Time Nick Message 12:36 sfan5 thought: should we formally drop support for LuaJIT older than 2021? we already have cmake print a warning if you use an older version since a few years. while I'm not aware of problems for typical windows/linux x86_64 usage I think there's really no excuse for anyone to ship ancient, potentially broken stuff 12:36 sfan5 (this would not exclude openresty's luajit to be clear) 12:36 sfan5 (warning in question: https://github.com/luanti-org/luanti/commit/af38bae57f4f0f77bf2c0cf029928567e554b60b) 12:37 rubenwardy Agreed 13:07 red-001 why not just pull in LuaJIT as a submodule? plenty of distros still ship openresty 13:08 sfan5 distros will unvendor deps 13:09 red-001 if Luanti is going to drop support for system LuaJIT it would seem to make sense to integrate it as a submodule so people don't have to hunt down whatever version they need separately, I'm sure distros might break build systems but distros will also ship a 3 year old version of Lunati, more so thinking about how to make things easy to anyone on a mainline version of Debian or Ubuntu 13:10 sfan5 I did not suggest that??? 13:10 red-001 "this would not exclude openresty's luajit to be clear" maybe I misunderstood that 13:10 sfan5 oh 13:10 sfan5 "not exclude" as in it would pass the check and you can still use it 13:11 red-001 ah okay I'm sorry I thought you meant that as in openresty would still be blocked since it's forked off an old version 13:11 red-001 my bad 13:11 red-001 yeah no, no complaints then, this seems reasonable 13:13 sfan5 only issue is currently the "old luajit" check is failing in msvc CI - false positive? not looking forward to fixing that 13:15 red-001 does cmake print out the failed compilation/symbol check output? might not be that bad to fix 13:15 red-001 as in the compiler/linker/whatever it runs to do the check 13:15 sfan5 it dumps a log file somewhere 13:17 red-001 I guess I could try and repro this on local windows, but with the luck with CI recently it might pass the check perfectly there, I guess need to really pull that file from CI somehow 13:20 sfan5 much more important question: have we decided which two-letter-acronym is right for the new name? old: MT, new: LT? LU? LN? LI? 13:21 MTDiscord obviously LT 13:22 sfan5 🇱🇹 13:22 MTDiscord if you split Luanti into syllables, you get Luan Ti, which is also a real name apparently, so LT it must be 13:29 rubenwardy We should give out LuanTea at the next fosdem 13:29 rubenwardy Custom printed teabags 13:30 rubenwardy Hopefully Yorkshire tea 13:38 MTDiscord while wearing LuanTies 13:39 Sheriff_U3 rubenwardy: don't forget the voxel tea cup to go with your tea. 13:41 red-001 TI 13:44 MTDiscord what we kind of lack is a cute mascot 13:44 Sheriff_U3 Sam 13:44 rubenwardy celeron55 in cat ears 13:45 MTDiscord Sheriff_U3: pssh everyone knows Sam is a recursive acronym for Sam ain't mascot 13:45 Sheriff_U3 I didn't... 13:45 pgimeno Mr Green? 13:46 MTDiscord Sheriff_U3: well it isn't, i was joking. it has sometimes been said that Sam stands for Sam ain't Minecraft though. but that doesn't stop Sam from giving off major MC vibes. 13:47 MTDiscord pgimeno: Dev Test is not exactly the cutest candidate out there :p 13:48 sfan5 who could say no to this guy https://x0.at/MkfT.png 13:48 sfan5 (girl?) 13:48 MTDiscord gender: wizard 13:53 [MatrxMT] is that a mouth or the face 13:53 rubenwardy Yes 13:54 [MatrxMT] BREAKING: the black patch in `player.png` is a Yes [LORE-ACCURATE] 13:57 red-001 person of unknown gender suspected of mischief against the realm 13:58 [MatrxMT] 3 counts of privilege escalation and 4 counts of abusing //lua 14:01 [MatrxMT] +1 for c55 w/ cat ears as a mascot 14:02 [MatrxMT] a talking luanti logo like it's ripped from an ad featuring a talking product 14:08 user333_ i definitely support the idea of having a mascot but don't have any ideas of what it should be 15:07 MTDiscord <.zenonseth> since Luanti is a game all about building, how about a beaver? like Lutti the Beaver? 15:33 red-001 for #11832 sfan I think it might just be a bit obscure still, but passing the already open socket to the server seems like a nice fix, at a slight cost of the bind call not being done on another thread 15:33 ShadowBot https://github.com/luanti-org/luanti/issues/11832 -- quickly restarting singleplayer on windows sometimes fails to bind port 15:35 red-001 but that might be for the better, should make feedback faster for users hosting a game server from the local tab if they pick a port that's in use, and the server startup cost should be trivial compared to everything else happening 17:35 red-001 not happy #17512 adds code overall, was really hoping could be a net negative change in code size 17:35 ShadowBot https://github.com/luanti-org/luanti/issues/17512 -- Bind to socket before starting server by red-001 17:36 red-001 maybe if I add a constructor to UDPSocket that binds to an address 18:01 red-001 +115, -112 is better I suppose 18:15 red-001 there +106,-112 18:33 red-001 +104,-120 ready for review now I suppose 19:28 red-001 someone wanna merge #17481 and see what happens? 19:28 ShadowBot https://github.com/luanti-org/luanti/issues/17481 -- [CI] Make VCPKG ABI hash more stable by red-001 19:29 red-001 (you will need to manually delete the caches as well if you do) 20:05 MTDiscord merging #17504 #17496 #17481 #17478 #17355 #17305 in 10m 20:05 ShadowBot https://github.com/luanti-org/luanti/issues/17504 -- Fix autoforward regression with get_player_control().up by y5nw 20:05 ShadowBot https://github.com/luanti-org/luanti/issues/17496 -- [no-sq] Reserve RDIX entries used by the engine + cleanup function dumping by red-001 20:05 ShadowBot https://github.com/luanti-org/luanti/issues/17481 -- [CI] Make VCPKG ABI hash more stable by red-001 20:05 ShadowBot https://github.com/luanti-org/luanti/issues/17478 -- Make media progress text position depend on bar image size by cx384 20:05 ShadowBot https://github.com/luanti-org/luanti/issues/17355 -- Fix AbsolutePath returning non identical strings on semantically identical paths by DeltachangeOG 20:05 ShadowBot https://github.com/luanti-org/luanti/issues/17305 -- Add `listimages[]` element to allow styling inventory slots with textures by CrazyladMT 20:09 crazylad \o/ 20:14 sfan5 feel free to add #17503 20:14 ShadowBot https://github.com/luanti-org/luanti/issues/17503 -- [no squash] Warn if object with duplicate GUID + more changes by sfan5 20:18 MTDiscord sure 20:22 sfan5 thx 20:41 red-001 #17512 seems to work, not sure if anyone else had the issue other than me, I don't know if it would happen on a non hypervisor setup wsor4035 are you still on Windows? 20:41 ShadowBot https://github.com/luanti-org/luanti/issues/17512 -- Bind to socket before starting server by red-001 20:42 MTDiscord i have a windows machine 20:42 red-001 does that bug still happen on it? 20:42 red-001 if not you might want to enable hyperv when testing 20:42 MTDiscord yes 20:42 red-001 I would appreciate if you could test if the PR fixes it then 20:42 red-001 whenever you get a second 20:42 red-001 it works on my machine 20:43 MTDiscord ill look tonight 21:13 red-001 can someone delete https://github.com/luanti-org/luanti/actions/caches a push a commit to master the touches the Windows,yml file? 21:13 red-001 (in that order) 21:14 red-001 probably on me a little bit, I should have changed the cache prefix 21:36 sfan5 i can just re-trigger the last run? 21:39 red-001 yeah if you delete the caches first I think 21:40 red-001 sfan5: should work if the caches are deleted before doing that, forgot you could retrigger runs 21:40 sfan5 ok