Time Nick Message 16:29 sfan5 rubenwardy: ping regarding android 17:14 rubenwardy #11703 17:14 ShadowBot https://github.com/minetest/minetest/issues/11703 -- Backport android fixes by rubenwardy 17:14 rubenwardy how stable are the arm netcode fixes, sfan5? 17:14 rubenwardy might be worth including those given that Android is particularly affected 17:15 sfan5 at this point: very, there is zero behavioural change 17:15 sfan5 btw you'll need to edit build.gradle too to download an older version of https://github.com/minetest/minetest_android_deps_binaries 17:16 rubenwardy 342eb18b7512462585a33bc9eef0b68298087151 ? 17:17 rubenwardy which is before any Irrlicht 1.9.0 17:17 rubenwardy https://github.com/minetest/minetest_android_deps_binaries/commits/master 17:17 rubenwardy Revert "Android: Irrlicht update (Add Alpha fix)" 17:17 sfan5 sure 17:18 sfan5 the networking fixes are fd8a8501b + fe5cb2cdf 17:22 rubenwardy argh 17:22 rubenwardy it was using an outdated version of deps still, so I deleted it 17:22 rubenwardy now I can't build or clean because it just errors 17:23 rubenwardy I'll try manually running android[downloadDeps] 17:25 rubenwardy back to the same error 17:25 rubenwardy src/client/content_cao.cpp:23:10: fatal error: 'ITextSceneNode.h' file not found 17:25 sfan5 irrlicht still too new 17:25 sfan5 this file was only removed in irrlichtmt 17:28 rubenwardy I've changed the build.gradle to src 'https://github.com/minetest/' + folder + '/archive/342eb18b7512462585a33bc9eef0b68298087151.zip' 17:29 rubenwardy it's probably caching android/native/build/deps.zip 17:43 sfan5 doesn't it download deps.zip to somewhere in android/build ? 18:11 rubenwardy yeah, the path I gave 18:11 rubenwardy I needed to delete that and adjust another thing 18:11 rubenwardy it builds locally but not on CI 18:11 rubenwardy well, Android builds on CI 18:11 rubenwardy clang/gcc builds fail probably due to an external factor 18:12 rubenwardy postgresql-server-dev-all : Depends: postgresql-common (>= 117~) but it is not going to be installed // Depends: postgresql-server-dev-10 but it is not going to be installed 18:12 rubenwardy I haven't been watching commits, I imagine this was fixed in master? 18:13 sfan5 yes 19:23 nrz pg 10 ? which os do you run, we are in 14 :p 19:26 rubenwardy CI 19:26 rubenwardy not me 19:27 nrz it should be fixed, pg10 is quite old :p 19:27 nrz 5 years old 19:32 rubenwardy ContentDB still uses pg9, lol 19:32 nrz there is multiple pg optimisations since yuears, upgrade ! 19:33 rubenwardy I believe that updating is mildly to severely painful 19:33 nrz pg_dumpall > backup.sql 19:33 nrz insall new engine 19:33 nrz cat backup.sql | psql 19:34 nrz both as postgres user 19:34 nrz easy 19:34 kilbith sfan5: I'm tired of this shit on Windows -> https://i.imgur.com/QniQcQi.png 19:34 kilbith give me the DPI in get_player_information() 19:34 kilbith v-rob: this is addressed to you too 19:35 kilbith currently we can't do universally well designed GUIs without the DPI in get_player_information() 19:36 sfan5 if you had dpi you'd make certain elements smaller/bigger, right? 19:36 kilbith yes 19:37 sfan5 I was thinking about dpi-adjusted window size for a while but I think that'd be worthless here 19:37 sfan5 so in the end exposing the DPI will be necessary 19:37 kilbith please yes 19:38 sfan5 would you have any use for touchscreen or not-touchscreen? 19:38 kilbith personally not 19:39 sfan5 I see 19:40 sfan5 the dpi proposal has been shot down before but I'm sure other devs also see the necessity, I'll open an issue as reminder 19:40 kilbith https://github.com/minetest/minetest/issues/10632 19:40 kilbith the other devs don't do formspec mods like i3 tbh 19:41 sfan5 hm 19:41 sfan5 regarding style[] does it not have relative font sizes? 19:41 kilbith give me a way to know if some player uses linux or windows 19:41 MTDiscord could core.get_screen_info() from the main menu be exposed to modders? 19:41 sfan5 what for? 19:41 kilbith sfan5: I've tried that but eventually reverted to absolute sizes 19:42 sfan5 @Jonathon no, the mainmenu code is totally different 19:42 kilbith sfan5: https://github.com/minetest/minetest/issues/10632#issuecomment-767031245 19:42 kilbith relative font sizes have problems too on different platforms 19:43 sfan5 user font size might a problem, other than that fonts absolutely should look identical everywhere 19:44 kilbith Windows 10/11 are displaying fonts bigger than Linux for the same size 19:45 kilbith because the DPI is different 19:45 sfan5 no surprise then 19:46 sfan5 sounds like absolute sizing in formspecs is not absolute enough though 19:47 sfan5 ¯\_(ツ)_/¯ 19:48 kilbith please give me a solution because I don't want to dual-boot on Linux just to keep modding on Minetest 19:49 sfan5 if you know the DPI isn't that the solution? 19:49 sfan5 also if you build from source you can change the DPI to 96 (there ought to the a setting for that) 19:49 kilbith how do I know the DPI? 19:49 sfan5 hypothetically, I mean the API you're suggesting 19:49 kilbith I've changed the setting to 96 and it doesn't change anything 19:51 sfan5 ah there's display_density_factor now 19:52 sfan5 so if you were to set that to 96/yourdpi you should it looking the "normal" way 19:52 sfan5 +get 19:56 kilbith so it's normally 0.66 19:56 kilbith and it doesn't affect the font size at all 19:57 sfan5 the font scales as expected here 19:57 kilbith even though if it worked I'd prefer make it look good to the users out of the box without telling them to go to the settings 19:58 kilbith the font scales as expected here 19:58 kilbith no it doesn't for me, it only affect the element sizes 19:58 kilbith not the font 19:58 sfan5 I assume you're testing in i3? I only checked the mainmenu 19:58 kilbith yes 19:59 kilbith the mainmenu stays intact as well between 0.66 and 1.0 20:00 sfan5 you must have a font size set in the config 20:02 kilbith removed it and still the same 20:02 kilbith comparison: https://imgur.com/a/Y3qgfyD 20:02 kilbith it's the exact same 20:03 sfan5 you restarted MT, right? 20:03 kilbith ah 20:03 kilbith now it changed 20:04 v-rob Heh, those restart-to-change settings really should be labeled as such :) 20:08 kilbith I don't wanna instruct my Windows users to go deep down into the advanced settings and mind to restart their clients 20:08 kilbith it must be looking good by default and that's why I need this information in get_player_info 20:08 sfan5 I am already convinced that DPI info is useful in case you're still doing that 20:08 kilbith I'm not 21:09 kilbith v-rob: where are you with the formspec rewrite? 21:15 v-rob I don't entirely remember, I haven't had time to do serious work on Minetest for some time, although I've still been hanging around. Stuff draws, and I was busy fleshing out sizers, as I recall 21:17 kilbith man I thought you and hecks were going to decouple MT from Irrlicht in November 22:13 MTDiscord Giving modders and main menu Lua access to DPI and screen size (in pixels) is the holy grail in my mind, maybe OS if that somehow affect fonts. Just thinking that with those things, absolute positioning is possible, as well as the ability to send different formspecs for different screen sizes. 22:17 MTDiscord main menu already ahs access to it 22:17 MTDiscord *has