--- Log opened Sun Aug 12 00:00:59 2018 20180812 00:03:52-!- louis94 [~~louis94@109.49-136-217.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 256 seconds] 20180812 00:36:47< celmin> There's a slight confusion in the wiki about terrain globbing. 20180812 00:37:12< celmin> It's not clear whether Gg also matches eg Gg^Ii, for example. 20180812 00:37:31<+wesdiscordbot> It doesn't. 20180812 00:37:36<+wesdiscordbot> Not in SLFs, at least. 20180812 00:37:52< celmin> It's also not clear whether G* will match Gg^Ii. 20180812 00:37:58<+wesdiscordbot> Same thing applies. 20180812 00:38:17<+wesdiscordbot> In general if you don't specify the overlay filter you won't match terrains with an overlay. 20180812 00:38:21< celmin> So basically, leaving out an overlay means… right, that. 20180812 00:38:34< celmin> What if you specify ^Ii without a preceding *? 20180812 00:38:47< celmin> Would that match nothing because overlays can't appear alone? 20180812 00:38:53< celmin> Or would it be equivalent to *^Ii? 20180812 00:38:56<+wesdiscordbot> No, you'd destroy the universe instead. 20180812 00:39:00< celmin> Haha. 20180812 00:39:06<+wesdiscordbot> idk what would happen to be honest. It might be considered erroneous. 20180812 00:40:56-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection] 20180812 00:41:32< celmin> And w* would not match any water terrains, right? 20180812 00:41:40<+wesdiscordbot> Filters are case-sensitive. 20180812 00:42:20< celmin> According to the wiki, something like *d will not work, huh... 20180812 00:42:36< celmin> Is this still true I wonder? 20180812 00:42:41< celmin> Probably is, I guess. 20180812 00:42:42<+wesdiscordbot> I've never had a need or tried to use prefix globs. 20180812 00:42:49<+wesdiscordbot> Besides *^Whatever. 20180812 00:54:42< celmin> They probably wouldn't be very useful even if they did work. 20180812 00:54:51< celmin> Other than that one, obviously. 20180812 00:58:22<+wesdiscordbot> Just github cleanup, someone wanna close https://github.com/wesnoth/wesnoth/issues/2673 ? 20180812 01:05:01< celmin> So does !,Gg match anything except grass> 20180812 01:05:03< celmin> ^? 20180812 01:05:35< celmin> I don't know if I've got ! figured out correctly, but wouldn't !,Gg,! match nothing? 20180812 01:08:45< celmin> Oh yeah, does * match only terrains without overlays? 20180812 01:10:49<+wesdiscordbot> Inverted patterns always give me a migraine for some reason so I'll pass on this one. 20180812 01:12:02< celmin> Heh. 20180812 01:12:24< celmin> I guess what I should really do with this case is test it directly. 20180812 01:12:31< celmin> Blargh... 20180812 01:12:41< celmin> I guess I'll connect to Windows over VNC if it's still working... 20180812 01:14:53< celmin> (Surprisingly it is still working. Yay for things auto-starting on startup?) 20180812 01:15:09<+wesdiscordbot> Ya know, I've honestly really been enjoying this forced Wesbreak as I work on the website project I'm working on. 20180812 01:15:15< celmin> (Wish it supported keystroke encryption though.) 20180812 01:15:19< celmin> Oh hi Vultraz. 20180812 01:15:25<+wesdiscordbot> it's greatly relaxed me 20180812 01:15:39< celmin> Fun. 20180812 01:17:10< celmin> (Also it seems to be wrong about my keyboard layout. I can never type the yen character on Windows why can I suddenly type it over VNC.) 20180812 01:20:21< celmin> Okay so I got all the terrains matched by "!,Gg" and all the terrains matched by "Gg" and somehow they add up to more than the product of the map width and height? Does get_map_size() exclude the border maybe? 20180812 01:20:57< celmin> Well, if I consider the map 2 hexes wider in each dimension, I get the aforementioned sum. 20180812 01:21:03< celmin> Anyway it would appear that my guess is correct. 20180812 01:21:19< celmin> If you have an odd number of ! elements, reaching the end of the list without a match = success. 20180812 01:21:34< celmin> If you have an even number of ! elements, reaching the end of the list without a match = failure. 20180812 01:22:41< celmin> So for example !,Gg,!,! is equivalent to !,Gg 20180812 01:25:21< celticminstrel> New version: https://wiki.wesnoth.org/StandardLocationFilter#Filtering_Terrains 20180812 01:25:57-!- gfgtdf [~chatzilla@x4dbbe692.dyn.telefonica.de] has quit [Ping timeout: 240 seconds] 20180812 01:26:18< celticminstrel> Old version if you want to compare: https://wiki.wesnoth.org/index.php?title=StandardLocationFilter&oldid=59866#Filtering_Terrains 20180812 01:29:38-!- gfgtdf [~chatzilla@x2f6124b.dyn.telefonica.de] has joined #wesnoth-dev 20180812 01:55:43< gfgtdf> @Shadowm can you rebuidl the server please ? 20180812 01:55:54<+wesdiscordbot> The build passed? Okay. 20180812 01:56:53< gfgtdf> ye 20180812 02:03:02<+wesdiscordbot> Rebuilt and restarted. 20180812 02:06:35<+wesdiscordbot> Ugh. 20180812 02:06:36<+wesdiscordbot> https://cdn.discordapp.com/attachments/259976436490829825/478021452524683266/unknown.png 20180812 02:06:38<+wesdiscordbot> Not very usable with the tiny shrunk textboxes. 20180812 02:06:54<+wesdiscordbot> @Vultraz Any idea why they shrink? 20180812 02:08:01<+wesdiscordbot> AAAAAGGGHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH 20180812 02:08:31<+wesdiscordbot> i just 20180812 02:08:32<+wesdiscordbot> i dunno 20180812 02:08:35<+wesdiscordbot> maybe grow isn't set 20180812 02:08:49<+wesdiscordbot> maybe your res is too small and it chooses the wrong size 20180812 02:08:50<+wesdiscordbot> idk 20180812 02:10:18<+wesdiscordbot> after working with an actually competent layout system i don't want to touch GUI2 again. 20180812 02:10:33<+wesdiscordbot> It's 1280x749. 20180812 02:10:41<+wesdiscordbot> Wait, does that mean you're quitting? 20180812 02:10:48<+wesdiscordbot> I dunno. 20180812 02:11:15<+wesdiscordbot> i really dunno 20180812 02:12:26<+wesdiscordbot> it's just that GUI2 is painful 20180812 02:12:35<+wesdiscordbot> after seeing how much easier it can be 20180812 02:14:06<+wesdiscordbot> what's the other one? 20180812 02:14:15<+wesdiscordbot> "CSS Grid" 20180812 02:14:15<+wesdiscordbot> been working with css grids 20180812 02:14:23<+wesdiscordbot> Yep, my guess was correct. 20180812 02:14:26<+wesdiscordbot> Is that what's in master? 20180812 02:14:29<+wesdiscordbot> No. 20180812 02:14:34<+wesdiscordbot> He's been working with CSS Grid for a project of his. 20180812 02:14:51<+wesdiscordbot> CSS Grid is a (work in progress) CSS specification for use in web browsers. 20180812 02:15:05<+wesdiscordbot> I wonder if nanovg would be easier to use. 20180812 02:15:11<+wesdiscordbot> or less buggy than gui2 20180812 02:15:21<+wesdiscordbot> I'll be honest here for a moment. 20180812 02:15:53<+wesdiscordbot> "wesnoth for your web browser" 20180812 02:16:16<+wesdiscordbot> GUI2 wasn't a properly thought-out idea. It is an improvement over GUI1, yes, but it's always had a lot of issues that weren't even considered during the design phase. 20180812 02:16:22<+wesdiscordbot> Now hear me out. 20180812 02:16:29<+wesdiscordbot> ig GUI2 is ever replaced it will not be with a custom UI toolkit 20180812 02:16:43<+wesdiscordbot> The easiest solution for anyone working on this game that runs into difficulties with the GUI code is to quit working on this game. 20180812 02:16:44<+wesdiscordbot> it will be an existing library 20180812 02:16:51<+wesdiscordbot> :D 20180812 02:17:11<+wesdiscordbot> I'm putting it simply and plainly because no-one else dares spell it out in public even though it's always crossed our minds. 20180812 02:17:21<+wesdiscordbot> People just "disappear" and blame hardware issues and other stuff. 20180812 02:17:27<+wesdiscordbot> Okay, so that's the easiest solution. 20180812 02:17:32<+wesdiscordbot> No-one can argue against it. 20180812 02:17:42<+wesdiscordbot> Is it the most satisfying solution, though? 20180812 02:17:45<+wesdiscordbot> No, it is not. 20180812 02:17:51<+wesdiscordbot> Because the problem remains unsolved. 20180812 02:18:21-!- travis-ci [~travis-ci@ec2-54-242-206-169.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 02:18:22< travis-ci> gfgtdf/wesnoth#1213 (1.14 - b5a2b03 : gfgtdf): The build failed. 20180812 02:18:22< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415004300 20180812 02:18:22-!- travis-ci [~travis-ci@ec2-54-242-206-169.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 02:18:34<+wesdiscordbot> Is it impossible to use QT or TK for the gui as a replacement? 20180812 02:18:50<+wesdiscordbot> Tk looks awful. 20180812 02:18:52<+wesdiscordbot> Or is it just so much work no one wants to do it :D 20180812 02:18:58<+wesdiscordbot> I'm not even sure it's themable, is it? 20180812 02:19:11<+wesdiscordbot> I haven't been up to date with Tk for years. I've used Tkinter before, looked meh. 20180812 02:19:11<+wesdiscordbot> Also Qt is overkill. 20180812 02:19:20<+wesdiscordbot> I've worked with Qt before. 20180812 02:19:22<+wesdiscordbot> QT is a pain to work with, but those are the only toolkits I work with. 20180812 02:19:31<+wesdiscordbot> Well there's GTK, but I thought that was linux only 20180812 02:19:35<+wesdiscordbot> I wonder if wesnoth could be remade using electron 🤔 20180812 02:19:40<+wesdiscordbot> Gtk is not third-party-friendly. 20180812 02:19:49<+wesdiscordbot> Totally, and it'd probably be faster than the crappy codebase if done well. 20180812 02:19:55<+wesdiscordbot> It's designed for Gnome's use first, second, and last. 20180812 02:20:11<+wesdiscordbot> Third parties fit somewhere along the nth position in there. 20180812 02:20:20<+wesdiscordbot> They're not even second-class citizens. 20180812 02:20:33<+wesdiscordbot> Vultraz did you 20180812 02:20:34<+wesdiscordbot> Like 20180812 02:20:36<+wesdiscordbot> Oh my god. 20180812 02:20:38<+wesdiscordbot> Alright let's be real, there are probably no good toolkits available for UI in non-web languages., there are just workable ones. 20180812 02:20:47<+wesdiscordbot> Please, don't ever consider using Electron. 20180812 02:20:50<+wesdiscordbot> Please. 20180812 02:21:03<+wesdiscordbot> I have THREE WEB BROWSERS OPEN because everyone wants to use Electron. 20180812 02:21:17<+wesdiscordbot> I'd have a fourth if I launched VS Code right now. 20180812 02:21:31<+wesdiscordbot> And you know what the worst part is? 20180812 02:21:45<+wesdiscordbot> None of their memory is shareable (without kernel-based heuristics at least). 20180812 02:22:02<+wesdiscordbot> Do you realize how heavy weight web browsers are these days? 20180812 02:22:48<+wesdiscordbot> aw, and ram is so expensive these days 20180812 02:23:02<+wesdiscordbot> ram is cheap 20180812 02:23:06<+wesdiscordbot> Well I just had to set up a brand-new laptop with 4 GB of RAM and no ability to expand it. 20180812 02:23:09<+wesdiscordbot> Some factories in the indochina area flooded like 6 years ago and prices never bounced back 20180812 02:23:26<+wesdiscordbot> @Vultraz RAM is insanely more expensive now that it's been 20180812 02:23:52<+wesdiscordbot> I dunno about that vultraz, I bought 16gb for $75 in 2012 and the same kit (in ddr4 and appropriately faster now) is about $145-150 now. 20180812 02:24:07<+wesdiscordbot> I mean it's great that some people have the purchase power to get fancy 16 GB laptops, but if my desktop were to die right now I'd get stuck with an even older laptop with only 4 GB of RAM. 20180812 02:24:12<+wesdiscordbot> For possibly years. 20180812 02:24:40<+wesdiscordbot> Which struggles to render Twitter at more than 20 fps. 20180812 02:24:45<+wesdiscordbot> i wonder if my laptop is at max memory 20180812 02:24:47<+wesdiscordbot> it's at 16 20180812 02:25:02<+wesdiscordbot> You can find that out trivially. 20180812 02:25:06<+wesdiscordbot> how? 20180812 02:25:12<+wesdiscordbot> Look up the model's specifications. 20180812 02:25:26<+wesdiscordbot> They always mention whether the motherboard supports more RAM. 20180812 02:26:13<+wesdiscordbot> My previous laptop before that had 2 GB of RAM and supported up to 8 GB but the existing configuration meant that I had to make do with 4. 20180812 02:26:50<+wesdiscordbot> I got sidetracked though. 20180812 02:26:52<+wesdiscordbot> up to 16. ok, i have max 20180812 02:27:17<+wesdiscordbot> My point is, you can either keep torturing yourself with GUI2, quit Wesnoth, or replace GUI2 with an existing solution for SDL 2. There probably is one out there. 20180812 02:27:26<+wesdiscordbot> Your cpu might also (often does) support less ram than your motherboard 20180812 02:27:47<+wesdiscordbot> *whether the laptop 20180812 02:27:55<+wesdiscordbot> so who's in charge of that SDL port? 20180812 02:28:00<+wesdiscordbot> (Because laptop manufacturers don't really make a distinction between mobo and CPU support for obvious reasons.) 20180812 02:28:03<+wesdiscordbot> and will it use SDL sound also? 20180812 02:28:13<+wesdiscordbot> Oh, sure for laptops, shadow. 20180812 02:28:14<+wesdiscordbot> we already use sdl 20180812 02:28:17<+wesdiscordbot> and sdl_mixer 20180812 02:28:24<+wesdiscordbot> ah, right 20180812 02:30:54<+wesdiscordbot> The tl;dr is that NIH syndrome as perpetuated by GUI2 is hindering the project. It's not helping it at all, so the only viable solution going forward is to excise that tumour before it's too late. 20180812 02:31:18<+wesdiscordbot> Not In House? 20180812 02:31:28<+wesdiscordbot> Not Invented Here. 20180812 02:31:44<+wesdiscordbot> https://en.wikipedia.org/wiki/Not_invented_here 20180812 02:32:25<+wesdiscordbot> yeah definitely harder to get new devs in also if it's an in-house thing. 20180812 02:33:04<+wesdiscordbot> I don't know if any of you have ever gotten job offers to write in a private custom proprietary language and they ask you how many years of experience you have with it... 20180812 02:33:21<+wesdiscordbot> given that the Theme ingame UI was already removed from master anyway, now seems like as good a time to do it as any. 20180812 02:36:00<+wesdiscordbot> Well 20180812 02:36:02<+wesdiscordbot> https://en.wikipedia.org/wiki/List_of_widget_toolkits 20180812 02:36:21<+wesdiscordbot> GTK isn't great outside of (gnome) linux, QT is overkill for shadow, and Tk sucks. 20180812 02:36:38<+wesdiscordbot> I remember people mentioned CEGUI a couple of years ago. 20180812 02:36:57<+wesdiscordbot> the only reason i said electron is it would be nice to get the css layout engine 😛 20180812 02:37:07<+wesdiscordbot> hides 20180812 02:37:12<+wesdiscordbot> is not completely sure how she feels about people calling her shadow without the m. 20180812 02:39:01<+wesdiscordbot> sorreee 20180812 02:40:13<+wesdiscordbot> Wasn't sure if you had an _ in IRC to kill discord highlights 20180812 02:40:53<+wesdiscordbot> I keep those just in case. 20180812 02:41:24<+wesdiscordbot> But I don't look at them most of the time. 20180812 02:51:04<+wesdiscordbot> heh 20180812 02:51:07<+wesdiscordbot> Crazy Eddie's GUI 20180812 02:52:58<+wesdiscordbot> Wow, backend in Ogre and irrlicht. 20180812 02:57:58-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180812 02:58:04-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180812 02:59:25-!- travis-ci [~travis-ci@ec2-54-242-206-169.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 02:59:26< travis-ci> gfgtdf/wesnoth#1214 (1.14 - 7d40a7f : gfgtdf): The build is still failing. 20180812 02:59:26< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415011363 20180812 02:59:26-!- travis-ci [~travis-ci@ec2-54-242-206-169.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 03:08:53-!- gfgtdf [~chatzilla@x2f6124b.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.93 [Firefox 52.9.0/20180621064021]] 20180812 03:09:35< irker955> wesnoth/wesnoth:1.14 Jyrki Vesterinen 1eff24b58e Hide battery status if the device doesn' AppVeyor: All builds passed 20180812 03:20:45-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180812 03:20:50-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180812 03:20:55< celmin> [Aug 11@10:20:47pm] wesdiscordbot: Please, don't ever consider using Electron. 20180812 03:20:56< celmin> I second, third, fourth, and fifth this. After seeing how much VS Code lags when I have two instances plus (sometimes) UE4 Editor open, I really don't want to have to use another electron app. (I suppose this could be also partly Windows 10's fault since I don't seem to have any problems when working on Wesnoth stuff in VS Code, though.) 20180812 03:21:32< celmin> [Aug 11@10:30:54pm] wesdiscordbot: The tl;dr is that NIH syndrome as perpetuated by GUI2 is hindering the project. It's not helping it at all, so the only viable solution going forward is to excise that tumour before it's too late. 20180812 03:21:33< celmin> By "excise that tumour" do you mean "remove GUI2"? 20180812 03:21:41<+wesdiscordbot> Yes. 20180812 03:21:45< celmin> I can't say I'm fundamentally opposed to that idea, mind you... 20180812 03:22:44< celmin> It doesn't feel worthwhile to do it, but if someone managed to do it, I can't see the situation being any worse. 20180812 03:23:03< celmin> I guess it would be better, the problem is just that initial hump of doing it. 20180812 03:23:10<+wesdiscordbot> Well if you think you can fix the fundamental issues with GUI2's layout system, by all means. 20180812 03:23:15< celmin> ALSO 20180812 03:23:35<+wesdiscordbot> Most of them are pretty obvious once you use a translation (which most people in the world do because there are more non-English-speaking countries than English-speaking ones). 20180812 03:23:53< celmin> Does "removing GUI2" require using a completely new markup? If the backend (layout, events, etc) can somehow be swapped out without appreciably touching the frontend (WML, parsing, etc), that would be even better. 20180812 03:24:11<+wesdiscordbot> Dump it all into the ocean, yes. 20180812 03:24:36<+wesdiscordbot> I think we're at the point where we can justify a full API break in the name of not dooming the project to a slow and painful death. 20180812 03:24:57< celmin> Hmm. 20180812 03:25:01<+wesdiscordbot> A lot of people will say "no" but they aren't the ones who will take over if vultraz and other key people quit. 20180812 03:25:28<+wesdiscordbot> Therein lies the issue with taking a dogmatic approach to backwards compatibility. 20180812 03:26:06< celmin> I do think though that writing code to parse GUI2 WML into something a new backend understands might not be as hard as it sounds. 20180812 03:26:21< celmin> So if it's not too much extra effort, we should do it. 20180812 03:26:25<+wesdiscordbot> As well as with prioritizing the needs of people who are only users, not developers, and therefore will never need to be concerned with having to keep this whole machinery running. 20180812 03:26:44< celmin> At least as an interim measure so that we don't need to redesign every single dialog all at once. 20180812 03:26:47<+wesdiscordbot> That's honestly secondary. 20180812 03:26:50<+wesdiscordbot> celmin, I've never had issues with VS Code lagging with UE open. 20180812 03:26:57<+wesdiscordbot> I mean if someoen can figure it out after it's done, props to them. 20180812 03:27:06<+wesdiscordbot> VS Code is probably the least laggy editor I've used besides emacs. 20180812 03:27:09< celmin> @lilinitsy - I don't think UE is specifically the problem here TBH. 20180812 03:27:15<+wesdiscordbot> But by no means it should be a requirement. 20180812 03:27:22< celmin> Because it happens whether or not UE is open. 20180812 03:27:34< celmin> Usually when I'm trying to build. 20180812 03:27:50< celmin> More specifically it's usually when I'm trying to build with UBT. 20180812 03:28:04<+wesdiscordbot> i do get some lag in vs when compiling 20180812 03:28:10<+wesdiscordbot> but that happens everywhere 20180812 03:28:29<+wesdiscordbot> VS and emacs are the only editors I've never had lag on me. 20180812 03:28:33< celmin> This might be why I haven't experienced it yet when doing Wesnoth stuff at home, because so far that's been only campaign work and some Lua stuff. 20180812 03:28:34<+wesdiscordbot> (vs code, not vs) 20180812 03:28:42< celmin> Some lag, sure. 20180812 03:28:50<+wesdiscordbot> Windows is notoriously bad at scheduling processes so that interactive processes can keep working under heavy load (whether it is computational or I/O-bound). 20180812 03:28:54< celmin> But ten-minute freezes? 20180812 03:29:13<+wesdiscordbot> At least by default/the regular SKUs. 20180812 03:29:17< celmin> Yeah, I figured it had to be at least partly Windows's fault. 20180812 03:29:19< celmin> SKU? 20180812 03:29:34<+wesdiscordbot> I think server SKUs by default prioritize background processes and services in particular. 20180812 03:29:53<+wesdiscordbot> It's an option that can be enabled by users but it's probably a bad idea for obvious reasons. 20180812 03:29:58<+wesdiscordbot> 10 minute freezes? 20180812 03:30:03<+wesdiscordbot> sound like your computer is just bad 20180812 03:30:05< celmin> I'm not really sure but I'm thinking it's also possible I'm running into swap overflow or something. 20180812 03:30:21<+wesdiscordbot> You should be using task manager to keep an eye on that. 20180812 03:30:39< celmin> It's a somewhat high-end computer, I think? I don't really know. I have task manager open most of the time but once it freezes up that doesn't help. 20180812 03:31:08< celmin> Another thing that probably doesn't help is that every few minutes VS Code decides it needs to poll the perforce server. 20180812 03:31:10<+wesdiscordbot> You'd know whether it is memory thrashing then. 20180812 03:31:38<+wesdiscordbot> Also as a side-note, Electron applications are important GPU users. 20180812 03:31:39< celmin> Sometimes it gets close to 99% CPU, other times it's instead getting close to 99% memory whatever that means. 20180812 03:31:48<+wesdiscordbot> That's particularly problematic on Linux with NVIDIA. 20180812 03:32:00< celmin> But not on Windows with NVIDIA? 20180812 03:32:24<+wesdiscordbot> I don't care about Windows with NVIDIA. 20180812 03:32:32<+wesdiscordbot> I run Windows on NVIDIA like once a month. 20180812 03:32:37< celmin> Shrug. 20180812 03:33:00<+wesdiscordbot> If Microsoft didn't make their platform actively hostile towards me I'd feel more inclined to use it. 20180812 03:33:06< celmin> Heh. 20180812 03:33:14<+wesdiscordbot> And by hostile I mean it literally costs me money. 20180812 03:33:19< celmin> I'm not too inclined to use it either TBH, if I can possibly avoid it. 20180812 03:33:39< celmin> I feel like Windows 10 is half malware. 20180812 03:34:04<+wesdiscordbot> That's only 25% of my bone to pick with it. 20180812 03:34:16<+wesdiscordbot> The rest of the bone (yes this is a weird analogy now) is bandwidth usage. 20180812 03:34:21< celmin> It might be possible with liberal pruning to get it into a state that's reasonable. 20180812 03:34:37< celmin> But I have no idea which things are safe to prune and which things would break everything if I try. 20180812 03:34:37<+wesdiscordbot> "i will download 6 GB WORTH OF UPDATES A MONTH because i wuv you <3" 20180812 03:35:06< celmin> I managed to destroy the Windows Update service by liberally deleting things from the registry and then also deleting all the on-disk files related to them. 20180812 03:35:30< celmin> Which is good because it doesn't try to update without asking but also bad because now I can't update. 20180812 03:36:59<+wesdiscordbot> Why did you do that instead of installing Linux 20180812 03:37:14< celmin> Pretty sure that's not an option? It's technically not my computer. 20180812 03:37:23<+wesdiscordbot> Note that I'm not being a stupid Linux fangirl here. I just know exactly how much effort you went through to achieve that. 20180812 03:37:31<+wesdiscordbot> And it'd have been far easier to switch platforms. 20180812 03:37:44< celmin> Besides, Windows is the primary target platform, so developing on Linux seems like it would be silly. 20180812 03:37:54<+wesdiscordbot> Not to mention you risked bricking your Windows install. 20180812 03:38:25< celmin> There seems to be some sort of recovery partition in case of that scenario though. 20180812 03:38:34< celmin> But it's not like I deleted things blindly. 20180812 03:38:53<+wesdiscordbot> There are multiple components surreptitiously ensuring that certain system services are running at all times and that their configuration in the Registry exists and makes sense. 20180812 03:38:54< celmin> It was specifically things related to services and the scheduler, which the services manager and scheduling manager wouldn't let me delete. 20180812 03:39:07<+wesdiscordbot> The task scheduler doesn't have anything to do with updates. 20180812 03:39:12<+wesdiscordbot> I'm still running 7 without any compatability issues. 20180812 03:39:12< celmin> Surprisingly, it does. 20180812 03:39:28<+wesdiscordbot> Not really. It's just some kind of cron equivalent. 20180812 03:39:30< celmin> There's like five or six tasks in the scheduler in Windows 10 relating to updates 20180812 03:39:46<+wesdiscordbot> I think those are just maintenance tasks. 20180812 03:39:51<+wesdiscordbot> Like the thing that purges old updates. 20180812 03:39:51< celmin> @lilinitsy At home, so am I. 20180812 03:40:02<+wesdiscordbot> Or the thing that compresses (?) old updates. 20180812 03:40:05<+wesdiscordbot> well, on my desktop. I do my non UE/Unity dev on linux, like wesnoth 20180812 03:40:46< celmin> @shadowm I'm not quite sure what they were, but something was continuously resurrecting windows every few days, and if it's not something in services, I figure it had to have been something in the scheduler. 20180812 03:40:53< celmin> ^windows update 20180812 03:41:16< celmin> And even when I deleted those entries in the scheduler, something kept resurrecting them. 20180812 03:41:33<+wesdiscordbot> Well, that's where the surreptitious part comes into action. 20180812 03:41:36< celmin> This is why I compared it to malware. You try to delete it out of desperation, yet it keeps coming back. 20180812 03:42:12<+wesdiscordbot> I'll say, however, that this is in part precisely to deter attempts from actual malware to mess with the system. 20180812 03:42:31< celmin> Eventually, after deleting the actual files too, I got it to the point where it gave me a few alerts about "we can't seem to update, check MS to troubleshoot the issue" and then finally it even gave up on that. 20180812 03:42:38< celmin> Maybe. 20180812 03:43:29<+wesdiscordbot> And that's the thing. 20180812 03:43:37<+wesdiscordbot> You had to disable system file protection to delete the files. 20180812 03:44:00<+wesdiscordbot> And in the process you risked making several components freak out about failing to load required files during startup. 20180812 03:44:11< celmin> But when it has a system in place to manage the OS, it should really let that system apply to everything. Disable the update service? Sorry, you don't have permission. Disable the scheduled task? Sorry, you don't have permission. This is an administrator account, why don't I have permission? 20180812 03:44:31< celmin> I don't recall disabling file system protection, but I guess I did get an administrator prompt when deleting the files. 20180812 03:44:49<+wesdiscordbot> Even with an UAC prompt the files would've popped right back in. 20180812 03:44:52< celmin> It sort of worked out somehow, maybe I was a little bit lucky. 20180812 03:44:58< celmin> …huh? 20180812 03:45:09<+wesdiscordbot> That's how system file protection works. 20180812 03:45:44<+wesdiscordbot> A process (don't remember which, and I think these days it's moved into the kernel) actively monitors system paths to make sure neither you nor other unauthorized entities mess with them. 20180812 03:45:49< celmin> Anyway I do want to update at some point but now that I broke windows update it just feels like too much effort to fix it and update. 20180812 03:46:08<+wesdiscordbot> There's an intricate permissions system going on so that Windows Update itself can tamper with system files. 20180812 03:46:45< celmin> (There was an actual reason for not updating, one of the updates broke the version of UE we were using, but then we upgraded UE so that problem no longer exists, so I could update now if I wanted to.) 20180812 03:46:55<+wesdiscordbot> (I'm pretty sure back in the Windows XP days the process in charge of system file protection was winlogon.exe.) 20180812 03:48:41< celmin> (But I kinda figure updating now would require actually doing a reinstall from the recovery partition which just feels like way more effort than it's worth… but I guess I'll have to do it eventually…) 20180812 04:12:15<+wesdiscordbot> Apparently these days it's implemented using ACLs on system files taking advantage of the fact that Windows requires NTFS starting with Vista. 20180812 04:12:34<+wesdiscordbot> But working around that still takes some effort. 20180812 04:13:18<+wesdiscordbot> Namely changing permissions or transferring ownership of files to a different account. 20180812 05:18:36-!- celmin [~celticmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The minstrel departs, to spread the music to the masses!] 20180812 05:19:58-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] 20180812 05:34:58< irker955> wesnoth/wesnoth:1.14 gfgtdf 717a43c490 Revert "preserve traslatable strings in AppVeyor: All builds passed 20180812 06:50:58-!- iwaim [~iwaim@rasteenie.alib.jp] has quit [Ping timeout: 260 seconds] 20180812 06:55:22-!- iwaim [~iwaim@rasteenie.alib.jp] has joined #wesnoth-dev 20180812 07:08:46-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev 20180812 07:13:54-!- valdar [~atarocch@37.176.100.2] has joined #wesnoth-dev 20180812 08:35:15-!- irker955 [~irker@wesnoth/umc-dev/bot/shikadibot] has quit [Quit: transmission timeout] 20180812 09:16:35-!- valdar [~atarocch@37.176.100.2] has quit [Ping timeout: 240 seconds] 20180812 10:00:41<+wesdiscordbot> Nice discussion... I was sleeping during your discussion about GUI2. In my opinion, old GUI1 was easier to use. I am working with GUI2 on TinyGUI project and it is insane... No documentation, no debug options, nothing... IMHO if we want to keep GUI2 framework, we should FINISH it first. It seems to be unfinished alpha to me... (Ok, it's more stable than alpha, so let's call it GUI2 beta). 20180812 10:01:27<+wesdiscordbot> Another option is restore GUI1 but I totally don't know if [resolution] tags are supported in GUI1 20180812 10:02:17<+wesdiscordbot> Totally new framework? Well, maybe in few years with Wesnoth 1.15 and 1.16 but not tomorrow or next week with 1.14 20180812 10:03:19<+wesdiscordbot> In case of new GUI framework I would prefer something like this: https://stackoverflow.com/questions/17517249/how-to-build-native-c-apps-with-html-css-ui 20180812 10:03:50<+wesdiscordbot> I don't know cssgrids but basic CSS should be enough for wesnoth's needs 20180812 10:03:59<+wesdiscordbot> .:EOF:. 20180812 10:04:36<+wesdiscordbot> It's true that GUI2 is unfinished. 20180812 10:04:55<+wesdiscordbot> I improved it in a couple of places when such improvements were necessary. 20180812 10:05:49<+wesdiscordbot> GUI1 is not suitable for our use cases, that's why GUI2 was born. 20180812 10:05:59<+wesdiscordbot> So no, don't even think of "restoring" it. 20180812 10:11:14-!- nurupo [~nurupo.ga@unaffiliated/nurupo] has quit [Quit: nurupo] 20180812 10:11:55-!- nurupo [~nurupo.ga@unaffiliated/nurupo] has joined #wesnoth-dev 20180812 10:14:22<+wesdiscordbot> Do you know EAWebKit ? 20180812 10:14:36<+wesdiscordbot> https://gpl.ea.com/eawebkit.html 20180812 10:15:41<+wesdiscordbot> Is there any link where to point people who want to use GUI2, let's say for a shopsystem? 20180812 10:16:46<+wesdiscordbot> You want documentation for GUI2 ? 20180812 10:17:14<+wesdiscordbot> yes. not necessarily in-depth, more like 'how to start' 20180812 10:20:27<+wesdiscordbot> Well, GUI2 have really poor documentation. You have to regularly ask people here. You can start with studying existing WML sources: https://github.com/wesnoth/wesnoth/tree/1.14/data/gui/window (for example title_screen.cfg) 20180812 10:20:51<+wesdiscordbot> Then there is a wesnoth wiki: https://wiki.wesnoth.org/ 20180812 10:21:56<+wesdiscordbot> Interesting documents which I found: - https://wiki.wesnoth.org/GUIWindowDefinitionWML - https://wiki.wesnoth.org/GUIWidgetDefinitionWML - https://wiki.wesnoth.org/GUILayoutGuide - https://wiki.wesnoth.org/GUILayout - https://wiki.wesnoth.org/GUIWidgetInstanceWML 20180812 10:22:47<+wesdiscordbot> Or simply start here: https://wiki.wesnoth.org/GUIToolkit 20180812 10:31:02-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20180812 10:48:50<+wesdiscordbot> That looks quite good 20180812 11:53:10-!- louis94 [~~louis94@109.49-136-217.adsl-dyn.isp.belgacom.be] has joined #wesnoth-dev 20180812 11:56:29<+wesdiscordbot> https://developer.nvidia.com/transitioning-opengl-vulkan 20180812 13:12:34<+wesdiscordbot> If you will add SDL_LogSetAllPriority(SDL_LOG_PRIORITY_VERBOSE); into main(), you will got full log of this output: 20180812 13:12:45<+wesdiscordbot> 2018-08-12 15:11:30.518722+0200 Wesnoth[15270:4073310] DEBUG: Invalid renderer 2018-08-12 15:11:30.519635+0200 Wesnoth[15270:4073310] DEBUG: Invalid renderer 2018-08-12 15:11:30.519831+0200 Wesnoth[15270:4073310] DEBUG: Invalid renderer 20180812 13:12:59<+wesdiscordbot> Is it reproducible outside macOS? 20180812 13:14:03<+wesdiscordbot> Also this bug seems to be upstream: 20180812 13:14:04<+wesdiscordbot> https://cdn.discordapp.com/attachments/259976436490829825/478189424832806913/unknown.png 20180812 13:14:29<+wesdiscordbot> (You can reproduce it with DEBUG version of SDL 2.0.8) 20180812 13:58:42<+wesdiscordbot> I am on right track with issue number 1 in this PR: https://github.com/wesnoth/wesnoth/pull/3451#issue-207773065 This issue is NOT cause by Xcode 10 but by macOS 10.14 SDK. Please merge this PR as all problems with Xcode 10 are fixed and I'll open new issue for tracking macOS 10.14 specific issues. 20180812 14:28:03<+wesdiscordbot> Created a PR. Let's see if it compiles, and I go to sleep meanwhile. Tomorrow is a long day. Have a good one 😃 20180812 14:29:40-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev 20180812 14:43:10-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.] 20180812 14:51:30-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev 20180812 15:00:42<+wesdiscordbot> Is GUI1 still used anywhere? 20180812 15:01:32<+wesdiscordbot> I think the help uses it 20180812 15:07:23-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev 20180812 15:09:06-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc] 20180812 15:10:46<+wesdiscordbot> It sounds like it'd be even more worth it to move to a single non-custom toolkit then, seeing as right now that means there's three different and custom UI toolkits being used - ThemeWML, GUI1, and GUI2. 20180812 15:11:31<+wesdiscordbot> Unfortunately rewriting all GUI code in Wesnoth would require way too much effort. 20180812 15:15:49< celticminstrel> ThemeWML is on top of GUI2 mind you. 20180812 15:16:00< celticminstrel> And yeah that was my thought too Jyrki. 20180812 15:16:26<+wesdiscordbot> I don't see anything that uses GUI2 in the ThemeWML implementation. 20180812 15:16:31< celticminstrel> If someone can get the switch done it'd be really nice, but the effort of making the switch is so huge. 20180812 15:16:37< celticminstrel> Whoops, that was a typo. 20180812 15:16:41< celticminstrel> I meant GUI1. 20180812 15:25:02-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: KABOOM! It seems that I have exploded. Please wait while I reinstall the universe.] 20180812 15:29:53-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev 20180812 15:34:46<+wesdiscordbot> Based on the discussion yesterday though, there's effectively no redeeming qualities of GUI2 other than it's better than GUI1, and having to use them has ended up driving multiple other developers away from the project over the years. I don't really know what would be involved in replacing them, but the effort to keep maintaining it for however many years into the future also sounds non-trivial, especially if it 20180812 15:34:47<+wesdiscordbot> means that Vultraz is going to leave as well rather than have to deal with it anymore. 20180812 15:37:52<+wesdiscordbot> The maintenance effort isn't that big as long as we don't need to make GUI changes. 20180812 15:39:18-!- irker724 [~irker@wesnoth/umc-dev/bot/shikadibot] has joined #wesnoth-dev 20180812 15:39:18< irker724> wesnoth: Martin Hrubý (hrubymar10) wesnoth:Fix_Wesnoth_for_Xcode_10 734acb4082cc / / (8 files in 3 dirs): Fix CodeBlocks and Scons builds https://github.com/wesnoth/wesnoth/commit/734acb4082cca47ec1d7cd23ca3ad4fe79aa7bd2 20180812 15:39:18< irker724> wesnoth: Martin Hrubý (hrubymar10) wesnoth:Fix_Wesnoth_for_Xcode_10 43aa192ace0c / src/video.cpp: Include SDL_WINDOW_ALLOW_HIGHDPI flag https://github.com/wesnoth/wesnoth/commit/43aa192ace0c5678db604c9c7f5e03c991f4d323 20180812 15:39:19< irker724> wesnoth: Martin Hrubý (hrubymar10) wesnoth:Fix_Wesnoth_for_Xcode_10 eb4a09d331cf / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj src/video.cpp: Fix indentation and force using tabs in Xcode project https://github.com/wesnoth/wesnoth/commit/eb4a09d331cf40f0e6edf69a00c22f0206268673 20180812 15:39:41< irker724> wesnoth: Martin Hrubý (hrubymar10) wesnoth:1.14 49f673679884 / / (75 files in 26 dirs): Fix building with Xcode 10 by renaming image, version and settings in root of sr https://github.com/wesnoth/wesnoth/commit/49f6736798842c4613e2ab8a7c7c1dfafa734422 20180812 15:39:43< irker724> wesnoth: Jyrki Vesterinen wesnoth:1.14 ac43a2d3ed9b / projectfiles/VC12/ (4 files): Update Visual Studio project https://github.com/wesnoth/wesnoth/commit/ac43a2d3ed9b855f932278a6f88cab1cffae20e1 20180812 15:39:45< irker724> wesnoth: Martin Hrubý (hrubymar10) wesnoth:1.14 734acb4082cc / / (8 files in 3 dirs): Fix CodeBlocks and Scons builds https://github.com/wesnoth/wesnoth/commit/734acb4082cca47ec1d7cd23ca3ad4fe79aa7bd2 20180812 15:39:47< irker724> wesnoth: Martin Hrubý (hrubymar10) wesnoth:1.14 43aa192ace0c / src/video.cpp: Include SDL_WINDOW_ALLOW_HIGHDPI flag https://github.com/wesnoth/wesnoth/commit/43aa192ace0c5678db604c9c7f5e03c991f4d323 20180812 15:39:49< irker724> wesnoth: Martin Hrubý (hrubymar10) wesnoth:1.14 eb4a09d331cf / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj src/video.cpp: Fix indentation and force using tabs in Xcode project https://github.com/wesnoth/wesnoth/commit/eb4a09d331cf40f0e6edf69a00c22f0206268673 20180812 15:39:51< irker724> wesnoth: Celtic Minstrel wesnoth:1.14 543d35447f06 / / (87 files in 30 dirs): Merge pull request #3451 from wesnoth/Fix_Wesnoth_for_Xcode_10 https://github.com/wesnoth/wesnoth/commit/543d35447f06539f0e486c569dba503709d5a5d0 20180812 15:47:05<+wesdiscordbot> well, yeah, we could just stick with a horrible toolkit forever. 20180812 15:48:12<+wesdiscordbot> but it means we'll never be able to improve the UI 20180812 15:48:18<+wesdiscordbot> any further 20180812 15:48:33< irker724> wesnoth: Celtic Minstrel wesnoth:illuminating_campfire 1f8c7d3fef7d / data/core/terrain.cfg: Make the fire terrains actually illuminate their hex https://github.com/wesnoth/wesnoth/commit/1f8c7d3fef7d32d204bdb53ca3dc3e82f2e433b2 20180812 15:48:42<+wesdiscordbot> You're overexaggerating. 20180812 15:49:45<+wesdiscordbot> GUI improvements are still possible with GUI2. And depending on how much imporvements we're making, it may be significantly less effort to improve a GUI2 dialog than, you know, rewrite all UI code in the whole game. 20180812 15:50:06<+wesdiscordbot> Programmers tend to badly underestimate how much work rewriting is. 20180812 15:50:29<+wesdiscordbot> https://www.joelonsoftware.com/2000/04/06/things-you-should-never-do-part-i/ 20180812 15:50:38< celticminstrel> Yeah, this is a) why it feels like too much effort to switch backends and b) why if we do switch backends I think we need to retain the GUI WML parser. 20180812 15:51:05-!- louis94 [~~louis94@109.49-136-217.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 244 seconds] 20180812 15:52:49<+wesdiscordbot> Can you summarize all things that you hate on GUI2? In best case create issue for it or new wiki page. 20180812 15:53:19<+wesdiscordbot> Why? Because we don’t want to do same mostake twice 20180812 15:53:20< celticminstrel> The layout engine is generally quite bad at deciding whether a given element can fit in the available space. 20180812 15:56:48<+wesdiscordbot> no relative sizes, only absolute ones 20180812 15:57:00<+wesdiscordbot> stacking things is a bloated mess 20180812 15:57:09<+wesdiscordbot> things are hard to precisely position 20180812 15:57:12< celticminstrel> Grow factors are relative sizes in the same way you described for CSS. 20180812 15:57:42<+wesdiscordbot> i meant things beside pixel sizes for, say, dialogs 20180812 15:57:53<+wesdiscordbot> because that means everything's a different size on a different screen 20180812 15:58:09<+wesdiscordbot> and that resolutions need to be accounted for with separate tags. 20180812 15:58:16<+wesdiscordbot> there's no cell-spanning in grids 20180812 15:58:23<+wesdiscordbot> the markup takes ages to write 20180812 15:58:30<+wesdiscordbot> the boilerplate for new widgets is absurd 20180812 15:59:13<+wesdiscordbot> there are a million tiny little flaws everywhere 20180812 15:59:28<+wesdiscordbot> there's no way to position anything other than in a grid 20180812 15:59:36<+wesdiscordbot> there's no such thing as animations 20180812 15:59:50< celticminstrel> Pretty sure animations are doable, if not easy. 20180812 15:59:52<+wesdiscordbot> or any way to place an animated element on the screen in a UI context 20180812 16:00:04< celticminstrel> Though only within a widget. 20180812 16:00:07<+wesdiscordbot> the main menu code alone is a mess of magic numbers 20180812 16:00:29<+wesdiscordbot> there's the whole broken mess of interactions between gui2 and the map on master 20180812 16:01:26<+wesdiscordbot> too often, random scrollbars decide to appear because we only accounted for English 20180812 16:02:07<+wesdiscordbot> dynamic layouts that change are hard work with 20180812 16:02:11<+wesdiscordbot> like the mp lobby 20180812 16:02:14< celticminstrel> At least that's better than the text not showing at al like with GUI1. 20180812 16:02:21<+wesdiscordbot> the layout engine doesn't do well 20180812 16:02:25< celticminstrel> True, true 20180812 16:02:29<+wesdiscordbot> we had to hack in fixed-sizes 20180812 16:02:46<+wesdiscordbot> we sometimes get that fucking bonkers bug where text all renders on one line. 20180812 16:02:51<+wesdiscordbot> there's no high DPI support 20180812 16:03:23<+wesdiscordbot> aligning listbox headers to column contents is black magic 20180812 16:03:29<+wesdiscordbot> I know you maybe told it thousand times already, but why wesnoth should migrate to OpenGL? 20180812 16:03:47<+wesdiscordbot> because SDL alone is not enough 20180812 16:03:49<+wesdiscordbot> we need shaders 20180812 16:04:12<+wesdiscordbot> A much more important reason is performance. 20180812 16:04:52<+wesdiscordbot> Wesnoth's software rendering pipeline is designed for situations where the window is small and not much is changing on screen. 20180812 16:05:19<+wesdiscordbot> Those assumptions are badly breaking down, due to increasing screen resolutions, as well as things like animated water. 20180812 16:05:34<+wesdiscordbot> indeed 20180812 16:05:49<+wesdiscordbot> Ok, I get it. 20180812 16:05:53<+wesdiscordbot> (another random gui2 bug: 2 listboxes cannot sit vertically ) 20180812 16:05:55< celticminstrel> Vultraz's answer technically doesn't really say anything. 20180812 16:06:27<+wesdiscordbot> It's not a bug, just a result of the design. 20180812 16:06:34<+wesdiscordbot> we need shaders for tod lighting, celmin 20180812 16:06:56<+wesdiscordbot> Because there is no absolute sizing, two listboxescan't agree about how they should distribute available space. 20180812 16:07:26<+wesdiscordbot> surely, you of all people see the deep fissures in gui2's design 20180812 16:07:30<+wesdiscordbot> and its deficiencies 20180812 16:07:36<+wesdiscordbot> even if you don't support replacing it 20180812 16:07:37< celticminstrel> Are you talking to me? 20180812 16:07:41<+wesdiscordbot> jyrki 20180812 16:07:45< celticminstrel> Ah okay. 20180812 16:07:59<+wesdiscordbot> Yeah, sure, I do. 20180812 16:08:13<+wesdiscordbot> I just strongly dislike when people randomly label expected behavior as bugs. 20180812 16:08:33<+wesdiscordbot> i don't see it as expected. 20180812 16:08:35<+wesdiscordbot> but whatever 20180812 16:08:47<+wesdiscordbot> I was thinking how can I help (save) wesnoth. I want to because wesnoth is my childhood 😄 20180812 16:09:00<+wesdiscordbot> i don't know at this point 20180812 16:11:14<+wesdiscordbot> @hrubymar10 My advice is to simply focus on small things instead of these major issues. 20180812 16:11:24<+wesdiscordbot> Any commit that improves the game is welcome. 20180812 16:12:05<+wesdiscordbot> As long as we're moving forward, it's fine, even if there are major technical problems. 20180812 16:12:50<+wesdiscordbot> What is not okay to me is moving backwards, i.e. breaking things in the name of "improvements" (see: master branch). 20180812 16:13:29<+wesdiscordbot> My "playground" now is macOS package and iOS GUI 20180812 16:14:11< celticminstrel> Yeah, merging the OpenGL branch before it was done was... not the best of decisions. 20180812 16:14:21<+wesdiscordbot> Problem with iOS GUI is, that it is GUI2 based so if wesnoth will remove GUI2 I'll basically lost all of my work 😄 20180812 16:14:38<+wesdiscordbot> The OpenGL branch has not been merged.. 20180812 16:14:49< celticminstrel> I mean the accelerated rendering branch, sorry. 20180812 16:15:14< celticminstrel> (You still working on the OpenGL branch?) 20180812 16:15:32<+wesdiscordbot> Planning to return to it once the 1.14 branch is in good enough shape. 20180812 16:15:41< celticminstrel> Ah, okay. 20180812 16:16:24<+wesdiscordbot> And problem with macOS (thanks apple) is OpenGL... I was thinking about it whole day 😄 I want to keep macOS target but it will be hard if apple suddenly remove OpenGL legacy support 20180812 16:16:46<+wesdiscordbot> Maybe I am facing that problem already with blank screen issue -.- 20180812 16:17:32<+wesdiscordbot> nah, they won't remove it for a while 20180812 16:17:32< celticminstrel> Like I've already said, deprecated means that it should still work for the time being. It means we need to find a GL-over-Metal solution (even if that ends up being GL-over-Vulkan-over-Metal), but it's not exactly urgent. 20180812 16:17:40<+wesdiscordbot> and it's master that uses hw acceleration, not 1.14 20180812 16:17:49<+wesdiscordbot> The bar is already high enough with OpenGL support. I don't want to make the task even harder by adding OpenGL ES, Metal, ANGLE or something into the mix. 20180812 16:18:10<+wesdiscordbot> So it's indeed possible that we end up at least temporarily dropping macOS support. 20180812 16:18:46<+wesdiscordbot> Great 😄 20180812 16:18:55<+wesdiscordbot> I get it ofc 20180812 16:19:00< celticminstrel> But that's only if we aren't able to find a GL-over-Metal solution before they drop it, which to me seems unlikely. 20180812 16:20:11<+wesdiscordbot> Unfortunately Apple isn't Microsoft. They care much less about backwards compatibility. 20180812 16:20:35<+wesdiscordbot> we shall see 20180812 16:20:40<+wesdiscordbot> I wouldn't be surprised if they dropped OpenGL support soon. 20180812 16:20:56<+wesdiscordbot> me to. sadly 20180812 16:21:24<+wesdiscordbot> New small features and bug fixes for 1.14 are fine 20180812 16:21:53<+wesdiscordbot> But in long term (1.15) we should make decision what is more important GUI or OpenGL? 20180812 16:22:11<+wesdiscordbot> OpenGL. It's much less work. 20180812 16:22:18< celticminstrel> Jyrki, is there some reason why MoltenVK combined with a GL-over-Vulkan library wouldn't work? I'd imagine it to be somewhat inefficient, but... 20180812 16:22:19<+wesdiscordbot> It's a no-brainer. 20180812 16:22:43< celticminstrel> A direct GL-over-Metal solution (like MoltenGL but free) is definitely better. 20180812 16:23:24<+wesdiscordbot> celticminstrel: I believe ANGLE would be the best GL-over-Vulkan library. At this point their Vulkan backend isn't even complete. 20180812 16:23:53<+wesdiscordbot> On top of that, they have only started working on supporting OpenGL ES 2.0 on Vulkan. 20180812 16:24:33<+wesdiscordbot> I believe switching from desktop OpenGL to OpenGL ES wouldn't be that much work, but it's still additional work on top of the already huge task. 20180812 16:24:37< celticminstrel> MoltenVK apparently isn't quite a complete implementation of Vulkan, either. 20180812 16:25:07<+wesdiscordbot> I believe ANGLE only needs to subset that MoltenVK supports, though. 20180812 16:25:46<+wesdiscordbot> Google uses ANGLE for Chrome's rendering, and they want Chrome to remain available on macOS - they don't want to lose that market share. 20180812 16:28:14<+wesdiscordbot> wxWidgets? why not? 20180812 16:28:33<+wesdiscordbot> Because rewrite. 20180812 16:28:43<+wesdiscordbot> Any option that involves rewriting the UI is bad. 20180812 16:29:32<+wesdiscordbot> you mean User Interface I expect 20180812 16:29:37<+wesdiscordbot> Yes. 20180812 16:32:30<+wesdiscordbot> Isn't every GUI framework involes UI rewrite? 20180812 16:32:33< celticminstrel> I really wish we had a map diffing tool. 20180812 16:32:47< celticminstrel> I'm currently using a text diff tool to diff maps. 20180812 16:32:54< celticminstrel> It works but it's not very good. 20180812 16:33:02<+wesdiscordbot> @hrubymar10 Yes. That's the problem. 20180812 16:33:08<+wesdiscordbot> 😄 20180812 16:59:11<+wesdiscordbot> I already posted it here once. Can I try implement automatic update checking? It will be useful for SF versions. It will works like that if is internet connection available on start, then it will wget text file on f.w.o, parse it using src/game_version.hpp and compare it with actual runtime version. If there will be update available then it will show message here: 20180812 16:59:12<+wesdiscordbot> https://cdn.discordapp.com/attachments/259976436490829825/478246081801486348/unknown.png 20180812 16:59:47<+wesdiscordbot> Is it useful? Should I do that? 20180812 17:06:16-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 17:06:17< travis-ci> gfgtdf/wesnoth#1216 (1.14 - 836642f : gfgtdf): The build failed. 20180812 17:06:17< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415147818 20180812 17:06:17-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 17:08:33-!- vn971 [~vasya@94.158.103.15] has joined #wesnoth-dev 20180812 17:24:13-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180812 17:24:19-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180812 17:31:21<+wesdiscordbot> As long as that doesn't happen for the Linux versions… 20180812 17:32:30-!- valdar [~atarocch@37.177.11.203] has joined #wesnoth-dev 20180812 17:45:49<+wesdiscordbot> or steam 20180812 18:04:16-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 18:04:17< travis-ci> gfgtdf/wesnoth#1217 (1.14 - d4fbce3 : gfgtdf): The build was fixed. 20180812 18:04:17< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415155533 20180812 18:04:17-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 18:06:09< celticminstrel> @hrubymar10 - I think it would be useful under the conditions sevu and Pentarctagon mentioned - ie, that it doesn't happen if you have some other method of updating. So, only do it on Mac and Windows, and don't do it if it's a Steam install. (Not sure how you can detect the latter though?) 20180812 18:06:27< celticminstrel> (Worst-case, you could distribute a different build on Steam.) 20180812 18:11:52-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 18:11:53< travis-ci> wesnoth/wesnoth#19156 (illuminating_campfire - 1f8c7d3 : Celtic Minstrel): The build passed. 20180812 18:11:53< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/415133966 20180812 18:11:53-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 18:21:35<+wesdiscordbot> On macOS, I can make steam version identify itself by running specific command line argument. Or even better, I'll try add --no-update-check argument which will be used on steam builds by default 20180812 18:21:50<+wesdiscordbot> Linux will have this flag also set by default 20180812 18:22:44<+wesdiscordbot> and idk if windows version on steam can send command line arguments, but if so, then it will solve my problem, isn't it? 20180812 18:32:10<+wesdiscordbot> maybe you could add --no-update-check to the .desktop file, and maybe sth similat is possible with the windows/mac versions which sould not use it? 20180812 18:32:43<+wesdiscordbot> (Oh, and new command line options should get a manpage entry too, and somehow listed with --help) 20180812 18:33:49<+wesdiscordbot> For the former, the file is doc/man/wesnoth.d, for the later I don't know 20180812 18:34:03<+wesdiscordbot> *wesnoth.6 20180812 18:43:41-!- louis94 [~~louis94@109.49-136-217.adsl-dyn.isp.belgacom.be] has joined #wesnoth-dev 20180812 18:48:45-!- irker724 [~irker@wesnoth/umc-dev/bot/shikadibot] has quit [Quit: transmission timeout] 20180812 19:12:46-!- irker418 [~irker@wesnoth/umc-dev/bot/shikadibot] has joined #wesnoth-dev 20180812 19:12:46< irker418> wesnoth/wesnoth:1.14 Celtic Minstrel 543d35447f Merge pull request #3451 from wesnoth/Fi AppVeyor: vs2017/Release Failed 20180812 19:12:46< irker418> Details: https://ci.appveyor.com/project/wesnoth/wesnoth-605wt/build/Wesnoth-VS2017-1.14-4229 20180812 19:21:19-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev 20180812 19:33:58-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180812 19:34:04-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180812 19:49:07-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 19:49:08< travis-ci> gfgtdf/wesnoth#1218 (1.14 - ed603df : gfgtdf): The build was broken. 20180812 19:49:08< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415184902 20180812 19:49:08-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 20:01:18< irker418> wesnoth: Martin Hrubý (hrubymar10) wesnoth:Fix_Xcode_10_building_master 965aeb36adaf / / (77 files in 27 dirs): Fix building with Xcode 10 - #3458 https://github.com/wesnoth/wesnoth/commit/965aeb36adafe264a33127fe400a5ab274712b33 20180812 20:25:41< irker418> wesnoth: Martin Hrubý (hrubymar10) wesnoth:Fix_Xcode_10_building_master a421ab51a9b1 / src/video.cpp: Include SDL_WINDOW_ALLOW_HIGHDPI on apple platform https://github.com/wesnoth/wesnoth/commit/a421ab51a9b19f958ed22eb7ac02af8ebe4617ac 20180812 20:26:23< irker418> wesnoth: Martin Hrubý (hrubymar10) wesnoth:Fix_Xcode_10_building_master 62b56eb9b547 / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Force use tabs in Xcode project https://github.com/wesnoth/wesnoth/commit/62b56eb9b5472f280870463999f50f3a248d2c22 20180812 20:27:13< irker418> wesnoth: Martin Hrubý (hrubymar10) wesnoth:master b6507cbbf2d0 / src/macosx/SDLMain.mm: We no longer support OS X Lion - remove legacy code https://github.com/wesnoth/wesnoth/commit/b6507cbbf2d01ae84130e5db2a08796f98f1235c 20180812 20:29:10-!- valdar [~atarocch@37.177.11.203] has quit [Ping timeout: 272 seconds] 20180812 20:37:40-!- celmin [~celticmin@unaffiliated/celticminstrel] has joined #wesnoth-dev 20180812 20:44:49< irker418> wesnoth: Martin Hrubý (hrubymar10) wesnoth:1.14 fa144aab9cb4 / src/macosx/SDLMain.mm: We no longer support OS X Lion - remove legacy code https://github.com/wesnoth/wesnoth/commit/fa144aab9cb41118a276b70287d2cd52226162c7 20180812 20:49:38-!- travis-ci [~travis-ci@ec2-54-242-206-169.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 20:49:39< travis-ci> wesnoth/wesnoth#19159 (Fix_Xcode_10_building_master - 965aeb3 : Martin Hrubý (hrubymar10)): The build failed. 20180812 20:49:39< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/415192646 20180812 20:49:39-!- travis-ci [~travis-ci@ec2-54-242-206-169.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 20:50:52-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 20:50:53< travis-ci> gfgtdf/wesnoth#1219 (1.14 - a0639a0 : gfgtdf): The build is still failing. 20180812 20:50:53< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415197577 20180812 20:50:53-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 21:00:57< irker418> wesnoth: Martin Hrubý (hrubymar10) wesnoth:Fix_Xcode_10_building_master 0271b2f59ba2 / src/tests/ (4 files in 2 dirs): Fixed typo https://github.com/wesnoth/wesnoth/commit/0271b2f59ba21460f41c7c6abefcfbae5e510b5c 20180812 21:12:18< irker418> wesnoth: Elvish_Hunter wesnoth:master 9b5c0dda5c81 / data/tools/wmllint: wmllint: do not warn about unknown IDs in [aspect] https://github.com/wesnoth/wesnoth/commit/9b5c0dda5c819e5ac6a40a78241a36e022a84f75 20180812 21:12:20< irker418> wesnoth: Elvish_Hunter wesnoth:master 2dda3280c1b3 / data/tools/wmllint: wmllint: do not warn about unknown IDs in [remove_event] https://github.com/wesnoth/wesnoth/commit/2dda3280c1b3098aba146e2775e58577930d3e60 20180812 21:12:22< irker418> wesnoth: Elvish_Hunter wesnoth:master 2d045f2a6a47 / data/tools/wmllint: wmllint: do not warn about unknown IDs in [tunnel], except inside of [filter] https://github.com/wesnoth/wesnoth/commit/2d045f2a6a47c58b5bb1d77d3c836119e5978708 20180812 21:12:24< irker418> wesnoth: Elvish_Hunter wesnoth:master 6e696661fdc8 / data/tools/wmllint: wmllint: don't even try to convert _info.cfg https://github.com/wesnoth/wesnoth/commit/6e696661fdc8710a4310d3a2fd7228e3f135e5ed 20180812 21:12:26< irker418> wesnoth: Elvish_Hunter wesnoth:master b1ec52cab84a / data/tools/wmllint: wmllint: added [lua] to the allowed children of [if] https://github.com/wesnoth/wesnoth/commit/b1ec52cab84a5a18bd1518449fec374f4be3492b 20180812 21:12:44< celmin> Elvish_Hunter: Is there a version of the recognize directive that only applies to a single line? 20180812 21:12:55<+wesdiscordbot> And now, another round of cherry-picking... 20180812 21:13:10< celmin> Also I think there might be something else in the AI thing similar to 9b5c0dda... 20180812 21:13:16< celmin> Lemme check quickly... 20180812 21:14:03-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 21:14:04< travis-ci> gfgtdf/wesnoth#1220 (1.14 - 1b6b39f : gfgtdf): The build is still failing. 20180812 21:14:04< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415201031 20180812 21:14:04-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 21:14:08<+wesdiscordbot> celmin: no, there isn't. However, # wmllint: ignore should work for that purpose on a single line. 20180812 21:15:24< celmin> [candidate_action], [stage], [engine] IDs should also be ignored. [ai] can have an ID too though it normally won't (that's only for AIs chooseable in MP). I'm not sure if [goal] can have an ID. 20180812 21:15:54< irker418> wesnoth/wesnoth:1.14 Victor Sergienko 60d238d7e5 A cleaner fix for "iOS interfering with AppVeyor: All builds passed 20180812 21:16:21<+wesdiscordbot> [stage] and [goal] are already ignored. This leaves just [candidate_action] and [engine] to be added. 20180812 21:16:32< celmin> Ah okay. 20180812 21:19:04< celmin> (Is there a place where wmllint directives are documented?) 20180812 21:19:13< irker418> wesnoth: Elvish_Hunter wesnoth:1.14 03510c36c1a1 / data/tools/wmllint: wmllint: do not warn about unknown IDs in [aspect] https://github.com/wesnoth/wesnoth/commit/03510c36c1a100241304fee3863b0536a0f68f90 20180812 21:19:15< irker418> wesnoth: Elvish_Hunter wesnoth:1.14 6164980ebc79 / data/tools/wmllint: wmllint: do not warn about unknown IDs in [remove_event] https://github.com/wesnoth/wesnoth/commit/6164980ebc79cd581b648505aa8a4beadfd4a8dc 20180812 21:19:17< irker418> wesnoth: Elvish_Hunter wesnoth:1.14 0bcb0e00d34f / data/tools/wmllint: wmllint: do not warn about unknown IDs in [tunnel], except inside of [filter] https://github.com/wesnoth/wesnoth/commit/0bcb0e00d34f6f9140c4e97c0311a4b47ec00488 20180812 21:19:19< irker418> wesnoth: Elvish_Hunter wesnoth:1.14 d128e605a725 / data/tools/wmllint: wmllint: don't even try to convert _info.cfg https://github.com/wesnoth/wesnoth/commit/d128e605a725240106cafe8f10b939e922e7f514 20180812 21:19:21< irker418> wesnoth: Elvish_Hunter wesnoth:1.14 eb65a44de049 / data/tools/wmllint: wmllint: added [lua] to the allowed children of [if] https://github.com/wesnoth/wesnoth/commit/eb65a44de0494dad4a4c4f02317043e2898f01cb 20180812 21:19:55<+wesdiscordbot> Yes, at the start of wmllint's file there are a lot of commented lines. All the directives are documented here. 20180812 21:20:07<+wesdiscordbot> celmin, mattscm, you mentioned recenty sth which should not be converted by wmllint 20180812 21:20:19<+wesdiscordbot> is that chancged by the above commits? 20180812 21:21:26< celmin> I think most of it is? 20180812 21:21:38<+wesdiscordbot> Yes, I've seen about [leader] causing errors. I'm already testing a fix for it. 20180812 21:21:43< celmin> I'll have to rerun wmllint on my campaign with those changes and see how it goes. 20180812 21:22:08< celmin> [leader] is just a shorthand for [unit]canrecruit=yes (but only works in [side] IIUC, not in events). 20180812 21:22:21-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 240 seconds] 20180812 21:22:21< celmin> IOW it should be treated as [unit]. 20180812 21:23:46<+wesdiscordbot> The situation is a bit more complex than this. The "side without type=" check looks like it was introduced to avoid a crash in Wesnoth. 20180812 21:25:05<+wesdiscordbot> The simplest solution consists in removing this check. The more complex one consists in retrofitting a "children" list in the validate_on_pop() function, so wmllint can check if a [leader] tag has been used. 20180812 21:25:27< celmin> Ooh this reminds me, recruitment pattern can be specified in a new way too... 20180812 21:26:22< celmin> Pretty sure you mean "consists of" here. 20180812 21:27:17< celmin> I think the crash doesn't actually occur anymore? On the other hand it could be nice for the wmllint to continue to make sure a leader exists if you don't have no_leader=yes... 20180812 21:27:32< celmin> BTW "this" = reading the comments at the top of wmllint. 20180812 21:29:24< celmin> Okay so basically a recruitment pattern can be specified in several ways now: 20180812 21:29:25< celmin> - [ai]recruitment_pattern= 20180812 21:29:26< celmin> - [ai][recruitment_pattern]value= 20180812 21:29:40< celmin> - [ai][recruitment_instructions][pattern]type= 20180812 21:30:05< celmin> - [ai][recruitment_instructions][recruit]pattern,type=yes, 20180812 21:30:10< celmin> I think that's all of them? 20180812 21:30:52< celmin> I guess it's not strictly necessary to cover them all (the first will probably continue to be the most common). 20180812 21:31:06< celmin> But if you want to... 20180812 21:31:48<+wesdiscordbot> Yeah, it looks like the other ways aren't even documented in the wiki. 20180812 21:32:15< celmin> The last one definitely is, on the AI_Recruitment page. 20180812 21:32:48< celmin> I think the others are sort of documented, but not explicitly? Replacing some_aspect= with [some_aspect]value= is a general thing that works, and I'm pretty sure it's documented somewhere. 20180812 21:33:04< celmin> And I think the second-to-last one is mentioned on the AI_Recruitment page but I can't remember. 20180812 21:33:26< celmin> Mind you, I'm not sure how discoverable the AI_Recruitment page is... 20180812 21:33:36< celmin> It should be linked from AIWML… I'll go make sure it is. 20180812 21:33:56< celmin> Ah, yes, it is indeed, good. 20180812 21:36:31<+wesdiscordbot> There's still the problem that these AI tags used as array indexes to set aspects confuse wmllint... I found some lines that seems to cause the errors, but I don't have a good solution yet. 20180812 21:36:38< celmin> (FTR the reason you might want to replace some_aspect= with [some_aspect]value= is because the tag can then also include conditions about turns and time of day, as well as an ID to be able to edit the aspect later with [modify_ai]; the conditions can also appear right at [ai] level but the ID could be a reason to do it, I guess. TBH it might be easiest to just ignore any IDs nested in an AI tag? Unless they're in a filter, which I think 20180812 21:36:39< celmin> is currently just in [avoid] and [goal][criteria].) 20180812 21:39:59<+wesdiscordbot> No, I was talking about lines like path=stage[main_loop].candidate_action[healing]. wmllint here finds two opening tags when it shouldn't find anything. 20180812 21:40:11< celmin> So wmllint traverses files in alphabetical order and remembers stuff from previous files when it gets to the later files? 20180812 21:40:22< celmin> Yeah, I know what you were talking about, that comment predated yours. 20180812 21:40:32< celmin> I mean I started typing it before you said that. 20180812 21:40:36< irker418> wesnoth/wesnoth:1.14 Victor Sergienko 10e96936f1 A cleaner fix for "iOS interfering with AppVeyor: All builds passed 20180812 21:40:58<+wesdiscordbot> No, it doesn't remember the other files. 20180812 21:40:58< celmin> Hmm, maybe don't count it as a tag if there's an = earlier in the line? 20180812 21:41:02< celmin> Ah okay. 20180812 21:41:15< celmin> But the documentation of who and whofield makes it sound like it does? 20180812 21:41:32< celmin> And the very existence of unwho. 20180812 21:41:53<+wesdiscordbot> It remembers these directives, sure, but not the actual content of the files. 20180812 21:42:06< celmin> Okay. 20180812 21:44:46< celmin> What's this stuff about newlines in messages? 20180812 21:47:14-!- travis-ci [~travis-ci@ec2-54-242-206-169.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 21:47:15< travis-ci> gfgtdf/wesnoth#1222 (1.14 - 382bca7 : gfgtdf): The build was canceled. 20180812 21:47:15< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415211879 20180812 21:47:15-!- travis-ci [~travis-ci@ec2-54-242-206-169.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 21:56:56< Elvish_Hunter> celmin: you've got a forum PM. 20180812 21:57:25-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 21:57:26< travis-ci> wesnoth/wesnoth#19165 (master - b6507cb : Martin Hrubý (hrubymar10)): The build has errored. 20180812 21:57:26< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/415198714 20180812 21:57:26-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 22:08:29-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection] 20180812 22:18:22-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 22:18:23< travis-ci> gfgtdf/wesnoth#1224 (1.14 - f62148e : gfgtdf): The build was canceled. 20180812 22:18:23< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415218909 20180812 22:18:23-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 22:20:54-!- sevu [~sevu@p54854237.dip0.t-ipconnect.de] has joined #wesnoth-dev 20180812 22:29:33-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180812 22:29:39-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180812 22:33:05-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 22:33:06< travis-ci> wesnoth/wesnoth#19167 (Fix_Xcode_10_building_master - 0271b2f : Martin Hrubý (hrubymar10)): The build passed. 20180812 22:33:06< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/415205995 20180812 22:33:06-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 22:41:00<+wesdiscordbot> hmm, a replay file has after 7 turns 30k lines… 20180812 22:41:24-!- sevu [~sevu@p54854237.dip0.t-ipconnect.de] has quit [Remote host closed the connection] 20180812 23:06:06-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 23:06:07< travis-ci> gfgtdf/wesnoth#1225 (1.14 - e0c47d4 : gfgtdf): The build has errored. 20180812 23:06:07< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415220710 20180812 23:06:07-!- travis-ci [~travis-ci@ec2-54-167-243-48.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 23:06:26-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev 20180812 23:19:33-!- louis94 [~~louis94@109.49-136-217.adsl-dyn.isp.belgacom.be] has quit [Remote host closed the connection] 20180812 23:20:02-!- louis94 [~~louis94@109.49-136-217.adsl-dyn.isp.belgacom.be] has joined #wesnoth-dev 20180812 23:30:35-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection] 20180812 23:31:09-!- gfgtdf [~chatzilla@x2f6124b.dyn.telefonica.de] has joined #wesnoth-dev 20180812 23:31:18< gfgtdf> yes the replay is often the lagest part of a savefile 20180812 23:31:25< gfgtdf> are you investigating replays fora speficic reason ? 20180812 23:31:39-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20180812 23:35:45<+wesdiscordbot> I wanted to [allow_undo] an event which has messages with [option]s. It dodn't work, in the replay file I found this: https://bpaste.net/show/7604a6e48ce6 20180812 23:36:17<+wesdiscordbot> That lead me to the conclusion that it's generally not possible to undo [option]s as that is directly send to the other players 20180812 23:36:34-!- travis-ci [~travis-ci@ec2-54-160-231-174.compute-1.amazonaws.com] has joined #wesnoth-dev 20180812 23:36:35< travis-ci> gfgtdf/wesnoth#1226 (1.14 - 9c2482d : gfgtdf): The build passed. 20180812 23:36:35< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/415232264 20180812 23:36:35-!- travis-ci [~travis-ci@ec2-54-160-231-174.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180812 23:36:46< irker418> wesnoth/wesnoth:1.14 sigurdfdragon 899609af46 Help: Add Saurian race description (fix AppVeyor: All builds passed 20180812 23:40:21< gfgtdf> sevu: hmm no it shodul be possibel to undo actions with options, (unless they are nonundoable for another reason) 20180812 23:40:29< gfgtdf> can you shwo me your code ? 20180812 23:40:59< gfgtdf> show* 20180812 23:42:35< gfgtdf> @sevu: if you [option][command] does not change the gamestate then usning [unsynced] actionwml is also an option. 20180812 23:42:51<+wesdiscordbot> if a 2k loop doesn't scare you (it's about the code before and after the loop) 20180812 23:43:27< gfgtdf> i'll try not to be scared. 20180812 23:43:34<+wesdiscordbot> it's a shop system, which consist of one giant [while], in which a message with some options is shown 20180812 23:43:59<+wesdiscordbot> might need a few minuted to get the code back und somewhere up 20180812 23:47:23< gfgtdf> @sevu i tested with this sipme event https://pastebin.com/FLMvhgQM and the message does nto make undoing impossible 20180812 23:47:27< gfgtdf> simple 20180812 23:48:01< gfgtdf> i think you might have something else, or anothe event that makes undoing impossible. 20180812 23:53:54<+wesdiscordbot> The code is this: https://github.com/sevu/Col/blob/main/macros/upgrade.cfg#L77-L80 + the [if] clause at the very end of the file 20180812 23:54:22-!- louis94 [~~louis94@109.49-136-217.adsl-dyn.isp.belgacom.be] has quit [Ping timeout: 272 seconds] 20180812 23:55:29<+wesdiscordbot> there is one [message]tag which contains one [option] which let's you exit the loop. and other [options]s with a second layer of messages & options. 20180812 23:56:09<+wesdiscordbot> Assuming you go directly with the one to leave the menu (the first), is there sth. blocking? 20180812 23:56:34<+wesdiscordbot> (I should test the code again as I just readded it, in case I made sth different this time) 20180812 23:57:56-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection] 20180812 23:59:08< gfgtdf> hmm no i dont see anythign that'd block it in that case. --- Log closed Mon Aug 13 00:00:00 2018