Time Nick Message 04:42 sofar submitting a coverity build again to the coverity server 05:08 sofar https://scan.coverity.com/projects/minetest-official 05:08 sofar results available 05:16 sofar nerzhul: cov claims a new leak in m_player_database 05:16 sofar not a big deal I think, but would still be nice to plug it 10:49 Beerholder_ Hi all, quick C++ scripting engine question. Any Lua functions registered under InitializeClient are functions executed on the client only? 10:52 red-002 yes 10:53 red-002 We should probalay rename Initialize to InitializeServer 11:18 Beerholder_ Thanks :) 11:38 josema Hi all!. Thank you 13:26 kilbith it's me or the clouds are slightly red now? 13:27 kilbith especially the 2d clouds 13:32 nerzhul is this due to paramat fast merge on cloud API ? i think 13:59 red-002 #5689 doesn't seem to even be minetest-related 13:59 ShadowBot https://github.com/minetest/minetest/issues/5689 -- Recebo erro de memória insuficiente ao tentar compilar esse arquivo 14:50 Beerholder_ So ... Pretty much all the functions seem to be bound to core: lua_setglobal(m_luastack, "core") (and aliased via minetest = core in builtin). Is it generally considered a requirement that any function to be added to the API must be bound to the global variable core? 14:52 Beerholder_ Or is there room to define an additional global e.g. in which some functionality can be grouped together? 14:52 rubenwardy everything should be in core 14:52 rubenwardy unless it's a constructor 14:52 rubenwardy like ItemStack 14:52 rubenwardy or, vector.new() (ew 14:52 rubenwardy ) 14:56 Beerholder_ Okidoki :) This is re. the MinetestBrowser btw. I was considering whether to provide a different global in the Lua API so that you can e.g. do browser.create(name, url) or minetest.browser_create(name, url) and then pass the browser object. 14:56 Beerholder_ I think I'd only need create and delete, as other things (like load_url, back, forward, etc.) could be done on the browser object that the core function returns 14:57 Beerholder_ Looking for the cleanest solution from a Minetest perspective :) 14:57 rubenwardy Will you need to communicate with the C++ api when creating the browser? 14:57 rubenwardy if so, I'd do minetest.create_browser 14:57 rubenwardy or minetest.add_browser(Browser()) 14:58 Beerholder_ Under the hood I will be calling the C++ code controlling the browser yes 14:58 rubenwardy that latter one is inconsistent with the API, but it would be nice if we could migrate to that 14:58 rubenwardy actually, forget that 14:58 rubenwardy I'd do minetest.create_browser() which returns a user object 14:59 rubenwardy *userdata 15:03 Beerholder_ Ok :) I will have a look at the mainmenu code as an example as that hooks back into the GUIEngine and follow the approach you suggested. Thanks rubenwardy! 15:05 Beerholder_ P.s. I good chance that it will be named differently later as I am keeping other things besides the browser in the back of my mind (VNC like clients, Wayland, etc.) 15:06 Beerholder_ Then the other biggie is the V8 JavaScript engine that comes with Chrome 15:08 Beerholder_ I will do a proof of concept where I will expose a single function e.g. place_node to the V8 JavaScript engine so that one can click on a button in the browser and it will place a node in the world 15:10 Beerholder_ I will for now hack it into the client but later on I think it should be placed under script in its own directory (e.g. script/javascript or script/v8) 18:07 paramat #5675 is now well tested and good for merge +1 18:07 ShadowBot https://github.com/minetest/minetest/issues/5675 -- Sneak: Improve and fix various things by SmallJoker 18:09 Krock good to hear. thanks for testing :) 18:10 sofar Krock: too bad the sound fade pr has those fade issues :( 18:10 Krock the issues themselves are fading out, so no problem 18:12 sofar it completely locks up and crashes MT 18:13 Krock even with the most recent commit I pushed about 10 mins ago? 18:13 sofar hadn't seen those yet 18:14 Krock oh, I thought you started the discussion because you've seen it in the notifications :3 18:16 * sofar looks 18:17 sofar Krock: ok, I'll see if I can test tonight 18:17 sofar also, it is entirely unclear to me what are logical values for fade and step 18:17 Krock will do some tests aswell but rn the motivation isn't here :/ 18:17 sofar e.g. if I want a 3 second fade out, what value do I use? 18:18 Krock if you have a volume of 1, then it would be..... 18:18 paramat celeron55 if possible could you check parts of #5672 ? the functionality in game.cpp, sky.cpp and sky.h is fine, i'm just not sure about the changes that add the bool to 'set sky', especially server / network / packet stuff 18:18 ShadowBot https://github.com/minetest/minetest/issues/5672 -- Set sky: Add bool for clouds with custom skybox by paramat 18:18 Krock 0.33, perhaps. ugh, I don't know either 18:18 sofar Krock: hehe, lmao 18:19 Krock yes right, 0.33 18:19 Krock and for volume 2 to 0 in 3 seconds it's 0.667 18:20 Krock basically volume/time = step value 18:20 Krock also, negative for fading out, positive for fading in 18:21 Krock I used this one to test it locally: https://hastebin.com/erevatepoz.lua 18:25 paramat nore sofar Krock if i add correct sounds on merge what is your opinion on game#1668 ? 18:25 ShadowBot https://github.com/minetest/minetest_game/issues/1668 -- add snow and ice stairs and slabs by timsoftgit 18:26 paramat i might merge that, chests and game#1387 in a few mins 18:26 Krock LGTM 18:26 ShadowBot https://github.com/minetest/minetest_game/issues/1387 -- Gunpowder (and tnt.burn) will trigger the on_ignite of nodes by Ferk 18:26 sofar paramat: the snow sounds is just a minor thing, just add them and merge the footstep pr imho 18:26 paramat ok 18:32 paramat just realised the 'snow's should be 'snowblock' too 18:34 paramat ok will merge those 3 over the next few mins, will take a while 18:57 paramat merged! 19:00 sfan5 so what was this problem with clouds being red(?) 19:02 paramat are you referring to my screenshot in the clouds API PR? 19:02 sfan5 nah someone mentioned a problem here a few hours ago 19:02 paramat i see it 19:03 paramat i mean, the comment 19:07 paramat due to my 'fast merge'? PR opened 16 Sep 2016, 216 comments, 16 commits, 6 reviewers and multiple requested updates :] 19:15 paramat well someone can colour sample screenshots or window captures to check colours 19:20 sfan5 2017-05-01 21:20:47: ERROR[Main]: ModError: Failed to load and run script from /home/stefan/minetest/bin/../games/minetest_game/mods/stairs/init.lua: 19:20 sfan5 2017-05-01 21:20:47: ERROR[Main]: ...minetest/bin/../games/minetest_game/mods/stairs/init.lua:613: unexpected symbol near ')' 19:20 sfan5 um 19:24 rubenwardy it's the comma. Oops paramat 19:24 sfan5 there's more wrong than the comma 19:24 sfan5 i've pushed a fix 19:24 rubenwardy erm no 19:24 rubenwardy https://github.com/minetest/minetest_game/blob/master/mods/default/functions.lua#L18 19:25 rubenwardy node_sound_*_defaults can take a function 19:25 rubenwardy *table 19:25 rubenwardy :P 19:25 sfan5 woops 19:26 sfan5 fixed 19:37 paramat oops i thought that comma would not cause a problem, as it's correct to do this elsewhere 19:38 paramat thanks 19:39 sfan5 you can only do this in tables 19:39 sfan5 not in argument lists 19:39 paramat ok, just what i was going to ask 19:43 sofar luacheck ftw :) 20:55 sfan5 ~tell paramat the cloud complaint does seem to have some validity https://kitsunemimi.pw/i/d74613f259b40ce1.png 20:55 ShadowBot sfan5: O.K. 20:56 kilbith it's more obvious on the flat clouds, on the mainmenu 20:57 sofar sorry, what's the difference? 20:57 sofar color? 20:57 kilbith it has a subtle red tit 20:57 kilbith *tint 20:57 sofar hahahaha 20:58 kilbith ? 20:58 sofar nobody would complain if there was a subtle red tit in the clouds, trust me 20:58 sofar any idea where the tint comes from? 20:58 kilbith please note I wasn't complaining, just sharing an observation 20:58 sofar oh it's fine 20:59 sofar I wonder if the whole sky has a subtle red tint too 22:02 Fixer sofar: is it only me, or there is some slight delay when chest inv pop ups? 22:03 sofar there is 22:04 sofar there are actually 2 delays 22:04 sofar one is from lag 22:04 sofar the other is the formspec 22:04 sofar it's really short, though 22:04 sofar enough to render the chest open visually but nothing more 22:05 sofar before the formspec appears 22:08 Fixer could be 22:09 Fixer sofar: animation is independant of that? 22:09 sofar of what? 22:10 Fixer sofar: of percepted delay 22:11 sofar I don't understand, I just explained that the time that the formspec appears and the time that the chest shows open are directly related to (1) a general lag value and (2) a small delay in the code 22:11 sofar the chest opening is only related to server lag 22:11 Fixer sofar: so it is totally unrelated to chest animation, i see 22:12 sofar the length of the animation in time is 0 22:12 sofar mesh update, basically 22:12 sofar there is no animation, effectively 22:12 sofar it's just a node swap 22:31 Fixer nice 22:59 paramat hmm 23:03 paramat yes gimp shows a change in hue and saturation 23:03 paramat much lower saturation and a magenta hue 23:03 paramat instead of blue 23:04 paramat i'll open an issue 23:05 paramat the PR did rework cloud colouring and i asked if it was identical, but no answer, wish i had tested it 23:06 paramat tested colours, that is 23:26 paramat #5692 23:26 ShadowBot https://github.com/minetest/minetest/issues/5692 -- Cloud colours have changed due to the clouds API 23:32 nore paramat: maybe try to change the cloud default color to #f0f0ffe5 (instead of #fff0f0e5 currently) and see what the result is? 23:33 nore (can't test right now) 23:41 paramat ok 23:42 sofar well that certainly explains that 23:42 paramat i see, RGB 23:42 sofar I wouldn't do that even 23:42 sofar #f0f0f0e5 seems better 23:43 sofar or even #ffffffe5 23:44 paramat thing is, does the blue tint come from fog taking the colour of the blue skybox, or were the clouds actually blueish? 23:44 sofar the old color is very blue 23:44 sofar the new color is pink/red 23:45 paramat as far as i know clouds have a fog colour that takes it's hue from the skybox colour 23:45 sofar if the clouds are no longer being corrected in color from fog we need to remove the indirect color correction 23:46 sofar V being lower means it's just (slightly) darker 23:46 sofar S is saturation, so interestingly the old clouds had much more color 23:47 sofar S 10 -> 3/4 23:47 paramat i hope they still have fog with hue taken from the skybox, that is necessary to make them blend into the horizon 23:47 sofar but the hue changed a lot 23:47 paramat for example with my moonrealm mod the skybox is black, and there is black cloud fog 23:48 sofar so what I think needs to happen is that we need to reverse correct for fog color 23:48 sofar or, alternatively, make clouds white and if it gets shaded with fog just say "oh well, that's part of the mechanic" 23:48 sofar which actually seems reasonable 23:48 sofar just look at the skybox mod - it modifies the fog color 23:48 sofar and rightfully so 23:51 paramat looking at the PR diff, it looks like a possible accidental swapping of RGB components .. 23:51 paramat need to check though 23:53 paramat new code has RGB 255 240 240, old code has RGB 240 240 255 23:55 paramat i hope this is the cause anyway :] 23:55 paramat which is what nore suggests 23:56 nore paramat: yeah, that was what I guessed 23:57 paramat i will test, just tired 23:57 nore didn't look at the pr though, but it seems reasonable enough 23:57 nore sorry can't test, I'm at work :p 23:57 paramat no prob :]