Minetest logo

IRC log for #minetest-dev, 2023-04-17

| Channels | #minetest-dev index | Today | | Google Search | Plaintext

All times shown according to UTC.

Time Nick Message
00:30 Alias joined #minetest-dev
00:50 jeremy_soller[m] joined #minetest-dev
01:35 tekakutli joined #minetest-dev
01:46 Soni joined #minetest-dev
04:00 MTDiscord joined #minetest-dev
05:43 tekakutli joined #minetest-dev
06:02 calcul0n joined #minetest-dev
06:14 MTDiscord <luatic> sfan5: maybe because copy-paste on Discord gives you a mess when copying usernames? x)
06:39 Alias joined #minetest-dev
08:02 sfan5 does it?
08:21 proller joined #minetest-dev
08:41 proller joined #minetest-dev
09:37 Zughy[m] about the roadmap (https://github.com/minetest/minetest/blob/master/doc/direction.md), can't we remove the first point about rendering/graphics improvement? I mean, nobody has really asked for it (not even core devs) in the roadmap brainstorm issue https://github.com/minetest/minetest/issues/12746#issuecomment-1315569363
09:37 Zughy[m] This doesn't mean of course that x2048 and lhofanshl can't keep doing what they're doing
09:38 Zughy[m] *lhofhansl, sorry
09:38 rubenwardy no, we cannot - it needs to be in the roadmap as there's still more to do
09:39 rubenwardy people aren't asking for it because it's in-progress and so they assume it will continue to happen
09:39 rubenwardy the point of the roadmap is to community what sort of things are the priority to work on, rendering/graphics is definitely high on that list
09:39 rubenwardy *communicate
09:43 rubenwardy also that goal includes SDL stuff, which is likely to be a focus for 5.8
09:46 nrz tbh what i'd like to see is SDL replacing legacy model, and in 5.9 removal of legacy backends if we are confident. and also the new main menu UI (maybe with some refinements, but it's a first very very good change), and entity enhancements. I'd like to take time to improve this entity things, as we discussed before.
09:46 nrz another thing, maybe we can also enhance some voxel performance with offloading on GPU with OpenMP or OpenCL (the second won't work on Mac OSX)
09:47 rubenwardy I'd like to see SDL stabilised and used by default for 5.8. Would make sense to remove other backends at that point
09:47 nrz i think 5.8 is too early for removal, we should keep them one release, to see side effects and fix them before removing 😄
09:48 rubenwardy there's not really much benefit to that as using a different backend will require recompiling anyway, could be a patch release
09:49 rubenwardy we could keep it around to revert to non-sdl in a .1
09:50 nrz you mean we cannot have a config option to switch between SDL and non SDL ? if i remember correctly on irrlicht it's an option no ?
09:50 nrz if it's compile time, we can give end users a solution, or some platforms which break should use legacy and new ones only SDL
09:50 nrz i'm pretty confident about the SDL2 ones, as it's now a pretty old library
09:50 rubenwardy correct, it's a compile option currently
09:51 rubenwardy USE_SDL2 in Irrlicht
09:51 nrz ok right 🙂
09:52 rubenwardy the problem with making SDL2 optional is that existing distros won't change their configs, they'll continue to use non-SDL
09:53 nrz a question about roadmap, we create roadmap issue on the beginning of the previous cycle ? it's not better to do it during the RC cycle ?
09:54 nrz i mean open roadmap issue for X+1 when feature freeze for X is happening
09:54 rubenwardy no, roadmaps are longer term than a release
09:55 rubenwardy it's been 2 years since the last roadmap
09:56 nrz i mean, we can have long term roadmap, but middle term for next release ?
09:56 rubenwardy no
09:56 rubenwardy https://github.com/minetest/minetest/blob/master/doc/direction.md
09:56 rubenwardy the "middle term for next release" is a milestone
09:56 nrz fine 🙂
09:56 rubenwardy !title https://github.com/minetest/minetest/milestone/22
09:56 ShadowBot rubenwardy: 5.8.0 Milestone · GitHub
13:15 Desour joined #minetest-dev
13:22 Desour FYI, the SDL is not far from being stable. there's just one pair of PRs to be merged before all known issues are fixed.
13:24 Desour the rendering improvements on the roadmap probably also include things like using texture atlases or array textures, which goes in a completely different direction than the improvements made yet
13:34 MTDiscord <kimapr> when media unloading?
13:34 MTDiscord <kimapr> is that on the roadmap
13:36 Desour I'd count that to internal code refactor and improvements
13:53 MTDiscord <Warr1024> Media unloading is not just some "internal refactor", it dramatically affects the way we can use media in mods and games.  Right now we are very limited by the fact that everything eats memory and there's no way to get it back.  Without that limitation there are features I've held back that I would merge as soon as I EOS the old pre-unloading versions.
14:05 proller joined #minetest-dev
14:27 Fleckenstein joined #minetest-dev
14:53 pgimeno is dual wielding receiving top priority for this release? given it wasn't ready for the previous one
15:15 Fleckenstein well i kinda asked for review of my TODO/agenda and no one confirmed that this is the way to go yet
15:26 fluxionary joined #minetest-dev
16:24 rubenwardy will merge #12614 later unless anyone else wants to review it
16:24 ShadowBot https://github.com/minetest/minetest/issues/12614 -- Add ability to override item images using meta by rubenwardy
16:27 sfan5 dual wielding is on the high-priority pile but that is unfortunetely no guarantee
16:27 sfan5 unfortunately*
16:28 Zughy[m] I still think we should prioritise Jude's PRs before they become obsolete
16:45 tekakutli joined #minetest-dev
17:27 appguru joined #minetest-dev
17:57 diceLibrarian joined #minetest-dev
17:59 tekakutli joined #minetest-dev
18:17 rubenwardy merging #12614 in 10
18:17 ShadowBot https://github.com/minetest/minetest/issues/12614 -- Add ability to override item images using meta by rubenwardy
18:22 MTDiscord <luatic> rubenwardy: what happens in str_to_v3f if the string is not formatted properly?
18:23 rubenwardy same behaviour as in Settings:get_pos
18:25 rubenwardy good point though, I copied the implementation without thinking much about the specific behaviour. Should probably emit a warning and only set wield_scale if valid
18:50 Desour joined #minetest-dev
21:09 PrairieWind joined #minetest-dev
21:09 PrairieWind joined #minetest-dev
21:25 tekakutli joined #minetest-dev
22:03 LizzyFleck joined #minetest-dev
22:04 Fleckenstein joined #minetest-dev
22:27 tekakutl` joined #minetest-dev
22:41 tekakutli joined #minetest-dev
23:34 AliasAlreadyTake joined #minetest-dev

| Channels | #minetest-dev index | Today | | Google Search | Plaintext