--- Log opened Thu Apr 19 00:00:28 2018 20180419 00:03:32< celmin|sleep> @sinda — What main= key do you refer to? 20180419 00:04:40<+discordbot> main=$base {bridge_type}|$base{!}’s {bridge_type} 20180419 00:05:01< celmin|sleep> Soliton: Ah, I see, the village ones are missing the po-comment with link to the wiki page. 20180419 00:05:50< celmin|sleep> In that case, @sinda, I imagine you could somehow get gender agreement working in that mini-language, as it's basically a context-free grammar. 20180419 00:06:02< celmin|sleep> Not sure if it'll be easy though. 20180419 00:13:09-!- celmin|sleep is now known as celticminstrel 20180419 00:13:56< mattsc> @Vultraz @shadowm So, in my browser at least, typing ‘w’ opens the branch selection pull-down menu in the main page of the Wesnoth github website, and if you type in a name of a branch that doesn’t exist and then press return, it creates a branch of that name. 20180419 00:14:04< mattsc> While I have no memory of that happening, I must have copy/pasted the wmllint command into my brwoser rather than into a terminal yesterday. 20180419 00:14:18< mattsc> Kind of embarrassing, but at least it explains what happened. 20180419 00:14:42< irker110> wesnoth: Charles Dang wesnoth:master 42e69ee48390 / data/campaigns/Under_the_Burning_Suns/images/units/quenoth/flagbearer.png: UtBS: improved the Quenoth Flagbearer's flag https://github.com/wesnoth/wesnoth/commit/42e69ee48390fce41d22414bb468d192f2d36b3d 20180419 00:16:05<+discordbot> (my work ^ The updated flag, that is) 20180419 00:16:11< celticminstrel> :/ 20180419 00:16:48< irker110> wesnoth: Iris Morelle wesnoth:master fa1514215098 / changelog.md src/gui/dialogs/title_screen.cpp: gui2/title_screen: Make Version label formatted and translatable https://github.com/wesnoth/wesnoth/commit/fa15142150983162707c7455ae5cc5e917029e96 20180419 00:16:50< celticminstrel> Okay, uh. That commit adds an image only. 20180419 00:16:52< irker110> wesnoth: Iris Morelle wesnoth:1.14 91de2a4c9812 / changelog.md src/gui/dialogs/title_screen.cpp: gui2/title_screen: Make Version label formatted and translatable https://github.com/wesnoth/wesnoth/commit/91de2a4c9812d0cb0c963e85565e23778c071e32 20180419 00:17:00< celticminstrel> Is that what it's supposed to do? 20180419 00:20:39<+discordbot> celmin: modified image 20180419 00:21:00< celticminstrel> Right, if it's a modified image, why is it an add without a delete though? 20180419 00:22:28<+discordbot> it's not? 20180419 00:22:58 * celticminstrel looks again just to make sure I didn't get confused. 20180419 00:24:00< irker110> wesnoth: Charles Dang wesnoth:master 9148ce2b995b / data/campaigns/Under_the_Burning_Suns/images/units/quenoth/ (rider.png stalwart.png vanguard.png): UtBS: fixed some opaque shadows on rider units due to indexed pngs https://github.com/wesnoth/wesnoth/commit/9148ce2b995b05be2ee0606b63ffa26cb3da1ccb 20180419 00:24:16< celticminstrel> Hmm. I got confused by the fact that github wasn't showing me the before and after for some reason. 20180419 00:24:29<+discordbot> for some reason gh keeps showing the binary view 20180419 00:25:03< irker110> wesnoth: Charles Dang wesnoth:1.14 3cb1641d07ab / data/campaigns/Under_the_Burning_Suns/images/units/quenoth/flagbearer.png: UtBS: improved the Quenoth Flagbearer's flag https://github.com/wesnoth/wesnoth/commit/3cb1641d07ab3f744307316d9368aa9ce02f2c4f 20180419 00:25:06< irker110> wesnoth: Charles Dang wesnoth:1.14 12495f9070af / data/campaigns/Under_the_Burning_Suns/images/units/quenoth/ (rider.png stalwart.png vanguard.png): UtBS: fixed some opaque shadows on rider units due to indexed pngs https://github.com/wesnoth/wesnoth/commit/12495f9070afef4441fddaa1fd3341556a985756 20180419 00:25:40<+discordbot> I have n idea why 20180419 00:28:09<+discordbot> (also yes, I didn't update the Vanguard's flag yet) 20180419 00:28:20<+discordbot> or, no, rather 20180419 00:30:49<+discordbot> noticed yesterday the corrupted elf sprite is pretty bad... 20180419 00:31:00< celticminstrel> Both yes or no work equally well there. 20180419 00:31:07<+discordbot> it's based on an ancient zerovirus sprite afaik. 20180419 00:31:31<+discordbot> looks really weird 20180419 00:33:11<+discordbot> but's.... passable. for now. 20180419 00:33:18<+discordbot> but it's* 20180419 00:40:32-!- travis-ci [~travis-ci@ec2-54-198-17-152.compute-1.amazonaws.com] has joined #wesnoth-dev 20180419 00:40:33< travis-ci> wesnoth/wesnoth#17696 (master - fa15142 : Iris Morelle): The build is still failing. 20180419 00:40:33< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/368414159 20180419 00:40:33-!- travis-ci [~travis-ci@ec2-54-198-17-152.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180419 00:41:38<+discordbot> What is wrong with master's tests? 20180419 00:44:26<+discordbot> WARNING: Client prerequisites are not met. wesnoth cannot be built WARNING: Unit tests are disabled because their prerequisites are not me 20180419 00:44:48<+discordbot> loony probably broke them in https://github.com/wesnoth/wesnoth/commit/c51065e39c85bfd8e4705cefa381451385c29bf8 20180419 01:17:41<+discordbot> I hope that gets fixed tomorrow. 20180419 01:22:58-!- Ivanovic_ [~ivanovic@p4FC53504.dip0.t-ipconnect.de] has joined #wesnoth-dev 20180419 01:23:25-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Quit: Disconnecting from stoned server.] 20180419 01:24:17< mattsc> celticminstrel: I wrote you another forum PM 20180419 01:24:52-!- Ivanovic_ is now known as Ivanovic 20180419 01:28:49-!- timotei__ [~timotei@wesnoth/developer/timotei] has quit [Ping timeout: 260 seconds] 20180419 01:35:01<+discordbot> @Vultraz You're not going to consider #2912 a blocker I presume? 20180419 01:35:54<+discordbot> (Noting that it's easy to end up with a smaller window simply because you only have one screen and want to run the game windowed in order to be able to see something else behind or around it.) 20180419 01:51:09-!- timotei_ [~timotei@wesnoth/developer/timotei] has joined #wesnoth-dev 20180419 02:12:06< irker110> wesnoth/wesnoth:master Nils Kneuper ef7c0a7b64 updated Slovak translation AppVeyor: All builds passed 20180419 02:29:00< celticminstrel> IMO 2912 should be fixed for 1.14. 20180419 02:31:35<+discordbot> Same. 20180419 03:20:08<+discordbot> I shall ponder a fix 20180419 03:33:35< celticminstrel> So ponderous. 20180419 03:35:05<+discordbot> Indeed --- Log closed Thu Apr 19 03:46:36 2018 --- Log opened Thu Apr 19 03:46:39 2018 20180419 03:46:48-!- lobby [~wesnoth@wesnoth/bot/lobby] has joined #wesnoth-dev 20180419 03:46:48-!- Topic for #wesnoth-dev: 1.14.0 tag on April 25th (time TBD), announcement on May 1st (Steam launch 6 pm PST) | String and feature freeze on 1.14 branch | Wesnoth Developers Channel | >>> Want to help? Go here: https://r.wesnoth.org/t42911 (and thanks!) <<< | Discord Server: https://discord.gg/battleforwesnoth | Logs: http://irclogs.wesnoth.org | Bug tracker: https://bugs.wesnoth.org 20180419 03:46:48-!- Topic set by shadowm [~iris@wesnoth/developer/shadowm] [Tue Apr 17 05:44:42 2018] 20180419 03:46:48[Users #wesnoth-dev] 20180419 03:46:48[+discordbot ] [ ChipmunkV[m] ] [ Gambit ] [ Jetrel_bot ] [ minzbonbon] [ shadowm ] 20180419 03:46:48[ aeth ] [ commavir ] [ heirecka ] [ Kawa[m] ] [ nore ] [ Soliton ] 20180419 03:46:48[ AI0867 ] [ crimson_penguin] [ higgins ] [ lobby ] [ nurupo ] [ syrma[m] ] 20180419 03:46:48[ aidanhs1 ] [ DDR ] [ iceiceice] [ loonycyborg ] [ oldlaptop ] [ TC01 ] 20180419 03:46:48[ APic ] [ elias ] [ irker110 ] [ madmax28 ] [ Polsaker ] [ TheJJ ] 20180419 03:46:48[ Appleman1234 ] [ EliDupree ] [ Ivanovic ] [ matthiaskrgr] [ pydsigner ] [ timotei_ ] 20180419 03:46:48[ atarocch ] [ esr ] [ iwaim ] [ mattsc ] [ Ravana_ ] [ vihta ] 20180419 03:46:48[ celticminstrel] [ galegosimpatico] [ janebot ] [ midzer ] [ Rhonda ] [ vincent_c] 20180419 03:46:48-!- Irssi: #wesnoth-dev: Total of 48 nicks [0 ops, 0 halfops, 1 voices, 47 normal] 20180419 03:46:48-!- Home page for #wesnoth-dev: https://www.wesnoth.org 20180419 03:47:03-!- Channel #wesnoth-dev created Tue Jan 27 05:28:41 2009 20180419 03:47:57-!- Irssi: Join to #wesnoth-dev was synced in 77 secs 20180419 03:51:56< irker110> wesnoth: Iris Morelle valen:master 9ee6dacf5cd5 / bin/valen.pl: valen: Update wesnothd port assignments https://github.com/wesnoth/valen/commit/9ee6dacf5cd521d587d0397bd65cd8fba391389a 20180419 03:51:58< irker110> wesnoth: Iris Morelle valen:master 2e57519a834d / bin/valen.pl: valen: Stop monitoring gonzo https://github.com/wesnoth/valen/commit/2e57519a834d6ccc4a86aef24a7c84f04231d390 20180419 03:59:32-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.] 20180419 04:16:51-!- celticminstrel is now known as celmin|sleep 20180419 04:43:35-!- gallaecio [~quassel@188.79.96.255] has joined #wesnoth-dev 20180419 05:10:09<+discordbot> I want to learn theme WML and create some kind of mobile gui. Can you recommend me some resources where can I learn it? Also who is now WML Master? 20180419 05:14:12<+discordbot> You can look on the wiki, but it's not really worth it. It's an ugly mess, hard to work with, unintuitive, and doesn't handle edge cases such as excessively long labels (you can attest) very well. And it's being replaced in 1.15. You'll be better off waiting instead of trying to wrestle with themewml and coming up with a product that (through no fault of your own) is just sub-par. 20180419 05:15:32<+discordbot> @hrubymar10 https://wiki.wesnoth.org/ThemeWML . There's no "WML Master", that's never been a thing. 20180419 05:15:59<+discordbot> zookeeper's title on the forums is just cosmetic. There isn't a single person that knows everything. 20180419 05:16:37<+discordbot> Also what Vultraz said is true but do note he wants you to wait several years for a replacement. 20180419 05:39:35<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/436400446773657600/screenshot-2018-04-01-19-02-15.png 20180419 05:39:56-!- vn9711 [~vasya@94.158.103.15] has joined #wesnoth-dev 20180419 05:40:10<+discordbot> This is working version of Wesnoth on smallest supported iPhone (iPhone 5S with 4' inch display) 20180419 05:41:06<+discordbot> In my opinion it is fully usable but it is topic for long discuss... 20180419 05:41:19< vn9711> hmm, that was strange. I left wesnoth client open for a [long] while. Upon returning, I saw a pop-up "End of file", which can only be closed, and you are then returned to titlescreen. 20180419 05:41:46<+discordbot> In this scale, using this resolution, there are only few GUI bugs 20180419 05:42:01<+discordbot> vn9711: The server disconnected you because I intentionally killed the server. 20180419 05:42:16<+discordbot> about 'play screen', it looks like this: 20180419 05:42:16<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/436401120756236288/screenshot-2018-04-01-19-21-18.png 20180419 05:42:37<+discordbot> For some reason ( @loonycyborg ) the client code decides to report that as "EOF" instead of a more sensible messages that end users can understand. 20180419 05:42:44<+discordbot> *message 20180419 05:42:55<+discordbot> Yes, this one is little bit worse, so if I make ToD smaller, char picture smaller, so it should be good somehow 20180419 05:43:21< vn9711> shado wm okay. I guess it's a rare situation then. 20180419 05:43:38<+discordbot> Another screen, where should I make images smaller: 20180419 05:43:38<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/436401466261897227/screenshot-2018-03-31-21-37-53.png 20180419 05:43:56<+discordbot> I'd rather not comment on its frequency. The point is that it should display a message that people can actually understand instead of parroting API statuses. 20180419 05:44:03<+discordbot> For the latter there's stderr. 20180419 05:44:03<+discordbot> (and don't look at weird font, it is fixed already (see first screen) 20180419 05:45:14<+discordbot> Ok, so to summarise it: main problem on tiny screens are images, so I'll take default theme and make all images smaller, that's my plan 20180419 05:53:46-!- gallaecio [~quassel@188.79.96.255] has quit [Remote host closed the connection] 20180419 06:02:58<+discordbot> @hrubymar10 really, the images should fit... it's a high DPI problem >_> 20180419 06:05:58<+discordbot> yes, it is 😄 20180419 06:06:24<+discordbot> my iPhone 5S version using 1.5x workaround 20180419 06:06:57<+discordbot> but it doesn't matter if it is 2x... images doesn't scale 20180419 06:11:03-!- gallaecio [~quassel@119.red-83-34-169.dynamicip.rima-tde.net] has joined #wesnoth-dev 20180419 06:11:55-!- Ivanovic [~ivanovic@p4FC53504.dip0.t-ipconnect.de] has quit [Changing host] 20180419 06:11:55-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev 20180419 06:19:51< irker110> wesnoth: Nils Kneuper wesnoth:master 5bb1089abced / changelog.md players_changelog.md po/wesnoth/de.po: updated German translation https://github.com/wesnoth/wesnoth/commit/5bb1089abced303146885983a66263bd1b826f9f 20180419 06:19:55< irker110> wesnoth: Nils Kneuper wesnoth:1.14 57f773aa5a6b / changelog.md players_changelog.md po/wesnoth/de.po: updated German translation https://github.com/wesnoth/wesnoth/commit/57f773aa5a6be107df0bfcc08a8aea6c7cc47744 20180419 06:35:09-!- travis-ci [~travis-ci@ec2-54-198-17-152.compute-1.amazonaws.com] has joined #wesnoth-dev 20180419 06:35:10< travis-ci> wesnoth/wesnoth#17698 (master - 5bb1089 : Nils Kneuper): The build is still failing. 20180419 06:35:10< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/368492216 20180419 06:35:10-!- travis-ci [~travis-ci@ec2-54-198-17-152.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180419 06:38:00-!- grzywacz [~karol@89-70-226-147.dynamic.chello.pl] has joined #wesnoth-dev 20180419 06:50:58<+discordbot> Ideally, GUI WML could work with multiple units, like in CSS: pixels, display points, % of the enclosing area, inches, whatever. 20180419 06:52:27<+discordbot> Honestly, I’d think when it comes to Apple the DPI stuff should be a lot easier than windows 20180419 06:53:09<+discordbot> Is it not so that you get a canvas 2x the size of the normal window which is then rendered at usual size? 20180419 06:53:14<+discordbot> Or something like that? 20180419 06:53:57<+discordbot> Isn't that the exact opposite of what you want for DPI scaling? 20180419 06:54:27<+discordbot> Giving a larger canvas makes the program assume that there is a lot of space available, wasting it. 20180419 06:55:27<+discordbot> The program would just draw everything in the "normal size", and shrinking the canvas to the final size would make everything smaller. 20180419 06:56:06<+discordbot> Hm. Perhaps I misunderstand how they do it 20180419 06:56:34<+discordbot> As you’ve seen, I really have a hard time understanding the mechanisms of high DPI work 20180419 06:56:40<+discordbot> And regarding the layout system, CSS is a terrible example. Any web developer can tell you that CSS is a pain to work with. 20180419 06:56:55<+discordbot> Especially since, for example, on windows, pixels are the same as screen coordinates whereas on Mac they are not 20180419 06:57:01<+discordbot> The best layout system I've been able to work with has been Windows Presentation Foundation: https://docs.microsoft.com/en-us/dotnet/framework/wpf/controls/panels-overview 20180419 06:59:05<+discordbot> Really, for high-DPI scaling, there are no shortcuts. The only way to do it is for the GUI framework to make everything larger as a result of knowing that the DPI is higher. 20180419 07:02:14<+discordbot> @Vultraz It's even worse, as the display coordinates on Apple are in "points" which, multiplied by a non-integer factor, scale to a real pixels. 20180419 07:03:09<+discordbot> That's not necessarily a bad idea. It allows you to use the same coordinates, margins, etc. regardless of the DPI. 20180419 07:03:30<+discordbot> For example, WPF and CSS have the same thing (called "device-independent pixels"). 20180419 07:03:49<+discordbot> @jyrkive I mean, again, I’ve been able to consistently get stuff rendering at the right size in high dpi utilizing SDL_SetScaleFactor and a few number tweaks, but there’s always something that doesn’t work 20180419 07:04:32<+discordbot> Ie, text will be crisp but not as thick as it should be since the text dpi is still 72, or (always) mouse events don’t work properly 20180419 07:04:43<+discordbot> Buttons don’t stay hovered 20180419 07:05:02<+discordbot> And that’s not even considering the issue of image scaling or loading larger images 20180419 07:05:35<+discordbot> (The images get scaled, yes, but in cases like the logo we want to load the high dpi image...) 20180419 07:05:58-!- midzer [~quassel@p5B0515CA.dip0.t-ipconnect.de] has left #wesnoth-dev ["http://quassel-irc.org - Chat comfortably. Anywhere."] 20180419 07:06:37<+discordbot> My last attempt was trying to render on a 4 k canvas rendered at “scaled up” 96 dpi but was really 4K 20180419 07:06:46<+discordbot> SDL_RenderSetScale() isn't magic. Its usefulness is highly limited. 20180419 07:07:08<+discordbot> The only thing it does is altering the behavior of SDL's own rendering methods. 20180419 07:07:47<+discordbot> We're mostly blitting surfaces on screen, and thus SDL_RenderSetScale() shrinks them and corrects their positions. 20180419 07:08:00<+discordbot> But everything else needs to be done manually. 20180419 07:08:06<+discordbot> Increasing text size. 20180419 07:08:13<+discordbot> Loading high-DPI images. 20180419 07:08:22<+discordbot> Correcting coordinates in input events. 20180419 07:10:09-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev 20180419 07:12:00< irker110> wesnoth/wesnoth:master Charles Dang 9148ce2b99 UtBS: fixed some opaque shadows on rider AppVeyor: All builds passed 20180419 07:25:31<+discordbot> simple, this is not 20180419 07:25:46<+discordbot> times like this I wish we were using a third party UI API that did all this for us 20180419 07:25:54<+discordbot> else it's rather reinventing the wheel 20180419 07:26:39<+discordbot> This project has a history of reinventing wheels all over the place. In addition to our own game engine, we have a custom GUI framework and markup language. 20180419 07:26:55<+discordbot> I would like to move away from that going forward. 20180419 07:27:09<+discordbot> In general 20180419 07:27:45<+discordbot> Overreliance on custom systems means not only maintenance is harder, but that outside contributors have a much steeper learning curve when fixing bugs 20180419 07:27:48<+discordbot> Take yourself 20180419 07:28:47<+discordbot> If we had been using a proper, modern, accelerated rendering backend (OGL or even just SDL_Texture) you'd likely have had a lot easier time debugging and fixing a bunch of those bugs, such as the halo issues - that is, if they had even been present in the first place 20180419 07:32:04<+discordbot> Unfortunately, replacing custom code with third-party libraries would require pretty much a full rewrite of a large part of the game, if not everything. 20180419 07:32:16<+discordbot> I'm not signing up to help with such a rewrite. 20180419 07:32:31<+discordbot> Nor am I asking you to 20180419 07:33:27<+discordbot> That's too much to ask of anyone. 20180419 07:34:52<+discordbot> And it remains to be seen what exactly I'll do going forward. 20180419 07:35:38<+discordbot> You alone? 20180419 07:35:55<+discordbot> My advice would be to just pick reasonable goals and work towards them. 20180419 07:35:56<+discordbot> (To be fair, I will say working with Wesnoth's messy custom APIs has given me some insight on how this stuff works that using third party libraries likely wouldn't have) 20180419 07:36:05<+discordbot> (which is quite helpful) 20180419 07:38:47<+discordbot> @shadowm maybe. I might keep intermittently working on wesnoth as a hobby project/learning experience. Or I might be forced to cede leadership and move away from an active role. 20180419 07:39:10<+discordbot> Regardless of which one is the case it's not your decision alone. 20180419 07:39:10<+discordbot> And I can't really expect everyone to stick around forever. 20180419 07:39:58<+discordbot> I'm not planning to develop Wesnoth forever, either. 20180419 07:40:11<+discordbot> I'm not going to try to rope people into super long-term commitments like "rewrite the entire UI subsystem - again" or something of that scope. 20180419 07:41:24<+discordbot> It's important to ensure that the project can still be developed when we're gone. It involves, for example, avoiding code that's extremely difficult to understand (such as the template mess that was used to deliver GUI2 events), as well as not starting huge refactoring projects in master (what happens if the developer quits halfway through?) 20180419 07:42:57<+discordbot> I don't intend to quit before finishing this particular refactor. 20180419 07:43:21<+discordbot> That being said, if I were hit by a bus I'm sure someone with more experience could come in an finish it in like a month. 20180419 07:43:43<+discordbot> It's not that the problem I'm facing is inherently complicated. I just don't see the path to the solution. 20180419 07:49:40-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev 20180419 07:50:58<+discordbot> I will say the work that our various contributors has done over the past few years has been top-notch, though, despite the Brobdingnagian codebase. 👌 20180419 07:54:49<+discordbot> I give it a lot of flack, but it's certainly cleaner and more well-maintained now than before. 20180419 07:55:37< irker110> wesnoth: pentarctagon wesnoth:master 135921ad1033 / docker/ (Dockerfile-base-1604-master Dockerfile-base-1804-master): Re-add boost-thread. https://github.com/wesnoth/wesnoth/commit/135921ad10334ccf890dd10b7f086bc98597b9eb 20180419 07:56:55<+discordbot> @Pentarctagon can you re-add the entry in Install and remove the changelog entry? 20180419 07:57:29<+discordbot> sure 20180419 07:58:22<+discordbot> no need for that 20180419 07:59:01<+discordbot> confused 20180419 07:59:02<+discordbot> now boost.locale depending on boost.thread is implementation detail of boost since we don't depend on it directly 20180419 07:59:43<+discordbot> How does it matter? When SCons is used, the build fails if Boost.Thread headers and libraries aren't found. 20180419 07:59:53<+discordbot> Because of the check you restored. 20180419 08:00:20<+discordbot> no distribution has separate packages for boost headers 20180419 08:00:22< vn9711> A question about [multiplayer] tag. Does anybody know what the format for "force_modification" field is? https://wiki.wesnoth.org/ScenarioWML#The_.5Bmultiplayer.5D_tag 20180419 08:00:28<+discordbot> and look at this: https://packages.debian.org/stretch/libboost-locale1.62.0 20180419 08:00:36<+discordbot> boost.locale depends on boost.thread 20180419 08:00:51<+discordbot> thus it's handled automatically in case of dynamic linking 20180419 08:00:51< vn9711> I've tried to insert the following: 1. directory name 2. visual name 3. name from _server.pbl Neither works. 20180419 08:01:25<+discordbot> You should be looking at this: https://packages.debian.org/stretch/libboost-locale1.62-dev 20180419 08:01:55<+discordbot> vn9711: the value of [modification] id= 20180419 08:02:27<+discordbot> boost-locale-dev doesn't provide headers according to filelist 20180419 08:02:59<+discordbot> https://packages.debian.org/stretch/amd64/libboost-locale1.62-dev/filelist 20180419 08:03:11<+discordbot> scons fails if boost-thread isn't there, so it's a dependency of being able to use scons. 20180419 08:03:32<+discordbot> FOUR HUNDRED?! 20180419 08:03:34<+discordbot> there's no way it won't be there 20180419 08:03:43<+discordbot> Okay, so it only provides libboost_locale.a. I see. 20180419 08:03:48<+discordbot> 😱 20180419 08:04:03<+discordbot> But the check also verifies that libboost_thread.a is available. 20180419 08:04:26<+discordbot> no, it just checks if -lboost_tread works 20180419 08:04:46<+discordbot> And it doesn't work if you con't have libboost_thread.a. 20180419 08:04:48<+discordbot> without that dev package it can still link against .so directly 20180419 08:04:57<+discordbot> so it will still work 20180419 08:05:41<+discordbot> the thing is currently there is no good way to force scons to use static boost when both static and shared are available at same path 20180419 08:05:41<+discordbot> That's not happening according to Travis CI logs. 20180419 08:05:42<+discordbot> https://travis-ci.org/wesnoth/wesnoth/jobs/368492221 20180419 08:05:57<+discordbot> Checking for Boost thread library... (cached) no Checking for Boost locale library... (cached) yes Checking for Boost filesystem library... (cached) yes WARNING: Base prerequisites are not met 20180419 08:06:49<+discordbot> due to https://github.com/wesnoth/wesnoth/commit/52f8e8233c99c214d7409fd9f0b2cd4056590255, when it was said it wasn't needed anymore 20180419 08:07:36<+discordbot> @jyrkive re that bug above... my first instinct is that wesnoth just runs out of memory or something.. is that possible? 20180419 08:08:00<+discordbot> why does it say (cached)? 20180419 08:08:09<+discordbot> Not really, unless surfaces are stored in some special limited storage. 20180419 08:08:16<+discordbot> isn't the builddir recreated each time? 20180419 08:08:23<+discordbot> Yes, running out of memory is possible. A 32-bit program can only use up to 2 gigabytes of memory. 20180419 08:08:33<+discordbot> I can reproduce the crash, but only after image 600 or so 20180419 08:08:36<+discordbot> @loonycyborg travis has been caching builds for a while now 20180419 08:08:49<+discordbot> What's the memory usage with, say, 500 images? 20180419 08:09:00<+discordbot> it's why it can take less than half an hour to finish now 20180419 08:09:10<+discordbot> @jyrkive does readding boost_tread to apt line fix it? 20180419 08:09:15<+discordbot> I'd be concerned if somehow 400 animation frames was all it took for Wesnoth to reach the heap limit. 20180419 08:09:39<+discordbot> @loonycyborg We'll see. Pentarctagon already added it. 20180419 08:10:06<+discordbot> Hang on a moment. 20180419 08:10:09<+discordbot> These images are huge. 20180419 08:10:25<+discordbot> Each one is 800x600. 20180419 08:10:38<+discordbot> @jyrkive the game crashes at around 1.7 GB of memory usage 20180419 08:11:01<+discordbot> You are on 64-bit Windows though, right? 20180419 08:11:08<+discordbot> wesnoth is still built 32 bit 20180419 08:11:08< zookeeper> 400 800x600 frames? :D 20180419 08:11:23<+discordbot> Oh, right. 20180419 08:11:23<+discordbot> on windows 20180419 08:11:27<+discordbot> It's so much that it's likely indeed hitting the 2GB limit (not all of it can be used due to memory fragmentation). 20180419 08:11:28-!- travis-ci [~travis-ci@ec2-54-205-141-74.compute-1.amazonaws.com] has joined #wesnoth-dev 20180419 08:11:29< travis-ci> wesnoth/wesnoth#17700 (master - 135921a : pentarctagon): The build is still failing. 20180419 08:11:29< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/368520649 20180419 08:11:29-!- travis-ci [~travis-ci@ec2-54-205-141-74.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180419 08:11:43<+discordbot> to switch means rebuilding a bunch of hard-to-build-on-windows libs 20180419 08:11:47< vn9711> @Vultraz thanks! 20180419 08:12:20<+discordbot> Part of the 2 GB space is needed for mapping libraries as well, isn't it? 20180419 08:12:48<+discordbot> Yes, library code is in the same address space. 20180419 08:12:49<+discordbot> (is wesnoth 64 bit yet on linux? I know it is on macos) 20180419 08:12:59<+discordbot> Yes, it's 64-bit on 64-bit Linux. 20180419 08:13:04<+discordbot> I see... 20180419 08:13:05<+discordbot> Vultraz, Wesnoth has been 64-bit on Linux for over a decade. 20180419 08:13:13<+discordbot> Just checking, just checking... 20180419 08:13:28< Soliton> on windows it's 2GB userland and 2GB kernel. (on linux it's 3 and 1.) 20180419 08:13:28<+discordbot> it's 32-bit only on windows, IIRC 20180419 08:13:29<+discordbot> It has also been built on various platforms that are not x86-based. 20180419 08:14:29<+discordbot> We really need to get Windows builds up to 64 bit. 20180419 08:14:44<+discordbot> Do we know which libraries are the problem ones? 20180419 08:14:55<+discordbot> I disagree. In normal conditions we don't get anywhere near the memory usage limit. 20180419 08:15:07<+discordbot> Ie, I know boost is fairly easy to recompile, and SDL already provides 64 bit libs 20180419 08:15:33<+discordbot> Anyway, the most problematic libraries are definitely the GTK+ stack (Cairo, Pango, fontconfig, libpng, etc.) 20180419 08:15:39<+discordbot> oh I get it, -dev packages don't depend on it and runtime package provides only versioned .so 20180419 08:15:48<+discordbot> A 800x600 surface's pixmap (without any form of compression) would be around 469 KiB. 1000 frames of that is just around half a GB. 20180419 08:16:01<+discordbot> So something still seems off. 20180419 08:16:19<+discordbot> Wait hang on. 20180419 08:16:19<+discordbot> > (8006004*1000)/(2^20) ans = 1 831,054688 20180419 08:16:21<+discordbot> I did the maths wrong. 20180419 08:16:36<+discordbot> looks about right 20180419 08:16:36<+discordbot> Yeah I forgot each pixel takes up 4 bytes. 20180419 08:17:17<+discordbot> wishes the GTK people paid more mind to Windows. 20180419 08:17:37<+discordbot> libpng isn't part of the GTK+ stack? 20180419 08:17:37<+discordbot> It's not really caused by "not paying mind to Windows". 20180419 08:17:57<+discordbot> It's the reference implementation of the PNG library so it isn't tied to any single project. And in our case SDL_image is the primary user of it. 20180419 08:18:17<+discordbot> It's more that the GTK+ stack is very large, and it's not possible to use only one library (Pango in our case) without pulling in more libraries. 20180419 08:18:32<+discordbot> I mentioned libpng only because it's one of the dependencies. 20180419 08:18:32<+discordbot> We don't use libpng directly anymore 20180419 08:18:41<+discordbot> Do we? 20180419 08:18:52<+discordbot> No. SDL_image uses it I just said. 20180419 08:19:01<+discordbot> Right. 20180419 08:19:12<+discordbot> So we don't need to worry about that one. 20180419 08:19:27<+discordbot> It's the lightest library we use, really. 20180419 08:19:36<+discordbot> Its only dependencies are zlib and the C runtime. 20180419 08:19:43<+discordbot> hmm I guess I'm fine with readding thread to INSTALL then 20180419 08:19:54<+discordbot> We need to recompile it as part of the GTK+ stack anyway (although the script I used compiled everything automtically anyway). 20180419 08:20:13<+discordbot> Huh. 20180419 08:20:34<+discordbot> You built more than just pango/cairo? 20180419 08:20:50<+discordbot> Pango, Cairo, and all of their dependencies. 20180419 08:20:56<+discordbot> it needs buttload of gtk stuuf 20180419 08:21:04<+discordbot> gdk, pixman 20180419 08:21:05<+discordbot> etc 20180419 08:21:10<+discordbot> harfbuzz 20180419 08:21:15<+discordbot> If I replaced only Pango and Cairo and left other libraries to older versions, the game crashed on startup. 20180419 08:22:09<+discordbot> (Likewise, we'd have no hope of getting Wesnoth to run if we compiled 64-bit Pango and Cairo and tried to use it together with 32-bit builds of their dependencies.) 20180419 08:22:45<+discordbot> But... if you had all of this set up how come you didn't just build the 64 bit libs right then 🤔 20180419 08:23:09<+discordbot> Because it would have been additional work for little gain. 20180419 08:23:36< irker110> wesnoth: pentarctagon wesnoth:master f4db76fe0f80 / INSTALL.md changelog.md: Re-add boost-thread to INSTALL and changelog. https://github.com/wesnoth/wesnoth/commit/f4db76fe0f802af38bff36b88119109e7bc82382 20180419 08:24:02<+discordbot> But we still want to switch to 64 bit on Windows eventually. Do we not? 20180419 08:24:25<+discordbot> Sure, when someone is willing to spend the effort. 20180419 08:25:58<+discordbot> At least I have been focusing on much more important stuff (at the time, I was trying to get the credits screen not to be empty if the player had too many add-on campaigns installed). 20180419 08:26:27<+discordbot> ah, right, right 20180419 08:26:56<+discordbot> the end credits screen is actually one place where functionality REGRESSED with the transition from GUI1 to 2 >_> 20180419 08:27:14<+discordbot> #2915 highlights an issue with the image cache system in ??? (possibly 1.6) through 1.14. 20180419 08:28:32<+discordbot> IIRC it used to have an entry count limit or an LRU mechanism to drop entries (I don't remember which), I believe it's currently constrained only by the environment. 20180419 08:29:01<+discordbot> I wonder what issues we'll see with this situation and texture usage.. 20180419 08:29:23<+discordbot> Oh you want to know how OpenGL fares in memory shortage scenarios, cool. 20180419 08:29:31<+discordbot> Would it be worth introducing a mechanism to limit the size of the image cache in 1.14? 20180419 08:29:50<+discordbot> As opposed to, you know, just telling people to not make huge animations like inferno8 does. 20180419 08:29:58<+discordbot> Not in my opinion. We're going to hit the limit only in extreme conditions. 20180419 08:30:16<+discordbot> I kinda know what he's getting at — he's been using cinematic sequences in his campaign. 20180419 08:30:34<+discordbot> Wesnoth doesn't support embedding videos, so he's using the brute-force approach. 20180419 08:30:48<+discordbot> Embedded videos, you say... 20180419 08:30:58<+discordbot> 🤔 20180419 08:31:02<+discordbot> Incoming rant in 3... 2... 1... 20180419 08:31:18<+discordbot> I remember asking you about that years ago, funnily enough 20180419 08:31:36<+discordbot> He's not even the first to try, DDR was. 20180419 08:31:44<+discordbot> He just never did anything with it. 20180419 08:31:49<+discordbot> Could that even be implemented? 20180419 08:31:54<+discordbot> Why 20180419 08:31:58<+discordbot> I mean, seriously. 20180419 08:32:07<+discordbot> There are better uses for all of you people's time right now, such as fixing master. 20180419 08:32:31<+discordbot> Why do you assume when I mention an idea I mean drop everything and implement it IMMEDIATELY 20180419 08:32:39<+discordbot> I think if someone outside the development team wants to add video playback support they're more than welcome to do it, but if any of you do it I'll start to seriously worry about your priorities. 20180419 08:32:57<+discordbot> Vultraz, because your question only ever has one possible answer anywhere: "yes". 20180419 08:33:01<+discordbot> I'd say that at least it would be a better idea to explicitly support his original use case than try to deal with the fallout of the workaround. 20180419 08:33:02<+discordbot> But theory != practice. 20180419 08:33:21<+discordbot> It's a very absurd edge case. 20180419 08:33:33<+discordbot> My opinion is we should try to get 64 bit wesnoth working for 1.14.1. 20180419 08:33:35<+discordbot> Most content creators do not have the technical know-how to produce cinematics for their campaigns. 20180419 08:33:39<+discordbot> on windows 20180419 08:33:41<+discordbot> that is 20180419 08:34:28<+discordbot> inferno8's addon may be a corner case, but it's an extremely well-done and intricate corner case. 20180419 08:34:30<+discordbot> IMO, 64-bit builds don't have anywhere near large enough advantages to warrant doing it for 1.14.x. 20180419 08:34:32<+discordbot> we shouldn't dismiss it 20180419 08:35:00<+discordbot> I think it's worth 64 bit builds for that alone. 20180419 08:35:01<+discordbot> To clarify I was talking about video embedding support. IDC about IA32 vs. x86_64 builds. 20180419 08:35:34<+discordbot> If you think switching to x86_64 on Windows is viable without leaving significant parts of the userbase out in the cold, by all means go for it. 20180419 08:35:43<+discordbot> It's 2018 anyway. 20180419 08:36:04<+discordbot> However. 20180419 08:36:22<+discordbot> I'll note that raising the VA space limit to I don't remember how many gigabytes does not automatically give you more RAM. 20180419 08:36:23<+discordbot> I can just use 64-bit mingw instead of 32-bit mingw for windows releases 😛 20180419 08:36:25<+discordbot> If we want to allow such huge animations, the easiest way would be for the image cache manager to automatically remove old images when the cache grows too large. 20180419 08:36:34<+discordbot> And many people still run Wesnoth on 4 GB laptops. 20180419 08:37:08<+discordbot> @loonycyborg but don't you need to do the same rebuild of the libs? 20180419 08:37:09< grzywacz> 4GB is fairly typical unless you're a dev with macbook pro. 20180419 08:37:17<+discordbot> yes 20180419 08:37:18<+discordbot> Yes, he'll need to rebuild all libraries for that. 20180419 08:37:30<+discordbot> grzywacz: 4 GB in 2018? I'd say 8 GB is the typical minimum 20180419 08:37:31<+discordbot> All of them, no exceptions. 20180419 08:37:49<+discordbot> You clearly have never had to work under a budget. 20180419 08:37:53<+discordbot> I'm still not building my own gtk though 20180419 08:38:03< grzywacz> Vultraz: maybe if you're a gamer or buy new hardware. I know many people at 4GB (or below). 20180419 08:38:05<+discordbot> using their prebuilt sdk 20180419 08:38:12< aeth> 8 GB or even 16 GB is the typical minimum for computers. Desktop computers. 20180419 08:38:22<+discordbot> I have a 16 GB laptop. 20180419 08:38:24<+discordbot> My Linux laptop is still 4 GB because it's old and I can't afford replacing it. 20180419 08:38:26< aeth> But most computers are cheap laptops, probably 20180419 08:38:27<+discordbot> So maybe I'm biased 20180419 08:38:38<+discordbot> I can't afford buying any electronics atm. 20180419 08:38:58<+discordbot> I had to have 4Gb for long time because adding more was blocked by motherboard 20180419 08:39:02<+discordbot> It'd be hilariously sad if a free game of all things suddenly became unplayable for me. <_< 20180419 08:39:08< grzywacz> You're not biased, you're spoiled. :D :D :D 20180419 08:39:12<+discordbot> Without investing in new hardware with money I don't have. 20180419 08:39:14< aeth> My upgrade cycle is 5 years, and I like computers. 20180419 08:39:28< aeth> Assume 8 GB is the reasonable minimum now. What was it 5 years ago? 20180419 08:39:36< grzywacz> 4. 20180419 08:39:40<+discordbot> There's swap. As long as all the images aren't needed at the same time, using more than the physical RAM should be fine. 20180419 08:39:47<+discordbot> and I had swap thrashing happen a lot 20180419 08:39:47< grzywacz> 8GB is for gaming. 4GB is still plenty for web and facebook. 20180419 08:39:57<+discordbot> @loonycyborg I don't think it matters how you do it if the result is the Windows releases are 64 bit. Since we don't use VS for releases that can be handled later. 20180419 08:39:58< aeth> Not true. 20180419 08:40:11< aeth> The biggest use of RAM on my computer most days is my web browser. 20180419 08:40:19<+discordbot> yes 20180419 08:40:32< grzywacz> It is the biggest user, and 4GB is still sufficient unless you open 40 tabs for no good reason. :) 20180419 08:40:33<+discordbot> See, he confirmed what I said. 20180419 08:40:33<+discordbot> browser eats ram like a child candy 20180419 08:40:41<+discordbot> "game does not support videos (yet) so this is a workaround" 20180419 08:40:54< aeth> Just a web browser workload and you're pushing limits with 4 GB, especially if you're using Chrome on a heavyweight environment (as opposed to e.g. Firefox on a tiling window manager) 20180419 08:40:56<+discordbot> adds video playback to 1.16 mental TODO 20180419 08:41:00<+discordbot> ffs 20180419 08:41:05<+discordbot> lol 20180419 08:41:08<+discordbot> Vivaldi (the browser I use) is currently using 900MB. 20180419 08:41:12<+discordbot> Push it back into 1.18 instead. 20180419 08:41:13<+discordbot> Seven tabs. 20180419 08:41:14< grzywacz> Vultraz: embed webview and stream from youtube :-p 20180419 08:41:30<+discordbot> Steam hardware survey has 4 GBs at 3rd highest(~8%). 8 GBs is at ~42% and 12+ GBs is ~40%. 20180419 08:41:42<+discordbot> I think video playback wouldn't be such a big task. 20180419 08:41:50<+discordbot> Much easier than, say, DPI scaling. 20180419 08:42:04<+discordbot> @Pentarctagon I've never had a Steam hardware survey on Linux, for the record. 20180419 08:42:08< grzywacz> Pentarctagon: don't forget it's among gamers. My wife doesn't have steam on her low-end work laptop. :) 20180419 08:42:17<+discordbot> I don't understand their survey algorithm but it's bullcrap. 20180419 08:42:43< grzywacz> Don't they just get your hardware specs when you run steam? 20180419 08:42:43<+discordbot> Whereas on Windows I get them every once in a while. 20180419 08:42:50<+discordbot> @shadowm AFAIK, the hardware survey displays information for all users, regardless of whether they get the survey prompt or not. 20180419 08:43:06<+discordbot> (Except for the info that the survey asks, of course.) 20180419 08:43:20< aeth> grzywacz: It also might overcount weak hardware for gamers. I have Steam on my weak laptop and my strong desktop. I do maybe 95% of my gaming on my desktop, but there are some strategy games I play on my laptop from time to time. 20180419 08:43:35< grzywacz> Fair enough. 20180419 08:43:49< aeth> I do have Steam on my laptop, although I've only ever used it for CK2. You'd be surprised the games that don't play nice with Intel integrated. Simple, 2D puzzle games might not run. 20180419 08:43:54<+discordbot> Also, BTW, I saw just now that I indeed have the power to use @ everyone and @ here like you said yesterday (auto-complete suggests them). Apparently it changed at some point. 20180419 08:44:42< grzywacz> aeth: it's a sad state of things, actually. 20180419 08:45:20<+discordbot> grzywacz: spoiled...well, somewhat 😛 20180419 08:45:24< aeth> Well, KDE Sudoku works 20180419 08:45:33< grzywacz> facepalm.jpg 20180419 08:48:02<+discordbot> @loonycyborg could you add a cloud sync rule for the saves dir on linux? and likewise for @hrubymar10 on mac? 20180419 08:48:11<+discordbot> (on steam) 20180419 08:48:25<+discordbot> ok will look into it 20180419 08:48:31<+discordbot> you added it for windows? 20180419 08:48:33<+discordbot> and it works? 20180419 08:48:34<+discordbot> Yes. 20180419 08:48:43< grzywacz> btw. is there a pre-relase/beta/whatever steam version that can be tested at this point? 20180419 08:48:43<+discordbot> Hardcoded to the 1.14 userdata dir 20180419 08:48:50<+discordbot> ~/.local/share/wesnoth/1.14 should suffice on Linux (but it'll only be used by 1.14.0 and later). 20180419 08:48:58-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20180419 08:49:08<+discordbot> let me look at their environment prefixes.. 20180419 08:49:25<+discordbot> grzywacz: It's not public. Only accessible for few key developers (not even me). 20180419 08:49:39<+discordbot> @jyrkive I can add you 20180419 08:49:55<+discordbot> He's willing to invite anyone from the dev team who can help I'm pretty sure. 20180419 08:50:02< irker110> wesnoth/wesnoth:1.14 Charles Dang 12495f9070 UtBS: fixed some opaque shadows on rider AppVeyor: All builds passed 20180419 08:50:30<+discordbot> @shadowm do I want LinuxXdgDataHome or LinuxXdgConfigHome? 20180419 08:50:39<+discordbot> Former. 20180419 08:50:43< grzywacz> Well, I'd like to poke at it before it get released to wider public. Maybe I manage to break it in some creative way. 20180419 08:50:46<+discordbot> OK, I can DM my Steam details (dunno what info you need) when I get home (where I have Steam installed). 20180419 08:50:47<+discordbot> The latter would contain preferences. 20180419 08:51:01<+discordbot> He just needs your accountname and email address I believe. 20180419 08:51:15<+discordbot> Or possibly only the latter. 20180419 08:51:35<+discordbot> both 20180419 08:52:03<+discordbot> Note that the account name is not the one you can modify, it's the one you used during sign-up. 20180419 08:52:14<+discordbot> ok, sync path for Linux set to $XDG_DATA_HOME/wesnoth/1.14/saves 20180419 08:52:15<+discordbot> Well, lmk if I can be of any help with it. I have no problems sending these 2 pieces of information. 20180419 08:53:52<+discordbot> @Vultraz I don't have $XDG_DATA_HOME set in my env. Will it correctly default to $HOME/.local/share? 20180419 08:53:59<+discordbot> Yes. 20180419 08:54:05<+discordbot> ok. 20180419 08:54:08<+discordbot> https://standards.freedesktop.org/basedir-spec/basedir-spec-latest.html#variables 20180419 08:54:14<+discordbot> "$XDG_DATA_HOME defines the base directory relative to which user specific data files should be stored. If $XDG_DATA_HOME is either not set or empty, a default equal to $HOME/.local/share should be used." 20180419 08:54:27<+discordbot> Wesnoth complies with this requirement. 20180419 08:54:33<+discordbot> Ok. 20180419 08:55:21<+discordbot> The only weirdness wrt how Wesnoth handles XDG Base Directory is that it doesn't version the config file. 20180419 08:55:26<+discordbot> ok, so I just need @hrubymar10 to well me the path for macOS 20180419 08:55:43<+discordbot> It's always $XDG_CONFIG_HOME/wesnoth/preferences no matter what the game verison is. 20180419 08:56:10<+discordbot> (Window's is %USERPROFILE%/My Documents/My Games/Wesnoth 1.14/saves) 20180419 08:58:18<+discordbot> @hrubymar10 is ~/Library/Application Support/Wesnoth 1.14/saves correct? 20180419 09:00:37<+discordbot> nope, before we used Wesnoth_1.13 with underscore 20180419 09:00:48<+discordbot> So now it should be ~/Library/Application Support/Wesnoth_1.14/saves 20180419 09:01:23<+discordbot> ah, wait, the windows version has no space 20180419 09:01:23<+discordbot> ok 20180419 09:01:43<+discordbot> I'll create PR for it as I told you before. Do you talking about changing user's dir version, right? 20180419 09:02:04<+discordbot> you can commit the userdata change now 20180419 09:02:10<+discordbot> kk 20180419 09:02:24<+discordbot> inb4 RC 4. 20180419 09:03:04<+discordbot> now, hrub suggested I use TBfW for the app title instead of just BfW.... ponders.... 20180419 09:03:42<+discordbot> Official name of this project should be The Battle for Wesnoth IMHO 20180419 09:03:49<+discordbot> "On April 19th, the WesCherrybomb vulnerability was found allowing a malicious party to inject code into kernel mode by exploiting an unprecedented flaw in the game's image cache mechanism." 20180419 09:03:57<+discordbot> It should be changed on steam at least 20180419 09:04:04<+discordbot> "The development team released versions 1.12.7 and 1.13.15 addressing this issue 24 hours later." 20180419 09:04:21<+discordbot> (I couldn't think of a better marketing name for it, sorry.) 20180419 09:04:49<+discordbot> The Battle for Ring 0? 😛 20180419 09:04:56<+discordbot> o 20180419 09:04:57< grzywacz> ugh 20180419 09:05:05<+discordbot> is that a Halo reference 20180419 09:05:14< grzywacz> That's ® by Tolkien. 20180419 09:05:15<+discordbot> No. 20180419 09:05:27<+discordbot> How about "The One Ring"? 20180419 09:05:33<+discordbot> Ah 20180419 09:08:21<+discordbot> so, who prefers TBfW for the app name on steam? 20180419 09:08:31<+discordbot> Ehhhhhhhhhhhhhhhhhhhhhhh 20180419 09:08:42<+discordbot> I don't like the acronym that results from that. 20180419 09:09:15< irker110> wesnoth: Martin Hrubý (hrubymar10) wesnoth:1.14 ee500f57ece6 / players_changelog.md projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Update macOS user dir path https://github.com/wesnoth/wesnoth/commit/ee500f57ece60900060c606679a140a76b622944 20180419 09:09:18<+discordbot> I do 😄 20180419 09:09:28<+discordbot> can't be worse than THoT 😬 20180419 09:10:02<+discordbot> brb dinnnnnah 20180419 09:11:02< zookeeper> BfW looks nicer than TBfW, for sure. 20180419 09:12:16<+discordbot> We can always use BfW acronym in chats, ... but in official documents we should use full name 20180419 09:12:38<+discordbot> it's like Windows 10 vs Win10 20180419 09:15:18-!- travis-ci [~travis-ci@ec2-54-205-141-74.compute-1.amazonaws.com] has joined #wesnoth-dev 20180419 09:15:19< travis-ci> wesnoth/wesnoth#17700 (master - 135921a : pentarctagon): The build was fixed. 20180419 09:15:19< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/368520649 20180419 09:15:19-!- travis-ci [~travis-ci@ec2-54-205-141-74.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180419 09:23:16<+discordbot> @Vultraz @hrubymar10 Uh, I'm apparently unable to quickly fix the IME input in text_box. With IME, all of: cursor movenemt (cursor gets into the middle of utf8 code points), selection (same problem), Return handling in chat window, IME itself - are disasters. Whole IME input in SDL is a disaster, tbh. I signed up to examine a crash stack and fix what's UB or what, but see a need to rewrite half of text_box with consideration of 20180419 09:23:16<+discordbot> multibyte strings, and maybe a piece of SDL2 too. Still working on it, no idea if I can finish it by May. 20180419 09:23:30<+discordbot> seems 1.14 branch still keeps saves in /home/USER/.local/share/wesnoth/1.13 20180419 09:23:40<+discordbot> so steamcloud doesn't find anything yet 20180419 09:23:55<+discordbot> because it's set to 1.14 20180419 09:24:12<+discordbot> when is it planned to bump it to 1.14? 20180419 09:24:24<+discordbot> It will become 1.14 when the version number is 1.14.x. 20180419 09:24:42<+discordbot> I'm not entirely sure why @hrubymar10 is doing it by hand. 20180419 09:26:21<+discordbot> It looks like for some mysterious reason there isn't a macOS code path that will generate the correct suffix automatically? 20180419 09:26:57<+discordbot> filesystem::set_user_data_dir() uses filesystem::get_version_path_suffix() on X11 and Windows. 20180419 09:26:59<+discordbot> Yes, it is compiled inside boost at compile time 20180419 09:27:04<+discordbot> Boost? 20180419 09:27:05<+discordbot> Not runtime 20180419 09:27:09<+discordbot> What do you mean Boost 20180419 09:27:16<+discordbot> boost_filesystem 20180419 09:27:31<+discordbot> I think you mean filesystem_boost.cpp. 20180419 09:28:17<+discordbot> Which is not Boost. It's the FS API implementation which is just called that because it's the version that uses Boost.Filesystem (historical artifact from 1.12.x, which has a native alternative that doesn¿t use Boost.Filesystem). 20180419 09:28:56<+discordbot> For 1.15.x this really ought to be done dynamically instead. 20180419 09:29:03<+discordbot> You are right, I was wrong 🤔 20180419 09:29:30<+discordbot> Yes, it should be done at runtime 20180419 09:29:36<+discordbot> (Unless there's a really good reason for hard-coding the path on macOS and forcing the packager of the day to do extra work.) 20180419 09:30:49<+discordbot> I don’t know. I don’t have access to source right now. I can look at it later 20180419 10:45:28<+discordbot> @Vultraz Would it be possible to get command history in the lobby? :/ 20180419 10:45:46<+discordbot> Uhhhh 20180419 10:45:55<+discordbot> Or selection. 20180419 10:45:56<+discordbot> As in, chat command history? 20180419 10:46:10<+discordbot> I just typed a long ass command and now I need to retype it by hand because I can't repeat it using history. 20180419 10:47:21<+discordbot> I disabled history for the Chatbox since I thought it weird that PMs were being saved verbatim in your preferences 20180419 10:49:32<+discordbot> Hmmm. 20180419 10:49:59<+discordbot> You could disable it for whisper windiows as well as any windows where the line starts with /whisper surely? 20180419 10:52:57<+discordbot> Yes, but the API would need to be expanded 20180419 11:12:36<+discordbot> what are advantages of OpenMP in wesnoth? 20180419 11:12:52<+discordbot> None that have been benchmarked in a decade or so. 20180419 11:13:07<+discordbot> Speed, as far as I know. OpenMP parallelizes some operations. 20180419 11:13:38<+discordbot> But indeed, the improvement hasn't been benchmarked. We should assume the improvement to be negligible unless someone shows otherwise. 20180419 11:14:24<+discordbot> That's actually something I've been curious about - why were actual separate threads not used? 20180419 11:14:39<+discordbot> I am trying wesnoth on macOS with OpenMP enabled now. I know about crash risks. 20180419 11:14:56<+discordbot> So far I don't see much difference in speed tbh 20180419 11:15:08<+discordbot> OpenMP is easier to some extent. You only need to annotate a loop to be run in parallel. 20180419 11:15:26<+discordbot> Believe it or not, most devs until you took over were wary of using threading outside network code. 20180419 11:15:32<+discordbot> (Although that ease is also a disadvantage. It's very easy to shoot yourself in the foot this way.) 20180419 11:15:51<+discordbot> @hrubymar10 those crashes haven't been confirmed in years 20180419 11:15:56<+discordbot> Maybe my CPU is too fast? on older for example Core2Duo it will make more sense? 20180419 11:16:04<+discordbot> https://github.com/wesnoth/wesnoth/issues/1260 20180419 11:16:09<+discordbot> the discussion is from 2011 20180419 11:16:43<+discordbot> Most likely explanation is that the code OpenMP speeds up isn't the bottleneck to begin with. 20180419 11:18:11<+discordbot> just checked... the OMP code is commented out in display.cpp 20180419 11:18:24<+discordbot> In master? 20180419 11:18:43<+discordbot> cpp // openMP can't iterate over std::size_t const int omp_iterations = open_mp_list.size(); // #pragma omp parallel for shared(open_mp_list) // This loop must not be parallelized. Refresh is not thread-safe; for one, unit filters are not thread safe. // This is because, adding new "scoped" wml variables is not thread safe. Lua itself is not thread safe. // When this loop was parallelized, assertion 20180419 11:18:43<+discordbot> failures were reported in windows openmp builds. for(int i = 0; i < omp_iterations; i++) { open_mp_list[i]->anim_comp().refresh(); } 20180419 11:18:48<+discordbot> pretty sure it's like that in 1.14 20180419 11:19:09<+discordbot> that puts almost all the remaining usecases in image.cpp and the image cache 20180419 11:19:15<+discordbot> surface cache* 20180419 11:19:41<+discordbot> Yes, that loop is also serial in 1.14. 20180419 11:20:12<+discordbot> and no idea what this case is in units/frame.cpp 20180419 11:20:15<+discordbot> cpp #ifdef _OPENMP #pragma omp critical(frame_surface) // with the way surfaces work it's hard to lock the refcount within sdl_utils #endif //_OPENMP 20180419 11:20:25<+discordbot> there's no frame_surface anymore 20180419 11:20:46<+discordbot> and why is it trying to parallelize a surface? 20180419 11:20:48<+discordbot> Yeah, I would not be surprised if OMP suffered some coderot. 20180419 11:21:08<+discordbot> After boucman stopped being involved I don't think anyone did anything with it ever again. 20180419 11:21:55<+discordbot> Wonder if we should just remove it 20180419 11:22:11<+discordbot> I actually suggested that for 1.12 and then again in 1.13.x. 20180419 11:22:41<+discordbot> Rewriting restart_process(), which is used only in OpenMP-enabled builds, is probably the last OpenMP-related thing anyone did. (January 2016.) 20180419 11:22:41<+discordbot> I remember suggesting it at some point and someone telling me it was stupid to just rip it out without knowing it improved anything 20180419 11:23:28<+discordbot> There wasn't a frame_surface when that code was introduced either btw. 20180419 11:24:11<+discordbot> https://github.com/wesnoth/wesnoth/commit/6edcc85bc3ee282b4cf05bfc2008e91e4d5d17fe 20180419 11:24:44<+discordbot> In commits like that, it would be nice if the developer had performance measurements in the commit message... 20180419 11:24:48<+discordbot> if no one has objections I can remove the OMP code from master. 20180419 11:25:23<+discordbot> Only if you actually remove the CMake and SCons code for it atomically or do it in a branch where the relevant people can help you with that. 20180419 11:25:37<+discordbot> Because, honestly. 20180419 11:25:50<+discordbot> The libpng options lingered around for way longer than they should've. 20180419 11:25:52<+discordbot> Or how about removing the CMake/SCons stuff first? 20180419 11:26:23<+discordbot> Or how about removing it all at once 😛 20180419 11:27:04<+discordbot> That's why I said "atomically". 20180419 11:27:06<+discordbot> I found the email but I don't think jyrkive will be impressed 20180419 11:27:24<+discordbot> (Just call me Jyrki.) 20180419 11:29:02<+discordbot> boucman's OP: https://pastebin.com/5tv47G8G 20180419 11:29:37<+discordbot> "A non-scientific measure showed a reduction of the average AI turn on a AI vs AI game of 8% on my 4core HT computer. So this is not neglectable." 20180419 11:31:15<+discordbot> It's a whole thread but I don't feel like pasting the entire thing right now. 20180419 11:31:19<+discordbot> Eight percent? Seriously? I have achieved improvements of couple dozen percent with simple optimizations. 20180419 11:31:54<+discordbot> bye bye omp 20180419 11:32:30<+discordbot> I found a message from myself complaining about Wesnoth using 100% CPU as the result of OMP's default of busy-waiting on a second thread. 20180419 11:33:38<+discordbot> And that's why Wesnoth restarts itself on startup if OpenMP is enabled. 20180419 11:34:07<+discordbot> And why I ended up rewriting that restart code (since originally it didn't work if Wesnoth's path contains spaces). 20180419 11:43:23<+discordbot> Is font scaling supposed to be available in 1.13.14? 20180419 11:43:44<+discordbot> I can't find it for some reason. I'm pretty sure it was there in 1.13.12 20180419 11:44:14<+discordbot> It was disabled AFAIK 20180419 11:44:16<+discordbot> It was removed in 1.13.14. See the changelog. 20180419 11:45:11<+discordbot> Ah, first entry, I see. Thank you. 20180419 11:59:30<+discordbot> what should be the green number ...: 20180419 11:59:31<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/436496057132646420/unknown.png 20180419 11:59:33<+discordbot> ? 20180419 12:00:25< Soliton> HP? or the stuff besides the movement points? 20180419 12:00:58<+discordbot> I mean green 6... 20180419 12:01:03< Soliton> visibility changes are displayed somewhere there. not sure it's exactly that line. 20180419 12:01:19<+discordbot> it should be defense? 20180419 12:01:36< irker110> wesnoth: Martin Hrubý (hrubymar10) wesnoth:1.14 4ba1811f03bf / data/themes/default.cfg: Fix MP label for #2855 https://github.com/wesnoth/wesnoth/commit/4ba1811f03bfd6b9706a281e05c316be86cbe97c 20180419 12:07:06< irker110> wesnoth: Charles Dang wesnoth:master 3792612fb759 / / (13 files in 4 dirs): Removed OpenMP-related code https://github.com/wesnoth/wesnoth/commit/3792612fb759942e6eb5de51364c2cca4ed16b40 20180419 12:07:09<+discordbot> yay, wesnoth now uses the console from which it's spawned again \ o / 20180419 12:14:52<+discordbot> @loonycyborg should I removed the unused ping timeout pref? 20180419 12:17:40< Soliton> in what way is it unused? 20180419 12:18:15<+discordbot> it doesn't do anything 20180419 12:18:28<+discordbot> I broke defense label... I am working on it RN but I have problem 20180419 12:18:28<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/436500830204461056/unknown.png 20180419 12:18:55<+discordbot> just shift it a few pixels to the left. 20180419 12:19:02<+discordbot> I need XP to be rect="=,+14,+63,+14" instead of rect="=,+14,+60,+14" 20180419 12:19:22<+discordbot> but this brokes defense label 🤔 20180419 12:21:56<+discordbot> Now you see why themewml is horrible 20180419 12:24:15<+discordbot> 😄 20180419 12:31:09< irker110> wesnoth/wesnoth:master Nils Kneuper 5bb1089abc updated German translation AppVeyor: All builds passed 20180419 12:37:46<+discordbot> Do I have to restart game because of every WML theme change? 20180419 12:38:55<+discordbot> Also, isn't that some kind of other bug? 20180419 12:38:55<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/436505973671067670/unknown.png 20180419 12:39:04<+discordbot> there should be red 20% 20180419 12:56:23<+discordbot> no for the themewml you can just press f5 at the titlescreen 20180419 13:23:37<+discordbot> @Vultraz if it doesn't do anything then sure 20180419 13:24:19< irker110> wesnoth: Martin Hrubý (hrubymar10) wesnoth:1.14 47bd5ade3726 / data/themes/default.cfg: Fix for defense label which I broke in 4ba1811f03bfd6b9706a281e05c316be86cbe97c https://github.com/wesnoth/wesnoth/commit/47bd5ade3726675c0d6e4e68af82004a01ddb63d 20180419 13:24:30<+discordbot> Fixed 20180419 13:25:11<+discordbot> about openmp, maybe ping boucman, he's the one who made that code iirc 20180419 13:25:11<+discordbot> Should I cherrypick it to master? 20180419 13:25:37<+discordbot> Don't. We don't have the ThemeWML sidebar in master. 20180419 13:25:45<+discordbot> kk 20180419 13:25:56< boucman> loonycyborg: that was a long time ago, and a workaround to squeeze some performance without adding proper multithreading 20180419 13:26:05< boucman> if it's more trouble than use, remove it... 20180419 13:31:50<+discordbot> Just did so 20180419 14:02:32< irker110> wesnoth/wesnoth:1.14 Nils Kneuper 57f773aa5a updated German translation AppVeyor: All builds passed 20180419 14:04:50-!- gallaecio [~quassel@119.red-83-34-169.dynamicip.rima-tde.net] has quit [Ping timeout: 256 seconds] 20180419 14:05:13-!- gallaecio [~quassel@38.red-83-46-126.dynamicip.rima-tde.net] has joined #wesnoth-dev 20180419 14:14:57-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180419 14:15:03-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180419 14:48:35-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180419 14:48:42-!- janebot_ [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180419 14:48:42-!- janebot_ is now known as janebot 20180419 15:07:08-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180419 15:07:15-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180419 15:12:16-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180419 15:12:22-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180419 15:53:32-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180419 15:53:37-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180419 15:58:20< irker110> wesnoth/wesnoth:master pentarctagon f4db76fe0f Re-add boost-thread to INSTALL and chang AppVeyor: All builds passed 20180419 16:13:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection] 20180419 16:37:42-!- Appleman1234 [~quassel@124x38x163x22.ap124.ftth.ucom.ne.jp] has quit [Read error: Connection reset by peer] 20180419 16:41:19-!- Oebele [~quassel@143.177.58.202] has joined #wesnoth-dev 20180419 17:04:11-!- vn9711 [~vasya@94.158.103.15] has quit [Quit: Leaving.] 20180419 17:04:13-!- vn971 [~vasya@94.158.103.15] has joined #wesnoth-dev 20180419 17:37:49-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20180419 17:44:17<+discordbot> @jyrkive ? Do you have time to try reproduce this on windows? The Battle for Wesnoth version 1.13.14 Running on Apple macOS 10.13.5 Game paths ========== Data dir: /Users/USER/Do_not_backup/wesnoth_1.13.14/Wesnoth.app/Contents/Resources User config dir: /Users/USER/Library/Application Support/Wesnoth_1.14 User data dir: /Users/USER/Library/Application Support/Wesnoth_1.14 Saves dir: 20180419 17:44:17<+discordbot> /Users/USER/Library/Application Support/Wesnoth_1.14/saves Add-ons dir: /Users/USER/Library/Application Support/Wesnoth_1.14/data/add-ons Cache dir: /Users/USER/Library/Application Support/Wesnoth_1.14/cache Libraries ========= Boost: 1.66 OpenSSL/libcrypto: 1.1.0h (runtime 1.1.0h) Cairo: 1.14.12 (runtime 1.14.12) Pango: 1.40.1 (runtime 1.40.1) SDL: 2.0.8 (runtime 2.0.8) 20180419 17:44:18<+discordbot> SDL_image: 2.0.3 (runtime 2.0.3) SDL_mixer: 2.0.2 (runtime 2.0.2) SDL_ttf: 2.0.14 (runtime 2.0.14) Features ======== Experimental OpenMP support: no JPG screenshots: yes Lua console completion: yes Legacy bidirectional rendering: no Cocoa notifications back end: yes 20180419 17:45:24<+discordbot> No, I don't. 20180419 17:45:44<+discordbot> kk, no problem 20180419 17:45:54<+discordbot> And please don't ping me unless the issue has something to do with me in particular (for example, if it seems to be caused by a commit of mine). 20180419 17:46:33<+discordbot> Ok, sorry. I just wanted to ask 20180419 17:51:47-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180419 17:51:54-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180419 17:57:47< Ravana_> after canceling game, I see different old game in gamelist for a moment https://www.sendspace.com/file/wn29vs Log has from before 20180419 20:43:54 warning engine: User mobile has unknown game_id: 21, but nothing new during testing 20180419 18:05:05<+discordbot> Ravana_ user mobile is iOS version which is not created from tag 20180419 18:05:36<+discordbot> It wasn't me, I just telling you why there was unknown game_id 20180419 18:08:28< Ravana_> wouldn't it have +dev and use trunk server then? 20180419 18:10:36-!- Oebele [~quassel@143.177.58.202] has quit [Remote host closed the connection] 20180419 18:17:52<+discordbot> nope 20180419 18:18:25<+discordbot> For some reason sinda doesn't include these +dev tags to iOS version 20180419 18:18:57-!- gallaecio [~quassel@38.red-83-46-126.dynamicip.rima-tde.net] has quit [Remote host closed the connection] 20180419 18:39:02-!- gallaecio [~quassel@188.79.96.255] has joined #wesnoth-dev 20180419 19:01:04<+discordbot> He shouldn't. There's not development versions. 20180419 19:01:08<+discordbot> *They're not 20180419 19:01:29<+discordbot> They're not unreleased versions I mean, argh. 20180419 19:10:15-!- irker110 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout] 20180419 19:41:41-!- irker198 [~irker@uruz.ai0867.net] has joined #wesnoth-dev 20180419 19:41:42< irker198> wesnoth: Nils Kneuper wesnoth:1.14 9a59c9627be0 / / (7 files in 6 dirs): updated Scottish Gaelic translation https://github.com/wesnoth/wesnoth/commit/9a59c9627be0aa7d84dc8098d650dba0c9fa08ae 20180419 19:41:42< irker198> wesnoth: Nils Kneuper wesnoth:master 767c9aa10178 / po/ (5 files in 5 dirs): updated Scottish Gaelic translation https://github.com/wesnoth/wesnoth/commit/767c9aa10178e4a6b31dbe750f5b7ff3fe2e6118 20180419 19:52:02-!- higgins [~higgins@68.ip-149-56-14.net] has quit [Quit: Leaving] 20180419 19:55:26-!- higgins [~higgins@68.ip-149-56-14.net] has joined #wesnoth-dev 20180419 20:06:08<+discordbot> @hrubymar10 What was that? Does iOS version make trouble on MP server? 20180419 20:08:21<+discordbot> yes, I also saw user mobile on 1.13.14 MP server. There was some updates to server before 'desktop' 1.13.14 AFAIK 20180419 20:09:19<+discordbot> And this user mobile used old version of wesnoth which caused some warnings in Ravana's log 20180419 20:11:08<+discordbot> You should better keep iOS version synced with desktop version. I already told you that it would be good, if you will be able to update wesnoth iOS to latest tag (1.13.14). Not to latest commit possible, because it will cause errors if you won't use +dev suffix in version 20180419 20:11:15<+discordbot> @sinda ^^ 20180419 20:16:29-!- Appleman1234 [~quassel@124x38x163x22.ap124.ftth.ucom.ne.jp] has joined #wesnoth-dev 20180419 20:34:57-!- gallaecio [~quassel@188.79.96.255] has quit [Ping timeout: 240 seconds] 20180419 20:35:01< irker198> wesnoth/wesnoth:master Charles Dang 3792612fb7 Removed OpenMP-related code AppVeyor: All builds passed 20180419 20:35:41-!- gallaecio [~quassel@188.79.96.255] has joined #wesnoth-dev 20180419 21:17:18< grzywacz> \o/ 20180419 21:17:42<+discordbot> hm? 20180419 21:52:09<+discordbot> https://github.com/wesnoth/wesnoth/blob/1.14/src/units/attack_type.cpp#L429 movement should probaby be not in the part here, commit c8e63168 20180419 21:52:28-!- Skyfaller [~skyfaller@wikipedia/Skyfaller] has joined #wesnoth-dev 20180419 22:10:29-!- gallaecio [~quassel@188.79.96.255] has quit [Ping timeout: 265 seconds] 20180419 22:32:51< irker198> wesnoth/wesnoth:1.14 Martin Hrubý (hrubymar10) 47bd5ade37 Fix for defense label which I broke in 4 AppVeyor: All builds passed 20180419 23:01:47<+discordbot> hm 20180419 23:01:49<+discordbot> debatable 20180419 23:06:32<+discordbot> at least in the other cases only the variable is in the span tag 20180419 23:06:56<+discordbot> Oh, I see 20180419 23:07:00<+discordbot> ok 20180419 23:07:02<+discordbot> then yes 20180419 23:07:17<+discordbot> Please commit the change to master. 20180419 23:07:41<+discordbot> that said, the block above has no span tags ... 20180419 23:07:43<+discordbot> will do 20180419 23:13:02<+discordbot> can't backport it to 1.14 just yet since that would be breaking the string freeze... though something like that could probably be po-fixed? 20180419 23:14:06<+discordbot> @sinda ok, so, you cannot fix the IME issue? 20180419 23:20:43<+discordbot> No vultraz it can't be pofixed. 20180419 23:20:55<+discordbot> Both the msgid and msgstr need to be changed. 20180419 23:21:07<+discordbot> 'k 20180419 23:24:18<+discordbot> @Vultraz I got some result last late, late night 😄 It works more or less, but needs some more debugging. 20180419 23:24:21< irker198> wesnoth: Charles Dang wesnoth:master f2a870b97ca9 / src/ (game_display.cpp game_display.hpp): Game Display: removed unused variable https://github.com/wesnoth/wesnoth/commit/f2a870b97ca9dcda1bd0d4bbe07b2b9379aa0a53 20180419 23:24:42<+discordbot> @sinda without refactoring text_box? 20180419 23:33:55< irker198> wesnoth: Severin Glöckner wesnoth:master beaddeee64ff / data/campaigns/tutorial/scenarios/01_Tutorial_part_1.cfg: Revert "tutorial: add translation note" https://github.com/wesnoth/wesnoth/commit/beaddeee64ff6db897aa00e47b83d50ef8f398b7 20180419 23:33:57< irker198> wesnoth: Severin Glöckner wesnoth:master f0f664b636ef / src/units/attack_type.cpp: Fix misplaced tag https://github.com/wesnoth/wesnoth/commit/f0f664b636efdaf788b77f15726dbc329b4d8413 20180419 23:40:01<+discordbot> @Vultraz Yes, I think. The code looks like it used to work, but was broken over time. Looks like it's possible to recover it. 20180419 23:40:36<+discordbot> 👍 20180419 23:40:54<+discordbot> Would be nice to get @RatArmy 's review on it. 20180419 23:47:11-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection] 20180419 23:47:28-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [] 20180419 23:49:13< irker198> wesnoth: Charles Dang wesnoth:master e583c470be59 / src/ (7 files in 3 dirs): Display: cleaned up overlay map getter interface https://github.com/wesnoth/wesnoth/commit/e583c470be597b486e447d0a0abfea820b1afc43 --- Log closed Fri Apr 20 00:00:22 2018