--- Log opened Mon May 21 00:00:06 2018 --- Day changed Mon May 21 2018 20180521 00:00:06< Ravana_> during last 5 years I have watched about 10 movies 20180521 00:00:22< celticminstrel> Maybe about the same if you count DVDs or blurays. 20180521 00:05:39< zookeeper> i bet _some_ of those superhero movies are at their core good movies, but it's hard to have any inclination to watch them when the dressing is so off-putting what with all the silly costumes and superpowers. 20180521 00:07:19< celticminstrel> XD 20180521 00:14:34<+discordbot> .................. 20180521 00:14:40<+discordbot> celmin: yes 20180521 00:16:16< celticminstrel> Stop with the ellipses, it's insulting. 20180521 00:16:56<+discordbot> [perplexed silence] 20180521 00:18:52<+discordbot> question: which starting area is better? 20180521 00:19:11<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/447916227923542027/unknown.png 20180521 00:19:25<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/447916289894383617/unknown.png 20180521 00:21:31< celticminstrel> That's obvious. 20180521 00:22:18<+discordbot> Okay, someone did a bad. 20180521 00:22:28< celticminstrel> Only possible problem I see with the second one is, what if all those imprisoned units are actually important. 20180521 00:22:35<+discordbot> wesnoth.wml_actions.color_adjust { red = 500, green = -500, blue = -500 } does not make the terrains pure red in a cave map. 20180521 00:22:51<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/447917151026937866/unknown.png 20180521 00:22:51< zookeeper> @Yumi, partially depends on what sort of place it is. if it's a fancy elaborate well-maintained hall, then the old one, and if it's more of a gloomy and mouldy dump then the latter. 20180521 00:23:28< celticminstrel> zookeeper: I'd say that if it's a fancy elaborate well-maintained hall then maybe you shouldn't use cave wall. 20180521 00:24:06< zookeeper> well, okay, maybe not elaborate 20180521 00:24:37<+discordbot> Oh for god's sake. 20180521 00:24:43< celticminstrel> The first one is undeniably boring. The second is interesting, which is an improvement; however I can't say whether it's suitable. 20180521 00:24:45<+discordbot> cpp void display::adjust_color_overlay(int r, int g, int b) { color_adjust_ = tod_color(r, g, b); update_tod(); } 20180521 00:25:04<+discordbot> The tod_color ctor clamps those values to the [-255, 255] range. 20180521 00:25:24< celticminstrel> What's wrong with that? Wouldn't larger absolute values be meaningless? 20180521 00:25:43<+discordbot> They are not when you have [color_adjust] interacting with ToD. 20180521 00:25:57< celticminstrel> Oh I think I see what you mean... 20180521 00:26:14<+discordbot> For example, if the ToD shift is 100,100,100, you can't use -255,-255,-255 to make the screen pure black anymore. 20180521 00:26:30<+discordbot> (The latter with [color_adjust], the former with [time].) 20180521 00:26:44< celticminstrel> Basically the two are added, huh. 20180521 00:26:59<+discordbot> @Vultraz Talking of undocumented expectations being broken again. 20180521 00:27:13< zookeeper> anyway, my point is that aside from gameplay, half of map design is about how aesthetically pleasing it looks and the other half about what sort of things it communicates to the player. 20180521 00:27:26< celticminstrel> IMO it makes more sense for -255 to just bring that channel to 0 though... 20180521 00:27:34<+discordbot> Not really. 20180521 00:27:49<+discordbot> 100-255 = -155. 20180521 00:27:56< celticminstrel> Like, instead of combining the two adjusts into one adjust that is then applied, why not separately apply each adjust? 20180521 00:28:03<+discordbot> Celmin: actually the ellipses were for zookeeper this time 20180521 00:28:19< celticminstrel> @Vultraz Still insulting though. 20180521 00:28:21<+discordbot> Either you ignore basic arithmetics, or explicitly add a way for [color_adjust] to ignore the ToD shift. 20180521 00:28:41<+discordbot> zookeeper: it's only stated in the campaign that there are "underlevels" but not what they exactly are 20180521 00:28:54<+discordbot> Anyway, I guess this is not too much of a deal breaker, but I'd still like to see it fixed because I used this stuff quite extensively in my campaigns, so I'm going to file a bug. 20180521 00:28:58< celticminstrel> So first it applies the ToD colour, so eg 200+100 brings it to pure white (255 because of clamping), and then it applies the color_adjust which brings it to pure black. 20180521 00:29:07<+discordbot> And mark it as a regression. 20180521 00:29:11<+discordbot> since they seem to be somewhat of a secret, hidden place, I was thinking that it would be less well-maintained and more hideout-like 20180521 00:29:26< zookeeper> @Yumi, sure, i'm not saying the new one isn't fitting. i've never actually even played that scenario myself :p 20180521 00:29:37<+discordbot> that's fine, just giving my reasoning 20180521 00:38:31<+discordbot> https://github.com/wesnoth/wesnoth/issues/3144 for the IRC people. 20180521 00:51:45<+discordbot> celmin: you're the author of the problematic commit: https://github.com/wesnoth/wesnoth/commit/c77faf2223b535badbe56fdd30bc404e283e7c29 20180521 00:51:50<+discordbot> what suggest you? 20180521 00:52:40< celticminstrel> I guess for 1.14, we can clamp to -511,511 instead? 20180521 00:52:58<+discordbot> Or not clamp it at all. 20180521 00:53:01< celticminstrel> And for 1.15, add an override_tod key in [color_adjust]. 20180521 00:53:17< celticminstrel> Well yeah, but more than 511 won't do anything extra anyway. 20180521 00:53:45<+discordbot> blah... 20180521 00:54:36<+discordbot> Shouldn't it be [-510, 510] then? 20180521 00:54:57<+discordbot> 255*2 = 510 20180521 00:55:10< celticminstrel> Well, 255 is 2^8-1. 20180521 00:55:22< celticminstrel> And 511 is 2^9-1. 20180521 00:55:39< celticminstrel> But I suppose that extra 1 might not make a difference, indeed. 20180521 00:55:56<+discordbot> The point is that if 255 - x = -255, x = 510. 20180521 00:56:04< celticminstrel> Yeah. 20180521 00:56:08< celticminstrel> Sure. 20180521 00:56:49<+discordbot> @shadowm could you confirm the wider clamp range fixes the issue? 20180521 00:57:26<+discordbot> I had just finished making a build without clamp. :p 20180521 00:57:33<+discordbot> Building again. 20180521 01:02:48<+discordbot> Uhh. 20180521 01:04:17<+discordbot> Suddenly a custom team colour isn't working anymore. 20180521 01:04:25< celticminstrel> ... 20180521 01:04:32-!- gfgtdf_ [~chatzilla@x4e363320.dyn.telefonica.de] has joined #wesnoth-dev 20180521 01:04:59<+discordbot> I doubt it has anything to do with the colour shift clamping though. 20180521 01:05:42< celticminstrel> Indeed. 20180521 01:06:28<+discordbot> But yeah, the colour effect is fixed... 20180521 01:06:29-!- gfgtdf [~chatzilla@x4e363280.dyn.telefonica.de] has quit [Ping timeout: 260 seconds] 20180521 01:06:42-!- gfgtdf_ is now known as gfgtdf 20180521 01:07:44<+discordbot> Quickly checking to see if it's actually related just in case, otherwise it must be a recent change. 20180521 01:08:19<+discordbot> Yeah, completely unrelated. 20180521 01:08:36<+discordbot> I'll commit the fix for #3144. 20180521 01:11:22<+discordbot> celmin: do you think you could add changelog entries for significant campaign changes for 1.14.12? 20180521 01:11:24<+discordbot> 2* 20180521 01:11:39< celticminstrel> No? 20180521 01:12:06<+discordbot> 😦 20180521 01:12:26<+discordbot> alright... I'll do it myself 20180521 01:12:49< irker078> wesnoth: Iris Morelle wesnoth:master ae127ea73a27 / changelog.md src/time_of_day.hpp: Clamp tod_color values in the [-510, 510] range instead of [-255, 255] https://github.com/wesnoth/wesnoth/commit/ae127ea73a27ac1df37b065169810833774452de 20180521 01:13:34< irker078> wesnoth: Iris Morelle wesnoth:1.14 711e9a287d0a / changelog.md src/time_of_day.hpp: Clamp tod_color values in the [-510, 510] range instead of [-255, 255] https://github.com/wesnoth/wesnoth/commit/711e9a287d0a578de4daa4820866af6ff4321315 20180521 01:13:38<+discordbot> i was hoping you could do your thing where you look through the commit log for important things 20180521 01:14:18< celticminstrel> I mean, you could do that too you know. And if you only want campaigns, "git log data/campaigns" or some such should work. 20180521 01:14:32< celticminstrel> I guess ideally you'd need to also specify the revision range too... 20180521 01:14:44< celticminstrel> So as not to accidentally add duplicates from 1.14.1 or earlier. 20180521 01:14:47<+discordbot> i can do it 20180521 01:15:03<+discordbot> Okaaay there are no relevant changes since that build that wasn't showing this issue with custom team colours. 20180521 01:15:41<+discordbot> Or so goes the theory. 20180521 01:16:07<+discordbot> Uhh. 20180521 01:16:17<+discordbot> I reverted to that build and the team colour isn't working either. 20180521 01:16:23<+discordbot> @shadowm could you test your remaining [change_theme] issue by adding rebuild_all(); right before redraw_everything() 20180521 01:16:28<+discordbot> in display::set_theme 20180521 01:16:54<+discordbot> I swear I saw a yellow unit ellipse when testing 30 minutes ago. 20180521 01:17:05<+discordbot> i just saw a yellow ellipse 20180521 01:17:35<+discordbot> the rebuild_all call seems to fix the issue for me, but I'd like to confirm since you have a bigger screen 20180521 01:17:51<+discordbot> I'm more concerned with the team colour issue right now. 😦 20180521 01:18:49<+discordbot> Is there any way to get team colour definitions from Lua? 20180521 01:18:56<+discordbot> no 20180521 01:20:20<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/447931616598425611/unknown.png 20180521 01:20:24<+discordbot> It must be being read. 20180521 01:20:54<+discordbot> tell me what exactly happens 20180521 01:21:03<+discordbot> The yellow color range doesn't work. 20180521 01:21:11<+discordbot> in what context 20180521 01:21:17<+discordbot> tell me exactly what happens and where 20180521 01:21:17<+discordbot> AtS E3S8A. 20180521 01:21:37<+discordbot> Instead of yellow you get pure black colours (which makes the ellipses pretty much invisible). 20180521 01:21:59<+discordbot> Wait, it works in E3S0. 20180521 01:22:14<+discordbot> It works if I reload after that. 20180521 01:22:17<+discordbot> What is going on??? 20180521 01:22:30<+discordbot> is this [side] color=yellow? 20180521 01:22:35<+discordbot> The only difference is that I loaded E3S0 by launching the campaign. 20180521 01:22:43<+discordbot> And E3S8A by loading a SOS save. 20180521 01:23:10<+discordbot> Is this like the music bug, where Wesnoth behaves differently when launching a campaign vs. loading a save? 20180521 01:23:36<+discordbot> very possibly 20180521 01:23:38<+discordbot> Also vultraz that question is so dumb I feel like I'm talking to low-tier tech support right now. 20180521 01:23:46<+discordbot> You know I'd know if it weren't yellow. 20180521 01:24:02<+discordbot> no 20180521 01:24:13<+discordbot> I mean, was this color set in a side definition 20180521 01:24:17<+discordbot> or by some other method 20180521 01:24:20<+discordbot> ie, objects 20180521 01:24:26<+discordbot> It's set in [side] in both cases. 20180521 01:24:33<+discordbot> Seriously. 20180521 01:24:42<+discordbot> alright, that helps me.. 20180521 01:24:49<+discordbot> Launching the campaign from the Campaigns menu is the only way to get the color_range registered it seems. 20180521 01:25:24<+discordbot> Actually, wait. 20180521 01:25:25< celticminstrel> So a call to add_color_info() is missing somewhere, I guess? 20180521 01:25:35< celticminstrel> (I think that's what the funciton was called.) 20180521 01:25:36<+discordbot> I just launched it again from the Campaigns menu and it did not register the color_range. 20180521 01:25:44< celticminstrel> ^function 20180521 01:25:46<+discordbot> On a new session. 20180521 01:26:30<+discordbot> agh. this is difficult since master color handling has already diverged... 20180521 01:26:33<+discordbot> It then works if I use F5 and relaunch. 20180521 01:26:36<+discordbot> (slightly) 20180521 01:26:48<+discordbot> Vultraz please tell me you have a 1.14 worktree. 20180521 01:26:53<+discordbot> I do 20180521 01:27:29<+discordbot> I think there's nondeterministic behaviour at work here. 20180521 01:28:24<+discordbot> Ok, I need you to do some debugging. Can you check two things: * Whether the default return statement is hit in team::get_side_color_range * Whether the out_of_range exception is caught in team::get_side_color_id 20180521 01:29:12<+discordbot> I need a debug build for that. 20180521 01:29:19<+discordbot> just add cerr logging 20180521 01:29:34<+discordbot> I only need to know if it's hit. 20180521 01:30:11<+discordbot> I'm expecting the answer to the first question to be yes and the second to be no 20180521 01:30:35<+discordbot> I'm building a debug build with UBsan just in case. 20180521 01:30:57<+discordbot> There's something very weird going on. 20180521 01:33:11<+discordbot> hmmmmmm 20180521 01:33:15<+discordbot> celmin may be right 20180521 01:37:32<+discordbot> I know it can be reliably reproduced by loading a SOS save in a fresh session. 20180521 01:37:55<+discordbot> I'm more unsure as to how to reliably reproduce getting the color_range correctly registered since it seems completely random right now. 20180521 01:38:22<+discordbot> well, it goes through game_config::add_color_info 20180521 01:40:06<+discordbot> wait... 20180521 01:40:55<+discordbot> Also, neither of the things you wanted me to check are hit when the color_range doesn't work. 20180521 01:41:08<+discordbot> ....this surprises me 20180521 01:41:25<+discordbot> but ok, it's possible that function is never called at all 20180521 01:41:27<+discordbot> (the first one) 20180521 01:41:39< celticminstrel> Why would it not be called at all? 20180521 01:42:37<+discordbot> UBsan doesn't report anything out of the ordinary. 20180521 01:42:45<+discordbot> ok 20180521 01:42:49<+discordbot> shot in the dark but 20180521 01:42:53<+discordbot> see what happens if you add 20180521 01:44:16<+discordbot> ah, wait, no, that's called 20180521 01:45:08<+discordbot> next thing you could try is printing out the ID of every color range hit in add_color_info 20180521 01:47:51<+discordbot> Apparently there's a gold CR in mainline now? 20180521 01:48:14<+discordbot> when did that get in there 20180521 01:48:26<+discordbot> add_color_info(): id = cyan add_color_info(): id = yellow add_color_info(): id = pink add_color_info(): id = red add_color_info(): id = lightred add_color_info(): id = darkred 20180521 01:48:27< celticminstrel> Color ranges are used for a bunch more things now. 20180521 01:48:37< celticminstrel> It might be one of the orb ranges, for example. 20180521 01:48:44<+discordbot> So add-on CRs are registered first, interesting. 20180521 01:48:54<+discordbot> And it's also interesting that's registered but it doesn't show up on the screen. 20180521 01:49:05< celticminstrel> Of course you can also use them for sides, if you want. 20180521 01:49:31<+discordbot> Not necessarily since that doesn't guarantee that the colour points are useful against magenta. 20180521 01:49:34<+discordbot> gold was added 5 years ago 20180521 01:49:37<+discordbot> https://github.com/wesnoth/wesnoth/commit/82b8d9ecdf5c0f7121703e34acfe8f95ea9ae489 20180521 01:50:49< celticminstrel> Why the heck was gold added in that commit? It's not related to the topic. 20180521 01:50:59<+discordbot> ¯_(ツ)_/¯ 20180521 01:51:03-!- travis-ci [~travis-ci@ec2-54-158-14-200.compute-1.amazonaws.com] has joined #wesnoth-dev 20180521 01:51:03< travis-ci> gfgtdf/wesnoth#1127 (1.14 - 222e731 : gfgtdf): The build was broken. 20180521 01:51:04< travis-ci> Build details : https://travis-ci.org/gfgtdf/wesnoth/builds/381482686 20180521 01:51:04-!- travis-ci [~travis-ci@ec2-54-158-14-200.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180521 01:51:07< celticminstrel> shadowm: What do you mean? 20180521 01:51:09<+discordbot> Don't over think commits from non-coders. 20180521 01:51:28< celticminstrel> "useful against magenta"? 20180521 01:51:32<+discordbot> It means that I can define a colour range that isn't very useful on magenta. 20180521 01:51:44<+discordbot> The magenta palette is a bit peculiar. 20180521 01:52:00< irker078> wesnoth/wesnoth:master Pentarctagon d02e16d712 Fixup: Vyncyn was already listed in the AppVeyor: All builds passed 20180521 01:52:03<+discordbot> You can see for example that blue doesn't have the same range (excuse the pun) as teal. 20180521 01:52:22<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/447939678805426188/magenta-palette-RC-magenta-9-teal.png 20180521 01:52:32<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/447939723026104320/magenta-palette-RC-magenta-2-blue.png 20180521 01:52:48<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/447939788658704404/magenta-palette-RC-magenta-8-white.png 20180521 01:53:28<+discordbot> Compare those three on a vertical column on your screen looking at it at a straight angle (unless it's an IPS panel) and hopefully you should see what I mean. 20180521 01:53:35< celticminstrel> Okay? 20180521 01:53:52<+discordbot> what's weird is yellow works 20180521 01:53:52<+discordbot> for me 20180521 01:53:55<+discordbot> in ats e1s9.1 20180521 01:53:59<+discordbot> loading the sos save 20180521 01:54:24<+discordbot> I have an idea. 20180521 01:54:35<+discordbot> Could the [change_theme] fix be causing this issue? 20180521 01:54:53<+discordbot> no 20180521 01:55:25<+discordbot> Okay, let me remove the tags from the scenario. 20180521 01:55:43<+discordbot> Okay, yeah, it's not that. 20180521 01:57:38<+discordbot> Well, idk then. 20180521 01:58:21<+discordbot> is it deliberate that the warning about GOD DAMN THIS O KEY AtS's port being experimental pops up in every scenario 20180521 01:58:28<+discordbot> Yes. 20180521 01:58:38<+discordbot> I can't exactly prevent that. 20180521 01:58:56<+discordbot> set global flag if shown once? 20180521 01:59:03<+discordbot> No? 20180521 01:59:11<+discordbot> k 20180521 01:59:57<+discordbot> Doesn't it take at least an hour or two per scenario if you're playing normally? 20180521 02:00:11<+discordbot> No. 20180521 02:00:22<+discordbot> Some "scenarios" take 1-3 minutes. 20180521 02:01:04<+discordbot> Okay, the only other special things AFAICT: the scenario uses a default-constructed ToD as its initial ToD, then the ToD schedule is replaced during prestart. 20180521 02:01:48< celticminstrel> Is that even related to colour ranges? 20180521 02:01:48<+discordbot> That's about the only thing I can think of that's relevant for this case. 20180521 02:01:52<+discordbot> wesnoth makes discord lag 20180521 02:01:53<+discordbot> agh 20180521 02:01:56<+discordbot> You tell me celticminstrel. 20180521 02:02:12< celticminstrel> Well, I don't know of any relation, but it's not the most familiar area of the code to me. 20180521 02:02:13<+discordbot> For all I know saying "poop" may case Wesnoth's display engine to engage ludicrous speed. 20180521 02:02:35<+discordbot> how is it that wesnoth can make my computer perform worse than when dota is running 20180521 02:03:11<+discordbot> Congratulations, you just realized something that a myriad of other people have realized since the launch on Steam! :p 20180521 02:03:27<+discordbot> this is unacceptable 20180521 02:04:31< celticminstrel> Pretty sure that's why certain things are being done in 1.15, so... 20180521 02:04:43< celticminstrel> You'll have to bear with it on 1.14. 20180521 02:04:55< pydsigner> Isn't 1.14 still mostly CPU-driven? 20180521 02:05:00<+discordbot> entirely 20180521 02:05:12<+discordbot> unless SDL decides to accelerate the framebuffer behind the scenes 20180521 02:05:18< pydsigner> Dota is using GPU (and probably more power overall) 20180521 02:05:31<+discordbot> Hardware rendering is only intended to reduce CPU usage, and Wesnoth is single-threaded. 20180521 02:05:40<+discordbot> I wonder if that could have something to do with it.. 20180521 02:05:54< pydsigner> SDL accelerating things behind the scenes doesn't sound very SDL-like 20180521 02:06:14<+discordbot> As fas as I can tell, if Wesnoth is causing lag, it can't really be caused by CPU usage. 20180521 02:06:31<+discordbot> "By default SDL tries to make a best guess whether to use acceleration or not on each platform." 20180521 02:06:39<+discordbot> https://wiki.libsdl.org/SDL_HINT_FRAMEBUFFER_ACCELERATION 20180521 02:06:45< pydsigner> Interesting 20180521 02:07:16< pydsigner> Certainly Wesnoth could cause CPU problems on machines with low core count or poor thermal envelopes 20180521 02:07:20< pydsigner> (laptops) 20180521 02:07:58<+discordbot> Still, I'd expect Dota to use even more total power and require more thermal throttling. 20180521 02:08:04<+discordbot> GPU power isn't free either. 20180521 02:08:33<+discordbot> The average user doesn't care as much about power usage as they do about UI responsiveness. 20180521 02:08:50<+discordbot> the performance i saw with that large map yesterday was beyond abysmal 20180521 02:09:01<+discordbot> it looked like 5-10 FPS at times 20180521 02:09:13<+discordbot> animations missing frames all over the place 20180521 02:09:20<+discordbot> lag when even mousing over units 20180521 02:09:31<+discordbot> Not surprising if Wesnoth is loading images from disk constantly. 20180521 02:09:46<+discordbot> Except it doesn't? 20180521 02:10:02<+discordbot> let us see 20180521 02:10:07<+discordbot> The main problem with that map is that memory usage kept increasing. 20180521 02:10:19<+discordbot> It loads them once from disk and then they stay in the cache until :refresh or F5 are used (or some other state condition I'm not aware of takes place). 20180521 02:10:26<+discordbot> Which is why TLU is a massive memory hog. 20180521 02:10:34< zookeeper> i didn't check out that map; it must have had a lot of water, right? 20180521 02:10:52<+discordbot> Constantly increasing memory usage suggests that the same images are being added to the cache over and over again. 20180521 02:10:53<+discordbot> not that much 20180521 02:11:39<+discordbot> That's assuming there isn't a memory leak somewhere. 20180521 02:13:02<+discordbot> Whether it's technically a memory leak or not is just semantics. 20180521 02:13:15<+discordbot> The only difference is whether the memory remains reachable. 20180521 02:13:20<+discordbot> I meant somewhere else and potentially not involving the images themselves. 20180521 02:13:47<+discordbot> I believe it would be much slower if images weren't involved. 20180521 02:14:05<+discordbot> Leaking a single object only costs a couple dozen bytes or so. 20180521 02:14:17<+discordbot> A single image can easily cost you kilobytes. 20180521 02:14:29<+discordbot> @Vultraz Can you think of anything else to inspect or should I move on to the change_theme fix you want me to check? 20180521 02:14:55<+discordbot> I checked image::set_team_colors() FTR, it seems good. 20180521 02:15:02<+discordbot> scrolling around that huge map added SO MUCH to the cache 20180521 02:15:14<+discordbot> it's a lot less on small maps, obviously 20180521 02:15:56<+discordbot> Where is this fabled memory blackhole generator anyway? 20180521 02:16:41<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/447945797267357696/Losowa_mapa-Autozapis19.gz 20180521 02:16:54<+discordbot> the guy reported upwards of 9 GB of memory usage 20180521 02:17:07< gfgtdf> is it the animated water ? 20180521 02:17:22<+discordbot> partly, probably 20180521 02:17:26<+discordbot> but also the huge map size 20180521 02:17:44<+discordbot> scroll around the entire thing and over 14,000 surfaces have been added to the cache 20180521 02:18:00<+discordbot> which increase (expectedly) as you attack and new images are needed 20180521 02:18:14< gfgtdf> heh 20180521 02:18:35<+discordbot> scrolling over water adds several thousand cache entries in a few seconds 20180521 02:18:54<+discordbot> I'm getting near 100% CPU usage as well but that might be because it's an -O0 build with UBsan. 20180521 02:19:11< gfgtdf> unfort i have no idea how this (map) drawing works 20180521 02:19:57-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 240 seconds] 20180521 02:20:23<+discordbot> nearing 20,000 cache entries 20180521 02:20:34<+discordbot> my laptop is as hot as it is playing dota 20180521 02:20:48< gfgtdf> is this 'fixed on ar' ? or does ar still use the same cahce mechanism 20180521 02:21:34<+discordbot> it does not 20180521 02:21:42<+discordbot> well, it does have a texture cache, so.... 20180521 02:21:46<+discordbot> probably the same damn issue 20180521 02:22:01-!- gfgtdf [~chatzilla@x4e363320.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.93 [Firefox 52.8.0/20180430140610]] 20180521 02:22:20<+discordbot> today it looks like it stabilized around 1 GB memory 20180521 02:22:23<+discordbot> It settled around 908 MB for me after scrolling everywhere. 20180521 02:22:32<+discordbot> hasn't broken 20k cache entries 20180521 02:24:12<+discordbot> also the cache is implemented in the most inefficient way possible 20180521 02:24:14<+discordbot> a vector 20180521 02:24:20<+discordbot> of surfaces 20180521 02:25:03<+discordbot> It's not the most inefficient if an adequate allocation can be reserved in advance. 20180521 02:25:09<+discordbot> which is not done 20180521 02:25:21<+discordbot> foo.reserve(20000) 20180521 02:25:54<+discordbot> It doesn't have a huge fingerprint by itself since surfaces are just glorified smart pointers IIRC. 20180521 02:26:26<+discordbot> (Which means that mostly they just have a SDL_Surface* member.) 20180521 02:28:37<+discordbot> Anyway tell me if you come up with new ideas for the color_range issue. 20180521 02:28:54<+discordbot> alright 20180521 02:28:58<+discordbot> can you test the change thing thing now 20180521 02:29:07<+discordbot> I am on it. 20180521 02:29:22<+discordbot> What was the test case again? 20180521 02:29:24<+discordbot> E1S9.1? 20180521 02:29:35<+discordbot> yes 20180521 02:30:03<+discordbot> memory usage stabilizes about 800 with preallocation 20180521 02:31:03<+discordbot> It mostly only helps with fragmentation and CPU overhead from the process of relocating the underlying storage if needed. 20180521 02:31:49<+discordbot> Up to a certain limit the latter isn't going to be very noticeable in the face of I/O overhead from fetching the images from disk. 20180521 02:32:55<+discordbot> yells "SPRITESHEETS" into the wind 20180521 02:33:17<+discordbot> The fix you proposed does help with the map edges. 20180521 02:33:29<+discordbot> looks like the memory overhead isn't much different without prealloc with the same just scrolling around the entire map test 20180521 02:34:25<+discordbot> Oh dear. 20180521 02:34:32<+discordbot> I just had a segfault. 20180521 02:35:09<+discordbot> I was in E1S9.1 and tried to load the SOS of E3S8A to test a more complex sequence but it died in the loading screen. 20180521 02:35:47<+discordbot> ;_; 20180521 02:35:49<+discordbot> "stable" 20180521 02:36:07<+discordbot> Don't make me explain to you, the release manager, what that really means. 20180521 02:36:26<+discordbot> it's not what most people expect it to mean 20180521 02:37:00<+discordbot> And now I used a debug build and nothing happened. 20180521 02:37:25<+discordbot> it happens 20180521 02:37:37<+discordbot> often stuff like oob errors don't happen with debug builds 20180521 02:37:55<+discordbot> Yeah I'm trying with the -O3 build with a debugger attached and nothing happens either. 20180521 02:38:52<+discordbot> are you running with ASan? 20180521 02:39:04<+discordbot> No. 20180521 02:39:48<+discordbot> Tell me. 20180521 02:39:58<+discordbot> Why are my campaigns always involved in bugs like this? 20180521 02:40:13<+discordbot> This has been going on for almost 11 years. 20180521 02:46:11<+discordbot> Because you’re one of the few people that pushes the engine to its limited limits 20180521 02:51:19< irker078> wesnoth: Iris Morelle wesnoth:1.14 6158b80c9400 / src/display.cpp: Fix [change_theme] not updating themed map borders correctly in some cases https://github.com/wesnoth/wesnoth/commit/6158b80c94005f4a9c0d31d2d54dba4eae96a082 20180521 02:52:28<+discordbot> the new gates really make everything look so much nicer 20180521 02:52:59<+discordbot> thougho 20180521 02:54:01<+discordbot> wonder if we should have a door opening sfx 20180521 03:02:42<+discordbot> # Simple dummy tiles for obsolete terrain codes {NEW:OVERLAY (*^Prg) portals/gate-rusty-long-n} {NEW:OVERLAY (*^Prgo) portals/gate-rusty-open-long-n} {NEW:OVERLAY (*^Pwd) portals/door-wooden-long-n} {NEW:OVERLAY (*^Pwdo) portals/door-wooden-open-long-n} 20180521 03:02:48<+discordbot> wat 20180521 03:02:59<+discordbot> These weren't ever in mainline, were they? 20180521 03:03:05<+discordbot> they were 20180521 03:03:08<+discordbot> for like 1 dev release 20180521 03:03:18<+discordbot> dunno why doofus added compatibility 20180521 03:03:19<+discordbot> They should've been dropped then. 20180521 03:03:23<+discordbot> please do 20180521 03:03:45<+discordbot> Ohhh they are still defined, but hidden. 20180521 03:04:07<+discordbot> That... complicates matters since odds are at least one crazy person out there will use them somehow. 20180521 03:04:28<+discordbot> meh 20180521 03:04:54<+discordbot> better to break it then make them tear their hair out wondering why certain terrain filters won't wokr 20180521 03:06:04< irker078> wesnoth: Charles Dang wesnoth:1.14 0c29a16a9e4a / src/random.cpp: Added a static assertion for IEEE 754 floating point compliance https://github.com/wesnoth/wesnoth/commit/0c29a16a9e4a3ce4651cf163a23fd938ab4d8ee8 20180521 03:06:31<+discordbot> ill leave it up to someone else to decide whether to dropthe cmake/scons tests 20180521 03:06:38<+discordbot> IIRC soliton advocated having both 20180521 03:06:42<+discordbot> not sure it's worth it 20180521 03:07:02<+discordbot> I'd prefer removing the CMake/Scons tests. 20180521 03:07:26< irker078> wesnoth: Charles Dang wesnoth:master 1bfa8e363cae / src/random.cpp: Added a static assertion for IEEE 754 floating point compliance https://github.com/wesnoth/wesnoth/commit/1bfa8e363cae458fb0fad80bd2adfbaf46e4cbe0 20180521 03:13:04< irker078> wesnoth: Iris Morelle wesnoth:1.14 fee407ee04e7 / changelog.md data/core/terrain-graphics.cfg data/core/terrain.cfg: Remove door terrains that were supposed to be removed by 1.14 https://github.com/wesnoth/wesnoth/commit/fee407ee04e7797c5e1e1c86d826eee6f11c018c 20180521 03:13:11< irker078> wesnoth: Iris Morelle wesnoth:master 6eddc6f978b0 / changelog.md data/core/terrain-graphics.cfg data/core/terrain.cfg: Remove door terrains that were supposed to be removed by 1.14 https://github.com/wesnoth/wesnoth/commit/6eddc6f978b0c3196bdba67b811459eaabbcaac6 20180521 03:32:35<+discordbot> if you're curious, the reason they existed was doofus originally attempted to have the code detect direction automatically 20180521 03:32:58<+discordbot> he eventually discarded that in favor of directional variant terrains 20180521 03:38:46< irker078> wesnoth: Charles Dang wesnoth:master d9315199ee44 / CMakeLists.txt SConstruct scons/ieee_754.py src/compile_time_tests/ieee_754.cpp: Remove the CMake/Scons compile-time check for IEEE 754 compliance https://github.com/wesnoth/wesnoth/commit/d9315199ee44dfa8a36e4e01a9332ef21505df3c 20180521 03:39:01<+discordbot> Ahh. 20180521 04:00:34< irker078> wesnoth: Charles Dang wesnoth:1.14 7904debd1426 / CMakeLists.txt SConstruct scons/ieee_754.py src/compile_time_tests/ieee_754.cpp: Remove the CMake/Scons compile-time check for IEEE 754 compliance https://github.com/wesnoth/wesnoth/commit/7904debd142671373232fe767d5c8fbe02116ec3 20180521 04:02:45-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Read error: Connection reset by peer] 20180521 04:07:58-!- travis-ci [~travis-ci@ec2-54-227-75-141.compute-1.amazonaws.com] has joined #wesnoth-dev 20180521 04:07:59< travis-ci> wesnoth/wesnoth#18272 (master - d931519 : Charles Dang): The build passed. 20180521 04:07:59< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/381507028 20180521 04:07:59-!- travis-ci [~travis-ci@ec2-54-227-75-141.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180521 04:10:10<+discordbot> What does "Scaled to hex" mean? I have trouble translating it 20180521 04:10:53<+discordbot> Same for "Scaled to zoom" 20180521 04:16:19<+discordbot> what to use when scaling an image to hex size/zoom level 20180521 04:21:34<+discordbot> https://developercommunity.visualstudio.com/content/problem/251769/invalid-syntax-error-when-building-variable-with-t.html 20180521 04:21:34<+discordbot> hm 20180521 04:21:39<+discordbot> an actual useful response this time.. 20180521 04:25:37<+discordbot> i wonder where the /P output goes 20180521 04:37:03<+discordbot> The link provided there explains. 20180521 04:37:15<+discordbot> yes 20180521 04:52:50<+discordbot> wow this output file has thousands of blank lines? 20180521 04:55:00<+discordbot> That's normal. 20180521 05:33:38<+discordbot> submits 20180521 05:38:03< irker078> wesnoth/wesnoth:1.14 Charles Dang 9c1da12ca0 Unit Recall: select the first row after AppVeyor: All builds passed 20180521 06:26:04<+discordbot> @Vultraz I don't see the preprocessed source code in the Visual C++ bug report. Is it awaiting for moderator approval or something? 20180521 06:48:19< irker078> wesnoth: Severin Glöckner wesnoth:1.14 9c4023f9284c / po/wesnoth/de.po: German: work around help bug by using non-breaking space https://github.com/wesnoth/wesnoth/commit/9c4023f9284c55e26ea9c4eb0bbbd639ba6bde8c 20180521 06:48:21< irker078> wesnoth: Severin Glöckner wesnoth:1.14 15aece7f46f9 / po/ (7 files in 7 dirs): updated German translation https://github.com/wesnoth/wesnoth/commit/15aece7f46f906d44c3329244f749de298b54bdc 20180521 06:48:23< irker078> wesnoth: Severin Glöckner wesnoth:1.14 93eb0337c767 / data/core/macros/names.cfg: name generators: update translation hint https://github.com/wesnoth/wesnoth/commit/93eb0337c767855330594a603837de1fbb4674cc 20180521 06:54:41< irker078> wesnoth: Severin Glöckner wesnoth:master 69f17c775d0f / data/core/macros/names.cfg: name generators: update translation hint https://github.com/wesnoth/wesnoth/commit/69f17c775d0fc73207f24cdac8db565c71b11ef2 20180521 06:54:43< irker078> wesnoth: Severin Glöckner wesnoth:master fd8473e3da01 / src/gui/dialogs/multiplayer/mp_create_game.cpp: GUI: add a translation prefix https://github.com/wesnoth/wesnoth/commit/fd8473e3da01bfabdd696883d1fabab599f62da4 20180521 07:02:15-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev 20180521 07:03:39<+discordbot> I replied to the comment 20180521 07:04:21<+discordbot> (Jyrki) 20180521 07:11:35<+discordbot> @jyrkive oh, it says my comment is visible to moderators 🤔 odd 20180521 07:14:52< irker078> wesnoth: Severin Glöckner wesnoth:master 36026a5ff791 / doc/man/wesnoth.6: manpages: clarify deprecation logdomain https://github.com/wesnoth/wesnoth/commit/36026a5ff7911f97fb60819d3da5e0daaf792725 20180521 07:21:24<+discordbot> Seems that the default behavior of [harm_unit] is changed to kill=no in 1.14 20180521 07:37:18<+discordbot> @jyrkive do you think the image cache would be faster if it used an unordered_map? 20180521 07:40:06<+discordbot> It didn't show up when I profiled rendering earlier. 20180521 07:40:10<+discordbot> It's not a bottleneck. 20180521 07:40:23<+discordbot> I see 20180521 07:40:25<+discordbot> alright 20180521 07:40:37<+discordbot> what was? 20180521 07:40:49<+discordbot> Blitting. 20180521 07:41:00<+discordbot> That is, copying surfaces to each other. 20180521 07:41:11<+discordbot> fuck. nothing we can do about that. 20180521 07:43:38<+discordbot> guess I really need to finish my a_r refactor 😦 20180521 07:49:22-!- Nobun [~user@host81-170-dynamic.24-79-r.retail.telecomitalia.it] has joined #wesnoth-dev 20180521 08:21:02< irker078> wesnoth/wesnoth:master nemaara 8add0261e9 Added ability for side 1 to see past gat AppVeyor: All builds passed 20180521 08:21:31-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20180521 08:42:21-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Ping timeout: 252 seconds] 20180521 09:11:15-!- vladimirslavik [vslavik@nat/redhat/x-wylmmebrbsmttvnn] has joined #wesnoth-dev 20180521 09:11:55-!- vladimirslavik [vslavik@nat/redhat/x-wylmmebrbsmttvnn] has quit [Client Quit] 20180521 09:12:17-!- vladimirslavik [vslavik@nat/redhat/x-fssobytltlplfbiy] has joined #wesnoth-dev 20180521 09:45:43< irker078> wesnoth: gfgtdf wesnoth:1.14 619704923ca6 / src/whiteboard/ (side_actions.cpp side_actions.hpp): wb: dont resend 'bump_later' commands https://github.com/wesnoth/wesnoth/commit/619704923ca692705607cc4820e50d4e6665990c 20180521 10:17:31-!- Oebele [~quassel@143.177.58.202] has joined #wesnoth-dev 20180521 10:27:03-!- aeth [~Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Read error: Connection reset by peer] 20180521 10:27:58-!- aeth [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev 20180521 10:28:52< irker078> wesnoth/wesnoth:1.14 doofus-01 0a82dcd4ed cosmetic changes to S6b and S7b maps, to AppVeyor: All builds passed 20180521 10:35:13< irker078> wesnoth: gfgtdf wesnoth:master 5dffb16d5eef / src/whiteboard/ (side_actions.cpp side_actions.hpp): wb: dont resend 'bump_later' commands https://github.com/wesnoth/wesnoth/commit/5dffb16d5eefd0bc014ffceeafd231c20a7d0791 20180521 11:03:58< irker078> wesnoth: Nils Kneuper wesnoth:1.14 bebf9348a8f2 / / (7 files in 6 dirs): updated Scottish Gaelic translation https://github.com/wesnoth/wesnoth/commit/bebf9348a8f2a19ba4fb2f8d63912b0d68a70f8d 20180521 11:04:00< irker078> wesnoth: Nils Kneuper wesnoth:1.14 b6a846c4b2f8 / po/wesnoth-trow/fr.po: updated French translation https://github.com/wesnoth/wesnoth/commit/b6a846c4b2f8e777175919d668eef2937c02621f 20180521 11:04:06< irker078> wesnoth: Nils Kneuper wesnoth:master d942bead1858 / po/ (5 files in 5 dirs): updated Scottish Gaelic translation https://github.com/wesnoth/wesnoth/commit/d942bead185828ac363f834d8b1cf379c1b33563 20180521 11:04:08< irker078> wesnoth: Nils Kneuper wesnoth:master a7dbdb4aa2f9 / po/wesnoth-trow/fr.po: updated French translation https://github.com/wesnoth/wesnoth/commit/a7dbdb4aa2f9377a6433408e0660e5a8ca57cbd3 20180521 11:06:25-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has quit [Ping timeout: 256 seconds] 20180521 11:07:50< irker078> wesnoth: Nils Kneuper wesnoth:master 6bfc5623e7d1 / po/ (9 files in 9 dirs): updated Japanese translation https://github.com/wesnoth/wesnoth/commit/6bfc5623e7d1023e5d0cdb38d6aff504898cfd33 20180521 11:07:56< irker078> wesnoth: Nils Kneuper wesnoth:1.14 115f45cae91b / / (11 files in 10 dirs): updated Japanese translation https://github.com/wesnoth/wesnoth/commit/115f45cae91b0d457f32ae4b5fe2066ebf3649bd 20180521 11:19:42< irker078> wesnoth/wesnoth:master newfrenchy83 7eb5ceb27c Update attack.cpp AppVeyor: vs2015/Release Failed 20180521 11:19:43< irker078> Details: https://ci.appveyor.com/project/wesnoth/wesnoth-7lnpw/build/Wesnoth-VS2015-master-3482 20180521 11:43:57-!- Oebele [~quassel@143.177.58.202] has quit [Ping timeout: 240 seconds] 20180521 11:50:58-!- Oebele [~quassel@143.177.58.202] has joined #wesnoth-dev 20180521 12:05:05<+discordbot> I don’t agree the gates need to be see-through... 20180521 12:13:48<+discordbot> Should I set travis xcode to DEBUG scheme? In DEBUG scheme it will show every warning 20180521 12:14:01<+discordbot> In RELEASE scheme are most of them mutted 20180521 12:14:24<+discordbot> Yes, I think it would be better to set to the Release scheme. 20180521 12:15:22<+discordbot> wait, you said 'to Release scheme' but we are using release scheme now. I was asking if I should change it to DEBUG 20180521 12:15:41<+discordbot> Ah, right. I meant Debug scheme. 20180521 12:16:00<+discordbot> I mixed them up while writing the message. 20180521 12:21:32< zookeeper> @Vultraz, hey, we agree on something, awesome. :] 20180521 12:21:49<+discordbot> \o/ 20180521 12:21:55< zookeeper> (i mean i don't think they need to be see-through in general; maybe it's warranted in some scenarios) 20180521 12:23:24< irker078> wesnoth: Martin Hrubý (hrubymar10) wesnoth:1.14 787bf47225c3 / utils/travis/steps/script.sh: Switch travis Xcode to DEBUG scheme https://github.com/wesnoth/wesnoth/commit/787bf47225c37f4dd5a8ec733d16529d8d9e2a1d 20180521 12:26:11< irker078> wesnoth: Martin Hrubý (hrubymar10) wesnoth:master dee670d2b588 / utils/travis/steps/script.sh: Switch travis Xcode to DEBUG scheme https://github.com/wesnoth/wesnoth/commit/dee670d2b5882c50464423b503477b0ae7593057 20180521 12:34:39-!- Oebele [~quassel@143.177.58.202] has quit [Remote host closed the connection] 20180521 13:16:09-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180521 13:16:23-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180521 13:27:39< irker078> wesnoth: Charles Dang wesnoth:1.14 a3ef80853caa / src/image.cpp: Preallocate memory for image/file presence caches https://github.com/wesnoth/wesnoth/commit/a3ef80853caac8fd1c115418b2474af07dd1cb14 20180521 13:29:08< irker078> wesnoth: Charles Dang wesnoth:master 949266b5aac3 / src/image.cpp: Preallocate memory for image/file presence caches https://github.com/wesnoth/wesnoth/commit/949266b5aac34dea0f5d0b4d14228a32cd0eb4c3 20180521 13:31:55< irker078> wesnoth: Jyrki Vesterinen wesnoth:1.14 bfe88ff19a32 / src/image.cpp: Revert "Preallocate memory for image/file presence caches" https://github.com/wesnoth/wesnoth/commit/bfe88ff19a3268322c83c0938989f57a858a8387 20180521 13:32:20<+discordbot> :thonk: 20180521 13:32:36<+discordbot> alright? 20180521 13:34:09< irker078> wesnoth: Jyrki Vesterinen wesnoth:master 1066de1a8273 / src/image.cpp: Revert "Preallocate memory for image/file presence caches" https://github.com/wesnoth/wesnoth/commit/1066de1a827338840cfe44a32824e6d1a2e9207f 20180521 13:35:02< irker078> wesnoth/wesnoth:1.14 Severin Glöckner 93eb0337c7 name generators: update translation hint AppVeyor: All builds passed 20180521 13:48:10< irker078> wesnoth: Charles Dang wesnoth:1.14 d30cf8ede3f9 / changelog.md: Updated changelog https://github.com/wesnoth/wesnoth/commit/d30cf8ede3f90cc24b4b573b5ab38d5ac7270a4d 20180521 13:50:27<+discordbot> we really should separate the changes that are on 1.14 in the master changelog... oh well 20180521 13:51:31-!- travis-ci [~travis-ci@ec2-54-158-234-7.compute-1.amazonaws.com] has joined #wesnoth-dev 20180521 13:51:31< travis-ci> wesnoth/wesnoth#18284 (1.14 - 787bf47 : Martin Hrubý (hrubymar10)): The build has errored. 20180521 13:51:32< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/381649274 20180521 13:51:32-!- travis-ci [~travis-ci@ec2-54-158-234-7.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180521 13:52:01<+discordbot> I also don't agree with the policy that everything in the 1.14.x changelogs is also considered notable enough to add to the 1.15.0 changelog. 20180521 13:52:39< irker078> wesnoth: Charles Dang wesnoth:master 998cf67e1a3b / changelog.md: Updated changelog https://github.com/wesnoth/wesnoth/commit/998cf67e1a3b14d9905a31b45cea7e8af4765d8c 20180521 13:53:03<+discordbot> The resulting changelog is so large that it's difficult to spot the actual notable changes in it. 20180521 13:53:11<+discordbot> @hrubymar10 oh dear, warnings everywhere 20180521 13:53:21<+discordbot> (and yes, that's the problem) 20180521 13:53:21<+discordbot> 😄 20180521 13:53:38<+discordbot> no one can tell what's actually new 20180521 13:54:28<+discordbot> @hrubymar10 probably need to drop -Wsign-conversion and -Wshorten-64-to-32 20180521 13:55:06<+discordbot> Ok, I'll drop these asap. I am at work now 20180521 13:58:25<+discordbot> Did you enable them explicitly? 20180521 13:59:44<+discordbot> The Clang project doesn't document which warnings -Wall enables. However, over at the GCC side, -Wall doesn't enable either -Wconversion or -Wsign-conversion. 20180521 14:07:06<+discordbot> In Xcode 'must' define which warnings are on and which off. When is warning set as off, then it add -Wno-* flag and when it is set as on -W-* flag 20180521 14:07:33<+discordbot> So I'll turn these 2 (-Wsign-conversion and -Wshorten-64-to-32) off 20180521 14:09:22<+discordbot> /Users/travis/build/wesnoth/wesnoth/utils/travis/ccache-clang -x c++ -arch x86_64 -fmessage-length=0 -fdiagnostics-show-note-include-stack -fmacro-backtrace-limit=0 -fcolor-diagnostics -std=c++11 -stdlib=libc++ -Wno-trigraphs -fpascal-strings -O0 -fno-common -Wmissing-field-initializers -Wmissing-prototypes -Wunreachable-code -Wnon-virtual-dtor -Wno-overloaded-virtual -Wno-exit-time-destructors -Wmissing-braces -Wparentheses 20180521 14:09:22<+discordbot> -Wswitch -Wunused-function -Wunused-label -Wunused-parameter -Wunused-variable -Wunused-value -Wempty-body -Wuninitialized -Wno-unknown-pragmas -Wno-shadow -Wno-four-char-constants -Wno-conversion -Wconstant-conversion -Wint-conversion -Wbool-conversion -Wenum-conversion -Wfloat-conversion -Wnon-literal-null-conversion -Wobjc-literal-conversion -Wsign-compare -Wshorten-64-to-32 -Wno-newline-eof -Wno-c++11-extensions 20180521 14:09:22<+discordbot> -DLOCALEDIR=\"translations\" -DFIFODIR=\"/var/run/wesnothd\" -isysroot /Applications/Xcode-9.2.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.13.sdk -fasm-blocks -fstrict-aliasing -Wdeprecated-declarations -Winvalid-offsetof -mmacosx-version-min=10.8 -g -Wsign-conversion -Winfinite-recursion -Wmove -Wcomma -Wblock-capture-autoreleasing -Wstrict-prototypes -Wrange-loop-analysis - 20180521 14:09:36<+discordbot> This much is only a fraction of the warnings Clang supports. 20180521 14:10:24<+discordbot> I think Xcode either doesn't expose them all, or doesn't bother to pass either flag if it aligns with Clang's default setting. 20180521 14:14:45-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev 20180521 14:15:31-!- travis-ci [~travis-ci@ec2-54-158-234-7.compute-1.amazonaws.com] has joined #wesnoth-dev 20180521 14:15:32< travis-ci> wesnoth/wesnoth#18285 (master - dee670d : Martin Hrubý (hrubymar10)): The build has errored. 20180521 14:15:32< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/381650068 20180521 14:15:32-!- travis-ci [~travis-ci@ec2-54-158-234-7.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180521 14:26:42<+discordbot> just confirmed a 13 year old bugreport 20180521 14:45:17-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20180521 14:48:12-!- travis-ci [~travis-ci@ec2-54-158-234-7.compute-1.amazonaws.com] has joined #wesnoth-dev 20180521 14:48:13< travis-ci> wesnoth/wesnoth#18286 (1.14 - a3ef808 : Charles Dang): The build has errored. 20180521 14:48:13< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/381674819 20180521 14:48:13-!- travis-ci [~travis-ci@ec2-54-158-234-7.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180521 15:04:12< irker078> wesnoth/wesnoth:master Severin Glöckner 36026a5ff7 manpages: clarify deprecation logdomain AppVeyor: All builds passed 20180521 15:12:16-!- travis-ci [~travis-ci@ec2-54-227-75-141.compute-1.amazonaws.com] has joined #wesnoth-dev 20180521 15:12:17< travis-ci> wesnoth/wesnoth#18288 (1.14 - bfe88ff : Jyrki Vesterinen): The build has errored. 20180521 15:12:17< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/381676296 20180521 15:12:17-!- travis-ci [~travis-ci@ec2-54-227-75-141.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180521 15:17:20-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection] 20180521 15:39:34-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.] 20180521 15:44:50-!- travis-ci [~travis-ci@ec2-54-158-234-7.compute-1.amazonaws.com] has joined #wesnoth-dev 20180521 15:44:52< travis-ci> wesnoth/wesnoth#18289 (master - 1066de1 : Jyrki Vesterinen): The build has errored. 20180521 15:44:52< travis-ci> Build details : https://travis-ci.org/wesnoth/wesnoth/builds/381676910 20180521 15:44:52-!- travis-ci [~travis-ci@ec2-54-158-234-7.compute-1.amazonaws.com] has left #wesnoth-dev [] 20180521 15:46:24<+discordbot> Oo, which one 20180521 16:08:32<+discordbot> there is also -Wall flag so... 20180521 16:08:56< celticminstrel> ? 20180521 16:11:03<+discordbot> sorry, I answered to jyrkive's post which is about 2 hours old 20180521 16:18:46< irker078> wesnoth: Martin Hrubý (hrubymar10) wesnoth:1.14 0eb38d526cc3 / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Xcode Disable -Wsign-conversion and -Wshorten-64-to-32 flags https://github.com/wesnoth/wesnoth/commit/0eb38d526cc3ab769b8d684dfb218712ef551d0b 20180521 16:19:38< irker078> wesnoth: Martin Hrubý (hrubymar10) wesnoth:master 6f2235088cf1 / projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Xcode Disable -Wsign-conversion and -Wshorten-64-to-32 flags https://github.com/wesnoth/wesnoth/commit/6f2235088cf11cd4359763567320b1a02f4e148e 20180521 16:19:42< irker078> wesnoth/wesnoth:master newfrenchy83 7eb5ceb27c Update attack.cpp AppVeyor: 1/2 builds failed 20180521 16:19:43< irker078> Details vs2015/Release: https://ci.appveyor.com/project/wesnoth/wesnoth-7lnpw/build/Wesnoth-VS2015-master-3482 20180521 16:24:28<+discordbot> Since someone was asking for an example for the issue with music crossfade blocking UI actions in my campaign ... 20180521 16:24:56<+discordbot> Just make sure you have music enabled and that the main menu music is playing at full volume, launch HttT, and spam Escape to skip all actions (including the story screen) as quick as possible. 20180521 16:25:23< celticminstrel> ... 20180521 16:25:35<+discordbot> In particular you'll notice that the story text fade in blocks on the first music crossfade. 20180521 16:26:02<+discordbot> And that pressing Escape on the first in-game message doesn't instantly dismiss it. 20180521 16:26:54<+discordbot> Also I reckon the other music bug isn't going to be fixed before 1.14.2 then. 20180521 16:28:33<+discordbot> OK, I can take a look later this week. 20180521 16:28:40<+discordbot> But it's indeed after 1.14.2. 20180521 16:29:38< irker078> wesnoth: Severin Glöckner wesnoth:1.14 d842aeea4595 / data/campaigns/Legend_of_Wesmere/scenarios/chapter2/07_Elves_Last_Stand.cfg: LoW 2 S7 - fix repeated endlevel execution https://github.com/wesnoth/wesnoth/commit/d842aeea45958851e900ac4cfe11b3a6cc94f3c6 20180521 16:29:45<+discordbot> As for 1.12, it doesn't happen there because (?) there isn't a music crossfade effect in the first place. 20180521 16:30:15< irker078> wesnoth: Severin Glöckner wesnoth:master f291deb6cfb6 / data/campaigns/Legend_of_Wesmere/scenarios/chapter2/07_Elves_Last_Stand.cfg: LoW 2 S7 - fix repeated endlevel execution https://github.com/wesnoth/wesnoth/commit/f291deb6cfb6aa3819391fb0fc242ab758d91bd3 20180521 16:32:39<+discordbot> Wow, biggie reported: https://github.com/singalen/wesnoth/issues/90 I wonder if it's a part of #2638. 20180521 17:30:30-!- sevu [~Shiki@p54854C72.dip0.t-ipconnect.de] has joined #wesnoth-dev 20180521 17:43:56< sevu> what is up with code like this: https://github.com/wesnoth/wesnoth/blob/master/data/campaigns/Legend_of_Wesmere/scenarios/chapter2/07_Elves_Last_Stand.cfg#L506-L510 20180521 17:44:31< sevu> scenario_end isn't listed in the wiki as predefined event name, though it doesn't seem to be called customly 20180521 17:44:59< celticminstrel> Ask gfgtdf maybe? 20180521 17:45:15< celticminstrel> I think he might've added an event like that. 20180521 17:46:01< sevu> [proceed_to_next_scenario] is not mentioned in the wiki as well - I say recently a commit where you removed such code 20180521 17:46:07< sevu> *saw 20180521 17:46:17< celticminstrel> Hm? 20180521 17:46:26< sevu> I thought it is maybe both legacy code 20180521 17:46:32< celticminstrel> No they're both new things... 20180521 18:02:40< zookeeper> i don't remember the details of why that stuff was added, but yes, gfgtdf did it so he'd know. 20180521 18:18:57-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180521 18:19:08-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180521 18:34:26< irker078> wesnoth/wesnoth:master newfrenchy83 cacb447dba Update abilities.cpp AppVeyor: All builds passed 20180521 18:42:27-!- mkdroid [~null@unaffiliated/matthiaskrgr] has joined #wesnoth-dev 20180521 18:45:31-!- sevu [~Shiki@p54854C72.dip0.t-ipconnect.de] has quit [Disconnected by services] 20180521 18:45:31-!- shiki [~Shiki@p5485481F.dip0.t-ipconnect.de] has joined #wesnoth-dev 20180521 18:47:26-!- sevu|shiki [~Shiki@p54854DBB.dip0.t-ipconnect.de] has joined #wesnoth-dev 20180521 18:47:55-!- sevu|shiki is now known as sevu 20180521 18:50:12-!- shiki [~Shiki@p5485481F.dip0.t-ipconnect.de] has quit [Ping timeout: 245 seconds] 20180521 18:55:12-!- sevu [~Shiki@p54854DBB.dip0.t-ipconnect.de] has quit [Ping timeout: 245 seconds] 20180521 18:58:50-!- sevu [~Shiki@p54854936.dip0.t-ipconnect.de] has joined #wesnoth-dev 20180521 19:03:57-!- mkdroid [~null@unaffiliated/matthiaskrgr] has quit [Quit: I'll be back!] 20180521 19:07:19-!- Bhoren [~Bhoren_wh@2a01:e0a:c:2150:f142:8c8e:f549:55d8] has joined #wesnoth-dev 20180521 19:16:20< celticminstrel> ...WTH is issue 3151? 20180521 19:17:07< celticminstrel> It sounds like "set_terrain does nothing on a generated map" but why would the map being generated have any effect on anything that happens once the scenario gets started? 20180521 19:28:47-!- shiki [~Shiki@p54854120.dip0.t-ipconnect.de] has joined #wesnoth-dev 20180521 19:28:47-!- sevu [~Shiki@p54854936.dip0.t-ipconnect.de] has quit [Disconnected by services] 20180521 19:28:55-!- shiki is now known as sevu 20180521 19:43:43-!- Bhoren [~Bhoren_wh@2a01:e0a:c:2150:f142:8c8e:f549:55d8] has quit [Quit: Leaving] 20180521 19:47:38< sevu> The LoW bug seems to be resolved .. apparently introduced by c086f0c and resolved by f7d5194 20180521 20:27:58< sevu> There is still this one blocker ... https://github.com/wesnoth/wesnoth/issues/3099 20180521 20:30:27< zookeeper> ah, i'll look into it 20180521 20:40:22-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20180521 20:42:24-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection] 20180521 20:49:10< irker078> wesnoth/wesnoth:1.14 Charles Dang d30cf8ede3 Updated changelog AppVeyor: All builds passed 20180521 20:51:41< zookeeper> so do we want to demand PR's to have proper commit messages (72 chars, campaign prefixes, etc) before merging or not (in cases where the person merging it can't use squash and thus rewrite the message)? 20180521 20:52:29<+discordbot> I used to demand that. 20180521 20:53:20<+discordbot> Except for the module prefix thing, since that's an informal convention I had which vultraz kind of adopted. 20180521 20:53:23-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev 20180521 21:10:59< zookeeper> celticminstrel, https://github.com/wesnoth/wesnoth/issues/3099#issuecomment-390783299 <- any quick thoughts? 20180521 21:21:37< celticminstrel> Commented. 20180521 21:21:58< celticminstrel> Worth a try, at least. 20180521 21:22:48< celticminstrel> I think we should demand the campaign prefixes on PRs, at least. 20180521 21:22:59< celticminstrel> I couldn't care less about character limits. 20180521 21:35:36< irker078> wesnoth: gfgtdf wesnoth:1.14 92f466ec81cc / src/playsingle_controller.cpp: don't generate a savegame snapshot unless needed https://github.com/wesnoth/wesnoth/commit/92f466ec81cc49d7461dec0a30a1ba818c444924 20180521 21:40:33< celticminstrel> Does that fix the on_save bug gfgtdf? 20180521 21:48:50-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection] 20180521 21:59:42-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180521 21:59:55-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180521 22:19:53< zookeeper> ok, got a fix for the SoF issue... 20180521 22:40:24<+discordbot> @zookeeper the prefix is good, but the 72 character limit isn't a hard one. I try to keep my commit messages generally around that range to prevent that running too long, but you will never see me insist on a strict 72 limit. In any context. 20180521 22:41:33<+discordbot> (I find that number way too low to be a limit for anything these days) 20180521 22:43:34< zookeeper> @Vultraz, okay. perhaps you'd wish to politely suggest to doofus that he may wish to adopt the prefix style? 20180521 22:45:16<+discordbot> I can mention it next time, but I don't think it's worth forcing in this case 20180521 22:46:00<+discordbot> Release is supposed to be this morning 🤔 20180521 22:46:02< zookeeper> better to mention it _before_ the next time, as then it'll be too late again 20180521 22:46:41<+discordbot> are you merging his utbs PRs? 20180521 22:48:10< zookeeper> they look okay at a glance, but i haven't _tested_ them 20180521 22:48:23< celticminstrel> I mean, anyone can suggest to doofus about adopting the prefixes. 20180521 22:48:27<+discordbot> since when did we test campaign changes 😛 20180521 22:48:29< celticminstrel> Doesn't have to be Vultraz. 20180521 22:48:52<+discordbot> is referring to himself since he never played through NR post-refactor 20180521 22:49:15< celticminstrel> Come on discordbot, why can't you send actions in a logical manner. 20180521 22:49:50<+discordbot> Ya know 20180521 22:50:13<+discordbot> Because there isn't an equivalent to CTCP ACTION (which isn't even part of IRC) on Discord. 20180521 22:50:21<+discordbot> The /me command merely italicizes the entire line. 20180521 22:51:23<+discordbot> For that matter, the IRC /me command is actually a client-side thing that actually sends a message containing a CTCP ACTION command, and it just so happens that most IRC clients support it. 20180521 22:52:32<+discordbot> oh, yeah....is gfgtsd around 20180521 22:52:34<+discordbot> gftdf 20180521 22:52:35<+discordbot> gfgtdf 20180521 22:53:04<+discordbot> 20180522 09:52:21 error engine: cannot handle require_scenario=yes because we don't know from which addon that scenario came from 20180521 22:55:37<+discordbot> also: 20180521 22:55:37<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/448257585280319488/unknown.png 20180521 22:55:42<+discordbot> ponders 20180521 22:56:16-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has quit [Disconnected by services] 20180521 22:56:47-!- Ravana_ [~Ravana@unaffiliated/ravana/x-2327071] has joined #wesnoth-dev 20180521 22:57:15<+discordbot> the giant-ass text doesn't look bad. 20180521 22:57:32<+discordbot> Language. 20180521 22:57:55<+discordbot> I see you're Captain America now 20180521 22:58:24<+discordbot> Well if you're going to put it like that then you're the bad guy who made me come out of retirement lol 20180521 22:58:44<+discordbot> :thonk: 20180521 23:01:51< irker078> wesnoth: Severin Glöckner wesnoth:1.14 dc100b392045 / po/ (7 files in 7 dirs): updated German translation https://github.com/wesnoth/wesnoth/commit/dc100b392045ba49361e9ffeb0f58fe7cf477f70 20180521 23:02:20<+discordbot> oooo, dunefolk translation 20180521 23:09:51-!- gfgtdf [~gfgtdf@134.76.63.8] has joined #wesnoth-dev 20180521 23:11:28< gfgtdf> vultraz: ye we should just silence that warning, it's not suprsing that i doesn't find the addon id for mainline scenarios 20180521 23:12:59< irker078> wesnoth: gfgtdf wesnoth:1.14 ad1105d7f9da / src/saved_game.cpp: remove 'cannot handle require_scenario=' warning https://github.com/wesnoth/wesnoth/commit/ad1105d7f9dacfbacdf952c79f8cec6b0f94b8dd 20180521 23:13:46-!- gfgtdf [~gfgtdf@134.76.63.8] has quit [Client Quit] 20180521 23:16:16< irker078> wesnoth/wesnoth:1.14 Severin Glöckner d842aeea45 LoW 2 S7 - fix repeated endlevel executi AppVeyor: All builds passed 20180521 23:37:13-!- Nobun [~user@host81-170-dynamic.24-79-r.retail.telecomitalia.it] has quit [Quit: WeeChat 2.1-dev] 20180521 23:38:26<+discordbot> gfgtfd: any thoughts on this? https://github.com/wesnoth/wesnoth/issues/3152 20180521 23:39:32< irker078> wesnoth: gfgtdf wesnoth:master 29120661f457 / src/saved_game.cpp: remove 'cannot handle require_scenario=' warning https://github.com/wesnoth/wesnoth/commit/29120661f457189def15788db8eb890f40c10e6c 20180521 23:40:31< irker078> wesnoth: ln-zookeeper wesnoth:1.14 1db98d87d85a / data/campaigns/Sceptre_of_Fire/ (11 files in 2 dirs): SoF S9: Fixed some issues with the volcano eruption (fixes #3099) https://github.com/wesnoth/wesnoth/commit/1db98d87d85a26968e88dd425c3256fbe65936c7 20180521 23:40:37< zookeeper> i hope i know what i'm doing... 20180521 23:46:25<+discordbot> what say you guys? 20180521 23:46:26<+discordbot> https://cdn.discordapp.com/attachments/259976436490829825/448270372446404618/unknown.png 20180521 23:47:04< zookeeper> @Vultraz, gonna tag on time? 20180521 23:47:15<+discordbot> ( @shadowm especially, since I don't know if it fits her campaign aesthetic) 20180521 23:47:37<+discordbot> @zookeeper somewhat 20180521 23:47:53<+discordbot> have you committed everything you want to commit? 20180521 23:48:12< zookeeper> i was just wondering if i had time to squeeze in something extra 20180521 23:48:21<+discordbot> yes 20180521 23:49:20< zookeeper> oh? well, in that case i'll check out that one UtBS bug... 20180521 23:51:16<+discordbot> take as much time asyou need 20180521 23:51:40<+discordbot> two years later 20180521 23:52:16<+discordbot> :thonk: 20180521 23:52:23<+discordbot> 😛 20180521 23:54:44< celticminstrel> Ah, so it's actually Discord's fault for having a dumb way of representing actions, got it. :/ 20180521 23:54:51< zookeeper> @Vultraz, uh, the Quenoth Shyde TC is broken... 20180521 23:57:01<+discordbot> @Vultraz To be honest, I'm not completely sure it works for IftU and AtS since they kind of slowly let go of the whole pseudo-medieval times aesthetic that permeates the game. 20180521 23:57:42<+discordbot> Which culminates in the last third of AtS completely ditching it. 20180521 23:59:28<+discordbot> You can actually tell that IftU scenario 13 is the turning point. --- Log closed Tue May 22 00:00:31 2018