Time Nick Message 00:48 MTDiscord not a core dev, but it does make sense to dehard code it and move it to lua 08:34 sfan5 @grorp i was thinking 15s or so 09:43 rubenwardy Might be safer to make the notification dismissable and then only reshow if it doesn't exist every X seconds 09:46 rubenwardy I remember Android introducing anti spam measures although that may not be a problem here 09:47 rubenwardy Maybe thinking of this, which just applies to sound https://www.androidauthority.com/android-15-notification-cooldown-great-3537037/ 09:49 rubenwardy So nevermind 10:20 [MatrxMT] Merging #15863 in 30m; doing this a bit earlier than usual in case CI wants to rebuild Windows dependencies 10:20 ShadowBot https://github.com/luanti-org/luanti/issues/15863 -- Add vcpkg.json by y5nw 13:14 [MatrxMT] sfan5: sure. now extended to 15s. 17:21 [MatrxMT] meeting in 40m 17:57 Krock Meeting in 3 minutes. Topic: feature freeze. Meanwhile, there's plenty of open features where we would need to decide their fate: https://github.com/luanti-org/luanti/milestone/26 18:00 [MatrxMT] o/ 18:01 [MatrxMT] Merging #16046, #16025, #15791 in 15m 18:01 ShadowBot https://github.com/luanti-org/luanti/issues/16046 -- Show SDL version in the About tab by y5nw 18:01 ShadowBot https://github.com/luanti-org/luanti/issues/16025 -- SDL: Send events for X1 and X2 mouse buttons by y5nw 18:01 ShadowBot https://github.com/luanti-org/luanti/issues/15791 -- Move keybinding settings to (Lua-based) setting menu by y5nw 18:02 Krock nice. 18:05 Krock I don't know who to ping. Nobody else reacted to the meeting. Anyone else present? 18:05 MTDiscord somewhat 18:05 Desour yes, hi p/ 18:05 Desour o/* 18:05 Krock perfect, so let's get started with the feature PRs minus the ones mentioned above by y5nw 18:06 Krock > #16042 18:06 ShadowBot https://github.com/luanti-org/luanti/issues/16042 -- Fix scrollbar on ContentDB grid by adding an area label by grorp 18:07 Krock already has an approval and could be merged in theory, unless someone would like to review it too. 18:07 pgimeno @y5nw could you hold 15791 a bit? got a related question 18:07 [MatrxMT] pgimeno: Sure; what question? 18:07 MTDiscord 16042 should go in imo. 18:08 Krock One remark: The 'bool has_size = parts.size() >= 3;' will make it impossible to add new arguments to the short label notation but considering that label[] hasn't seen a need for change in years 18:08 Krock I don't think this is a concern. 18:08 pgimeno @y5nw it seems impolite to ask in the middle of the meeting, I better ask later (and I can't PM you since you're not in IRC) 18:09 Krock pgimeno: we can discuss this right now if you'd like that. 18:09 Krock that's on the milestone too, so it fits. 18:10 pgimeno thanks, well, it's about whether it will affect the possibility of having two keybindings for the same action 18:10 MTDiscord merging #15872, #16045 in a couple mins so they're off the big PR pile 18:10 ShadowBot https://github.com/luanti-org/luanti/issues/15872 -- Make it more convenient to implement item/player/entity abilities that modify node drops by j-r 18:10 ShadowBot https://github.com/luanti-org/luanti/issues/16045 -- Cleanup miscellanea (don't squash) by appgurueu 18:11 [MatrxMT] pgimeno: That is up to a later PR: #15934 18:11 ShadowBot https://github.com/luanti-org/luanti/issues/15934 -- Implement secondary keybindings by y5nw 18:12 [MatrxMT] (Which I will need to rebase soon, but I don't think it will fit into the 5.12 cycle) 18:13 pgimeno @y5nw ok, sounds great! then 15791 greenlit from me :) 18:15 Krock perfect :) 18:15 Krock > #16039 18:15 ShadowBot https://github.com/luanti-org/luanti/issues/16039 -- Do SDL default on macOS by sfence 18:15 MTDiscord isn't a real feature 18:15 MTDiscord it's the feature / bugfix combo, we can do this later 18:16 Krock I see. The test reports are also positive so I guess we could just merge it soon. 18:16 MTDiscord yep 18:16 Krock > #16034 18:16 ShadowBot https://github.com/luanti-org/luanti/issues/16034 -- Handle texture filtering sanely to avoid blurriness by sfan5 18:16 MTDiscord similar story, and is already reviewed and approved; unproblematic 18:17 Krock I remembered this PR being larger but it seems to be doable for 5.12.0 18:17 Krock > #16025 18:17 ShadowBot https://github.com/luanti-org/luanti/issues/16025 -- SDL: Send events for X1 and X2 mouse buttons by y5nw 18:17 Krock merge pending 18:17 Krock > #15979 18:17 ShadowBot https://github.com/luanti-org/luanti/issues/15979 -- Add `inventory_image_animation` and `wield_image_animation` by cx384 18:18 Krock I still have a review pending there and haven't yet finished it yet. 18:18 MTDiscord yeah, same 18:19 Krock Also rather large; I'd prefer to remove the milestone from that one. It's definitely a feature and needs more reviewing. 18:19 Krock or rather move to 5.13.0 18:19 Krock objections? 18:20 [MatrxMT] as a modder, I'm fine with delaying it 18:21 MTDiscord realistically i don't think i'll be able to give it a thorough review before the FF starts 18:21 Krock sure. we'll enter feature freeze as soon there's no more true feature PRs in the milestone. Moved to 5.13.0 18:22 Krock next would be 15872 -> pending merge 18:22 Krock > #15055 18:22 ShadowBot https://github.com/luanti-org/luanti/issues/15055 -- Clip posteffect color to respective nodes by Desour 18:23 Desour doesn't look like anyone reviewed it, so just move it out of milestone :shrug: 18:23 Krock It's an interestingly large PR for what it achieves at the end. I've also got it on my TODO list for "better times (TM)" 18:23 MTDiscord same story (sorry, it's a cool PR implementation-wise) 18:24 Krock Moved to 5.13.0 for now 18:24 Krock > #14135 18:24 ShadowBot https://github.com/luanti-org/luanti/issues/14135 -- Add persistent unique identifiers for objects by sfence 18:24 [MatrxMT] THAT would be great 18:25 Krock open reviews and changes requested. It's unlikely it'll make it into 5.12.0 18:25 MTDiscord working on it right now 18:25 MTDiscord i think the requested changes are relatively simple in nature, well most of them 18:25 [MatrxMT] <3 18:25 Krock > Doesn't this require a mapblock version bump? 18:25 MTDiscord this one will require a bit more care yes 18:26 Desour (was open ended question) 18:26 Krock it it means what I think then it means that saved worlds will receive a bump too 18:26 Desour idk if when the doc says (=1) it means that the value version value is guaranteed to be 1 or just max 1 in the doc 18:26 Krock which means that by-file shared world cannot be opened any more in slightly older versions. 18:27 MTDiscord i mean i'm somewhat optimistic that there isn't really a problem due to the fact that we're just appending stuff and i think we're generally lax about this 18:27 MTDiscord but i'll test and look at the code to be sure 18:27 Krock oh nvm I introduced this variable such that it's kinda forwards-compatible. the data is simply discarded. 18:28 Krock maybe there used to be a hard check on version 1 in older Minetest versions 18:29 Desour ok, looks like no bump needed (looked at code) 18:30 Krock that's for version 5.0.0 compat 18:30 Krock https://github.com/luanti-org/luanti/commit/67049eba 18:30 Krock strict (version == 1) check, which defeated the purpose of versioning 18:32 Krock > #13125 18:32 ShadowBot https://github.com/luanti-org/luanti/issues/13125 -- Android: Persistent notification while ingame by srifqi 18:32 Desour (well, not completely defeated it. the new one makes old a prefix of new. but in the old you did merely get wrong entities, it looks like, but the rest of the block is fine) 18:32 Krock I could give this a try on Android 6.1 if that's desired 18:33 Krock Desour: yes. now we have a version for breaking changes and one for backwards-compatible ones ¯\\_ (ツ)_/¯ 18:35 Krock alright. end of PRs. The remaining ones should be doable in the next few days to enter feature freeze soon 18:35 Krock Are there any comments about specific issues/blockers in particular? Is something missing from the milestone? 18:37 Desour do we now use sdl by default in the next release? 18:39 Krock It used to be reverted due to #14545 18:39 ShadowBot https://github.com/luanti-org/luanti/issues/14545 -- SDL: Some keybinds broken due to missing character lookup 18:39 Krock Now the issue affects the Linux Irrlicht Device too, maybe others as well 18:39 Krock so going by that logic we should not ship and IrrDevice any more because all of them have this issue 18:40 Krock (or rather: it's the key handling in Luanti that changed towards SDL's favour) 18:40 [MatrxMT] I think non-SDL IrrlichtDevices are only partially affected (keybindings in shifted state can only be specified manually in minetest.conf) 18:42 Krock I see. Thanks for the correction. 18:43 Krock so the answer to the "Use SDL by default" question would then likely be "yes". The feature freeze will tell us whether the players are OK with that. 18:45 [MatrxMT] We still need a mainmenu dialog to inform users about the change, but I have started working on it 18:45 pgimeno I've been using it and I'm perfectly OK. The only trouble was that I used to use the slash in the keypad due to the conflict with the sneak binding, but I remapped it to that key. 18:46 [MatrxMT] I was wondering if any dev has a take on #15010 . Anything I can try, I don't know 18:46 ShadowBot https://github.com/luanti-org/luanti/issues/15010 -- OS doesn't receive keys captured by Minetest [SDL] 18:46 [MatrxMT] Since it's in the milestone 18:47 MTDiscord okay Desour, i should have addressed everything now. 18:49 [MatrxMT] Zughy: What version of SDL are you using? (use e.g. `sdl2-config --version` if you don't want to recompile latest master; asking since others don't seem to be able to reproduce it) 18:49 MTDiscord did the two merges i announced earlier 18:52 [MatrxMT] 2.0.20 18:53 [MatrxMT] Wow that's old 18:53 [MatrxMT] let's see if it's Debian's fault 18:53 MTDiscord if in doubt, blame debian (/s) 18:54 [MatrxMT] fastest way to update it? 18:54 [MatrxMT] I mean, they do at least have nice testing coverage 18:55 pgimeno my bookworm is in 2.26.5 18:55 [MatrxMT] Like, they even test locale files; just don't ask why certain things stand out: https://i18n.debian.org/l10n-pkg-status/l/luanti.html 18:56 Krock Offtopic: would it be possible to randomize the internal Android version number during development builds? I "upgrade" the app if the versions match. 18:58 [MatrxMT] if I try to update it (`sudo apt install libsdl2-dev`), it tells me I'm already on the latest version (2.0.20+dfsg-2ubuntu1.22.04.1) 18:58 pgimeno @Zughy Bookworm? 18:59 [MatrxMT] Pop!_OS 22.04 19:00 pgimeno well, oldstable is at 2.0.14 19:02 pgimeno @Zughy my best bet is to install it from source if your distro doesn't provide it, which is a bit of a PITA but I don't know of any good alternatives 19:02 [MatrxMT] the question is, are we sure it's SDL's fault? 19:02 [MatrxMT] Before I even try 19:03 [MatrxMT] Idk, but it seems likely given that nobody else is able to reproduce it 19:11 MTDiscord different question: why is pop os still on 22.04? 19:11 MTDiscord shouldn't it be 24.04 by now (i checked the website, that's not a thing; why not?) 19:14 [MatrxMT] no idea luatic, but it's a great question 19:15 [MatrxMT] @pgimeno: that was the smoothest installation from source I've seen in my whole life, not gonna lie 19:15 pgimeno wow, great 19:16 [MatrxMT] now I need to understand why it gets 2.0.20 instead of the latest one 🙃 19:16 [MatrxMT] I mean Luanti 19:18 pgimeno maybe you may need to specify SDL2_INCLUDE_DIR and SDL2_LIBS or something similar? 19:19 MTDiscord <.niklp> > why is pop os still on 22.04? afaik they'll release 24.04 once cosmic is ready 19:21 rubenwardy Krock: there's a callback for app installation, that could be used in debug mode to invalidate the install 19:22 rubenwardy Or the .zip could have a precomputed hash that is compared and stored 19:22 rubenwardy Computed at build time 19:22 Krock rubenwardy: false alert actually. Something must've changed in recent Luanti versions. I can now upgrade same versions without an issue. 19:22 Krock I noticed this while switching between two 5.10.0 or 5.11.0 development builds and couldn't get it to install 19:23 pgimeno an OS which lags behind Debian, that's something to behold 19:30 Krock Let's say this is the end of the regular meeting. Thanks for participating. Those who are still here can surely continue ;) 19:37 sfan5 did I miss anything? 19:37 [MatrxMT] thank you Krock for your work 19:39 Krock sfan5: I added a an easter egg into our code: a fork bomb. Aside from that nothing new 19:39 Krock jokes aside - just went through the feature PRs to see what makes it into 5.12.0 for a feature freeze soon (TM) 19:40 [MatrxMT] funnily enough, today is Easter 19:40 Krock thanks, that's the joke. 19:40 [MatrxMT] :( 19:41 [MatrxMT] New PR for the milestone: #16049 19:41 ShadowBot https://github.com/luanti-org/luanti/issues/16049 -- Migrate existing keycode-based keybindings; add warning dialog about migration by y5nw 19:42 [MatrxMT] y5nw: on SDL 2.33, nothing changes. Well, actualy now if I hit the screen key it freezes for not even half of a second, but it still doesn't launch the OS built-in screen capture tool 19:48 [MatrxMT] Zughy: I'm out of ideas then 22:34 pgimeno why is core.serialize() serializing functions? does that have a purpose? 22:44 MTDiscord someone thought it was a cool feature a long time ago 22:44 MTDiscord just don't pass data structures containing functions to it 22:45 MTDiscord if you want to be extra-cautious write yourself a simple recursive function that checks whether there are any functions in a table structure 22:46 pgimeno I was trying to use it as a quick inspection tool for nodes using //lua core.chat_send_all(core.serialize(core.registered_items["node name"])) 22:47 pgimeno however, since it dumps functions as binary blobs, it errs with "invalid UTF-8" on nodes which have an on_place or the like 22:48 MTDiscord ah well the problem is that it isn't really intended for inspection purposes 22:48 pgimeno since Lua bytecode, LuaJIT 2.0 bytecode and LuaJIT 2.1 GC64 bytecode are mutually incompatible, I wondered what purpose that served 22:48 MTDiscord for that we have dump / dump2 22:49 MTDiscord if you assume that you're gonna stay on the same lua impl, it's maybe mildly useful. but as said, it's not really a thing (it's not documented for a reason), it's just something we sorta continue to support. 22:49 MTDiscord but we should probably start deprecating it. 22:50 MTDiscord btw for a nice pretty printer, i wrote https://github.com/appgurueu/dbg/blob/master/src/pp.lua 22:50 MTDiscord should maybe look into making dump nicer