Time Nick Message 06:00 [MTMatrix] Krock: another feature that depends on USE_SDL=1 is toggling fullscreen mode at runtime via keyboard shortcut. probably not the most important thing though. 06:35 Mantar I can do that already with my window manager 08:45 [MTMatrix] celeron55: re https://forum.minetest.net/viewtopic.php?p=438143#p438143 "this is because it binds keys based on their position on the keyboard, instead of the symbol on the key": You know that we haven't switched keybindings to scancodes (key positions), right? They're still based on keycodes. SDL just doesn't give us keycodes that take modifiers into account. 08:51 [MTMatrix] It looks like your post caused follow-up confusion: https://forum.minetest.net/viewtopic.php?p=438191#p438191 08:52 [MTMatrix] > The key still works across layouts, it's just that the correct character is not displayed in the controls menu. 08:52 [MTMatrix] This is false. This would be the case if we had changed keybindings to use SDL scancodes, but had forgotten to show the actual layout-dependent key name in the controls menu. However, we haven't changed keybindings to use SDL scancodes. 08:57 sfan5 @luatic I think 14932 is simply a product of setRelativeMode not being available on !SDL 08:58 [MTMatrix] (replying here because I couldn't find my forum password) 08:58 sfan5 remember that we enabled it specifically for wayland #12636 08:58 ShadowBot https://github.com/minetest/minetest/issues/12636 -- Set and unset relative mouse mode where necessary by ndren 09:03 [MTMatrix] the issue only happens with SDL according to https://github.com/minetest/minetest/issues/14932#issuecomment-2276231213, so you're saying that relative mouse mode causes the problem? (?) 09:08 sfan5 yes 09:09 sfan5 of course could also be something that's broken with our use of relativ mouse mouse specifically 09:09 sfan5 but what I have observed if you have part of your display stack not handle relative mouse mode (e.g. vnc connection) that exact behavior OP is seeing happens 09:12 sfan5 merging #14937, #14929 in 5m 09:12 ShadowBot https://github.com/minetest/minetest/issues/14937 -- Revert macOS workflow to run on x86 by sfan5 09:12 ShadowBot https://github.com/minetest/minetest/issues/14929 -- Update credits for 5.9.0 by sfan5 09:14 sfan5 we're also not error-checking the result of SDL_SetRelativeMouseMode 09:14 sfan5 >If relative mode is not supported, this returns -1. 09:16 sfan5 however in principle nothing should happen if it doesn't work (except that we try to constantly re-enable it?) 12:46 [MTMatrix] this could possibly be the bug, considering the "Only change it when necessary, as it flushes mouse motion when enabled" comment 13:20 celeron55 @grorp well... assuming the release will be done now with the irrlicht backend, i'll not bother fixing the messages 13:22 celeron55 (at least i managed to make a number of users post their thoughts which is good enough) 13:22 MTDiscord Maybe its time to discuss how to merge https://github.com/minetest/minetest/pull/11843 - for example only type change without any actual big worlds support? - or full version vith compatible net BUT disabled 32bit by default ? or will wait few years more ? 13:27 MTDiscord I know OpenTTD uses a system of choosing the size at the beginning of a new game. Maybe that would be helpful for Minetest : most tests worlds could be on 8 bits, while gaming worls would be sized depending on the gamestyle 14:07 sfan5 does someone plan to prepare a PR to disble sdl by default? 14:25 [MTMatrix] uhm, the survey is going in the other direction 14:28 rubenwardy I think the survey is fundamentally flawed due to the sample 14:54 rubenwardy the disadvantages don't effect those on QWERTY english keyboards. Technical users will find the disadvantages easy to work around and the advantages more appealing 16:00 MTDiscord The +/- keys affect people with QWERTY english keyboards, I thought. 16:00 MTDiscord They don't affect me because I use Dvorak. 🙂 16:14 pgimeno I think it's vitally important that the keypad "/" works in all situations - I hope that's common to all proposed solutions 16:15 [MTMatrix] Yeah, it doesn't require shift 16:16 pgimeno yeah, my concern is that something like supporting shift + 7 makes the implementor forget about ensuring that support for keypad "/" still works 16:16 [MTMatrix] Another question should be: why players are expected to tweak fog, but that's gonna result in another long conversation 16:16 [MTMatrix] *fog distance 16:18 [MTMatrix] Not a core dev but I think that the regression is justified. It's not like we're top quality UX right now 16:19 [MTMatrix] PRs have reached almost the same levels we had a couple of years ago, so whatever the solution I think we should act fast 16:20 [MTMatrix] Also: are sfence and cx384 core devs now? I didn't get it, sorry 16:25 pgimeno @Zughy https://irc.minetest.net/minetest-dev/2024-07-29#i_6188880 https://irc.minetest.net/minetest-dev/2024-08-04#i_6189618 https://irc.minetest.net/minetest-dev/2024-08-04#i_6189607 https://irc.minetest.net/minetest-dev/2024-08-07#i_6190732 16:25 pgimeno may have been during Matrix bridge downtime 16:27 Desour welcome to both! 16:28 rubenwardy Adjusting draw distance is super important for managing performance 16:29 rubenwardy (Minetest used to support doing this automatically but it reducded performance and caused jitter) 18:43 MTDiscord We expect users to control view distance, but I agree fog* distance doesn't make much sense. It should just be latched to view distance in some standard way, and then fog can be toggle-able still. 19:05 sfan5 considering fog obscures the view what's the logical difference between view and fog distance? 19:06 MTDiscord Tbh, fog can be used for effects by the game 19:06 MTDiscord To intentionally obscure the view 19:07 MTDiscord What's the fog gradient along the view frustum is more important than where's the end of it exactly 19:07 celeron55 well, obviously it's dual purpose. ideally the game should be able to let the engine know what the artistic fog distance should be, and then from somewhere (currently from the user) should come a render distance value which depends on the performance of the hardware (or how much CPU and GPU the user wants to use for rendering) 19:07 MTDiscord Although both matter 19:09 MTDiscord Tbh it should be that of what's game-chosen and what's user-chosen, the smaller value is picked 19:09 celeron55 yeah the fog has many parameters, for artistic fog you can do many things, but for obscuring the rendering limit you want relatively "short" fog 19:09 celeron55 i suppose all these concepts and the smartness to decide what to do in the end should be added to the engine 19:11 Desour we already have a fog api 19:11 celeron55 there's at least this issue https://github.com/minetest/minetest/issues/14176 20:06 rubenwardy there's also volumetric fog 20:53 rubenwardy do the VM images no longer come with gl headers? 20:53 rubenwardy #14944 20:53 ShadowBot https://github.com/minetest/minetest/issues/14944 -- Disable SDL2 for 5.9.0 by rubenwardy 21:03 sfan5 which vm? 21:04 rubenwardy nevermind, fixed it - updated the windows.yml workflow vcpkg list 22:40 rubenwardy if decided, I suggest merging #14944 soonish and doing rc2 22:40 ShadowBot https://github.com/minetest/minetest/issues/14944 -- Disable SDL2 for 5.9.0 by rubenwardy 22:40 rubenwardy then release in a week 22:40 rubenwardy well depends how confident people are that non-sdl still work 22:40 rubenwardy +s