 23:04:32-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20160223 23:04:32-!- irker437 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20160223 23:04:32-!- aidanhs [~aidanhs@81.4.110.234] has joined #wesnoth-dev
20160223 23:04:32-!- AI0867 [~ai@wesnoth/developer/ai0867] has joined #wesnoth-dev
20160223 23:04:32-!- shikadibot [~shikadi@wesnoth/umc-dev/bot/shikadibot] has joined #wesnoth-dev
20160223 23:04:32-!- Soliton [~Soliton@wesnoth/developer/soliton] has joined #wesnoth-dev
20160223 23:04:32-!- EliDupree [~quassel@idupree.com] has joined #wesnoth-dev
20160223 23:04:32-!- prkc [~prkc@46.166.190.188] has joined #wesnoth-dev
20160223 23:04:32-!- vn971 [~vasya@91.247.233.240] has joined #wesnoth-dev
20160223 23:04:32-!- bumbadadabum [~bumbadada@wesnoth/developer/bumbadadabum] has joined #wesnoth-dev
20160223 23:04:32-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has joined #wesnoth-dev
20160223 23:04:32-!- molgrum [~molgrum@unaffiliated/molgrum] has joined #wesnoth-dev
20160223 23:04:32-!- legoktm [~quassel@wikipedia/Legoktm] has joined #wesnoth-dev
20160223 23:04:32-!- {V} [~V@105-70-ftth.on.nl] has joined #wesnoth-dev
20160223 23:04:32-!- Greywhind [~Greywhind@c-50-133-231-228.hsd1.ma.comcast.net] has joined #wesnoth-dev
20160223 23:04:45-!- loonycyborg [~loonycybo@wesnoth/developer/loonycyborg] has joined #wesnoth-dev
20160223 23:04:45-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has joined #wesnoth-dev
20160223 23:04:45-!- Jetrel_bot [~Jetrel@ec2.happyspork.com] has joined #wesnoth-dev
20160223 23:04:50-!- kodomazer [62f86f38@gateway/web/freenode/ip.98.248.111.56] has joined #wesnoth-dev
20160223 23:04:50-!- kidneb [kidneb@not.allthetime.xyz] has quit [Max SendQ exceeded]
20160223 23:04:54-!- kidneb [kidneb@not.allthetime.xyz] has joined #wesnoth-dev
20160223 23:04:57-!- mattsc [~mattsc@fw.hia.nrc.ca] has quit [Client Quit]
20160223 23:06:21-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 23:06:22-!- mjs-de [~mjs-de@f048098199.adsl.alicedsl.de] has joined #wesnoth-dev
20160223 23:06:22-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20160223 23:07:22-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev
20160223 23:07:28-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev
20160223 23:08:04< celticminstrel> Did I miss anything?
20160223 23:08:11< celticminstrel> eg, did vultraz or mattsc say anything?
20160223 23:08:14< mattsc> are we back yet?
20160223 23:08:21< vultraz> we're back
20160223 23:08:25< vultraz> didn't say anything
20160223 23:08:27< mattsc> I had started to say something before I got kicked out.
20160223 23:08:28< celticminstrel> Seems like it.
20160223 23:08:44< zookeeper> celticminstrel, i just realized: when you rewrote the leadership ability to use a single tag, you seem to have forgotten the do the same to THoT's inspire ability.
20160223 23:08:54< zookeeper> s/the/to
20160223 23:08:56< mattsc> celticminstrel: I would get rid of the version key and just make sure that both standard and composite aspect syntax always work
20160223 23:09:01< celticminstrel> zookeeper: I didn't know about it.
20160223 23:10:23-!- louis94 [~~louis94@91.178.241.57] has joined #wesnoth-dev
20160223 23:10:33< mattsc> Crab himself told me a long time ago that he sees no reason not to do so any more, he just never got around to changing the code.
20160223 23:13:28< vultraz> celticminstrel: from a purely design standpoint, would you say a bool argument to q_c::quit is sufficient or should I make it a named enum or something
20160223 23:13:38< celticminstrel> Depends what it's for.
20160223 23:13:51< vultraz> 	if(!full_exit) {
20160223 23:13:52< vultraz> 		throw_quit_game_exception();
20160223 23:13:54< vultraz> 	}
20160223 23:13:55< vultraz> 	throw CVideo::quit();
20160223 23:13:57< vultraz> (full_exit defaults to true)
20160223 23:14:03-!- gfgtdf [~chatzilla@f054060000.adsl.alicedsl.de] has joined #wesnoth-dev
20160223 23:14:09< celticminstrel> Eh.
20160223 23:14:15< celticminstrel> Well, do whatever you want, I guess.
20160223 23:14:16< gfgtdf> vultraz: why do you want to adda bolean parameter ?
20160223 23:14:36< gfgtdf> did someone recently changes config.hpp ot t_string.hpp ?
20160223 23:14:44< celticminstrel> Yes, I did.
20160223 23:14:58< celticminstrel> But I also recently reduced the number of headers that include config.hpp.
20160223 23:15:10< vultraz> gfgtdf: so using quit_confirmation::quit(false) would just quit from game not from app fully
20160223 23:16:09< gfgtdf> celticminstrel: hmm i'd still huess that most cpp files include config in one way or another
20160223 23:16:26-!- Appleman1234 [~Appleman1@KD119104016245.au-net.ne.jp] has joined #wesnoth-dev
20160223 23:16:46< celticminstrel> Maybe. It's kinda hard to say.
20160223 23:16:54< gfgtdf> celticminstrel: also it seems liek msvc thinks like "so many files were changed we should make a complete rebuild" when config.hpp gets changed.
20160223 23:17:10< celticminstrel> Really?
20160223 23:17:53< gfgtdf> celticminstrel: well at lest for the that the only explanation why it alsp rebuild the files in lua/ directoy
20160223 23:18:02< gfgtdf> celticminstrel: which dont use config objects afaik
20160223 23:18:13< celticminstrel> Some of them do, certainly.
20160223 23:18:21< gfgtdf> celticminstrel: which one for example ?
20160223 23:18:51< celticminstrel> Uhh, the file that deals with converting configs to tables.
20160223 23:18:59< gfgtdf> celticminstrel: those arent in lua/
20160223 23:19:06< celticminstrel> Ohhh, right, duh.
20160223 23:19:15< vultraz> gfgtdf: basically i want to add a bool value so q_c can handle more than one type of 'quitting'
20160223 23:19:17< celticminstrel> I get confused about that sometimes.
20160223 23:21:11< vultraz> gfgtdf: AFAIK there're only two 'quit modes', right?
20160223 23:24:18< gfgtdf> celticminstrel: well i dont know how the ditor quits
20160223 23:24:29< gfgtdf> vultraz: ^
20160223 23:24:48< vultraz> the editor has its own thing
20160223 23:25:02< vultraz> runs play_slice() as long as do_quit_ is false
20160223 23:25:05< vultraz> so I can't affect that
20160223 23:25:13< celticminstrel> ...what does turns= do in an aspect?
20160223 23:26:16< celticminstrel> mattsc: ^
20160223 23:26:56-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160223 23:27:09< mattsc> celticminstrel: if set, the aspect is only valid during those turns
20160223 23:27:17< celticminstrel> So like, turns=1,2,3?
20160223 23:27:31< mattsc> yes
20160223 23:27:34< celticminstrel> Or turns=12-20?
20160223 23:27:35-!- ancestral [~ancestral@97-116-184-84.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20160223 23:27:57-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 23:28:12< mattsc> I believe that both of those work, yes, but I have not confirmed it in a while.
20160223 23:28:28< mattsc> (It used not to work at all because of a bug, but I am pretty sure that that was fixed)
20160223 23:28:43< gfgtdf> oh it finshed compliling
20160223 23:29:15< celticminstrel> I'm looking over my code that expands simplified aspects.
20160223 23:29:49< mattsc> celticminstrel: ok; I’ll have to leave shortly; I’ll be back again later tonight
20160223 23:29:58< celticminstrel> I think it does (almost?) everything the upgrade_blah_blah_blah functions already did...
20160223 23:30:02< celticminstrel> Okay.
20160223 23:30:43< mattsc> celticminstrel: sounds great
20160223 23:30:50< celticminstrel> (And more.)
20160223 23:31:07< irker437> wesnoth: Charles Dang wesnoth:master ed9ead2ac9e4 / src/ (6 files in 3 dirs): Add flag to quit_confirmation::quit() to only quit out of game https://github.com/wesnoth/wesnoth/commit/ed9ead2ac9e414efe52cab34f62fe2652cd88270
20160223 23:31:27< vultraz> celticminstrel, gfgtdf ^
20160223 23:32:48< vultraz> gfgtdf: I didn't test the change in the synced_choice_wait dialog since I'm not sure how to get to it
20160223 23:33:08-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.]
20160223 23:33:37< celticminstrel> That would require MP play.
20160223 23:33:43< celticminstrel> On an actual server.
20160223 23:33:53< celticminstrel> With a scenario that actually does a synched choice.
20160223 23:34:10< celticminstrel> (Well, LAN would work too, I guess.)
20160223 23:35:43< irker437> wesnoth: Charles Dang wesnoth:master 27d4091ec5bf / src/quit_confirmation.hpp: Reapplied a fix from f067332d0ff5 that got lost in ed9ead2ac9e4 https://github.com/wesnoth/wesnoth/commit/27d4091ec5bf510a9f3a1dd2a50d46ac229cfb8c
20160223 23:36:07< gfgtdf> vultraz: you need to play a mp game, use wesnoth.synchonize_choice() (of just any random using action liek creating units)in a prestart event and have a huge lag
20160223 23:38:38< celticminstrel> Logically, you'd think [leader_goal] would be a goal, but it is in fact an aspect.
20160223 23:38:40< vultraz> gfgtdf: hm ... bug: if you have a quit confirmation dialog open and then click the window X, you have to click it two times to actually make it do anything
20160223 23:38:55< vultraz> and if you click it once then you have to click on stuff in the confirmation dialog twice
20160223 23:39:09< celticminstrel> vultraz: Get rid of the foreach loop and use back() instead.
20160223 23:39:43< vultraz> blockers_.back().prompt_() ?
20160223 23:39:48< celticminstrel> Yes.
20160223 23:40:23< gfgtdf> no if you ahve 2 quit_confirnation object we shodul rather go and search for for the second one.
20160223 23:41:07< celticminstrel> gfgtdf: Explain?
20160223 23:41:18< vultraz> hm
20160223 23:41:25< vultraz> celticminstrel: that doesn't seem to resolve it
20160223 23:41:56< celticminstrel> Huh, okay, never mind then.
20160223 23:42:05< gfgtdf> celticminstrel: well you say its casues by teh foreach loop which somehow imples that you think that there is more than one entry in that loop
20160223 23:42:18< celticminstrel> Oh, wait, it's something to do with open_ maybe.
20160223 23:43:29< celticminstrel> ...so... what vultraz said is...
20160223 23:43:52< celticminstrel> "when quit confirmation is open, clicking the window's X twice force-quits the program"?
20160223 23:43:54< celticminstrel> and
20160223 23:44:27< celticminstrel> "if you click the window's X only once in that case, you get a spare quit confirmation dialog making it seem like you have to click twice to dismiss it"?
20160223 23:44:59< gfgtdf> vultraz: how did you open the first quit dialog ?
20160223 23:45:29< vultraz> hm
20160223 23:45:31< vultraz> wait
20160223 23:45:43< vultraz> the doubling up only happens in the eidtor
20160223 23:45:44< vultraz> editor
20160223 23:46:14< vultraz> in-game, doesn't matter if you click Quit to Desktop or Quit to Titlescreen, if you click the window X when the prompt is open the game will quit
20160223 23:46:16< vultraz> to desktop
20160223 23:46:27-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160223 23:46:34< celticminstrel> Does the editor controller inherit from quit_confirmation?
20160223 23:47:06< celticminstrel> Oh, right, the controller base class does, right?
20160223 23:47:11< celticminstrel> So that would be a yes.
20160223 23:47:19< vultraz> well
20160223 23:47:26< celticminstrel> Unless there's no such thing as an editor controller.
20160223 23:47:29< vultraz> gfgtdf moved it to play_controller
20160223 23:47:46< vultraz> (which reminds me, should move the header include too)
20160223 23:47:47-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 23:47:47< gfgtdf> vultraz: the solution is the teh eduitor shodul also use qc::quit.
20160223 23:47:51-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160223 23:47:58< gfgtdf> vultraz: no i also added one in editor controller
20160223 23:48:00-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 23:48:14< vultraz> no you didn't
20160223 23:48:20< gfgtdf> vultraz: yes i did
20160223 23:48:35< vultraz> oh
20160223 23:48:50< vultraz> I meant move the header include from controller_base to play_controller
20160223 23:48:53< vultraz> since it's still in the former
20160223 23:49:04< vultraz> and play_controller includes controller_base
20160223 23:49:16< gfgtdf> vultraz: i didnt move teh header include
20160223 23:49:26< vultraz> I'm syaing I should
20160223 23:49:27< vultraz> saying
20160223 23:50:24< gfgtdf> hmm strign comparision didnt work as expected
20160223 23:50:31< gfgtdf> maybe i shodul rebuild boost with icu support
20160223 23:52:27< vultraz> blagh
20160223 23:52:34< vultraz> the editor's main loop is weird :|
20160223 23:55:05-!- Greg-Boggs [~greg_bogg@173.240.241.83] has quit [Remote host closed the connection]
20160223 23:55:39< vultraz> i mean
20160223 23:55:40< vultraz> just
20160223 23:55:42< vultraz> why
20160223 23:55:42-!- Greg-Boggs [~greg_bogg@173.240.241.83] has joined #wesnoth-dev
20160223 23:55:55< vultraz> does it work on a loop dependent on a variable 
20160223 23:56:02< gfgtdf> vultraz: i think you just make qc::quit() a function that doesnt throw and return  boolean. Then you should add function qc::quit_title,qc::quit_desktop which are defines as << if(quit()) throw execption >> and make editor use qc::quit()
20160223 23:56:03< vultraz> fabi y u do dis
20160223 23:56:14< fabi> vultraz: hi
20160223 23:56:18-!- mjs-de [~mjs-de@f048098199.adsl.alicedsl.de] has quit [Remote host closed the connection]
20160223 23:57:03< fabi> wa do dis me?
20160223 23:57:58< vultraz> fabi: why did you design the editor to use a loop that needs to be exited by means of setting a variable 
20160223 23:58:31< fabi> hmmmm, sure I did this?
20160223 23:58:48< vultraz> well you pretty much wrote the editor in its current state
20160223 23:58:50< vultraz> didn't you
20160223 23:58:55< fabi> no
20160223 23:59:01< fabi> the framework is all ilor's
20160223 23:59:21< fabi> and ilor did not write it from scratch either
20160223 23:59:32< gfgtdf> boost.zip is twice as big as boost.7z did thex skip files in the 7z version ir it is just much better comptrressed?
--- Log closed Wed Feb 24 00:00:01 2016