Time Nick Message 01:16 MTDiscord Yes I think my phone might be using that to accelerate discord as I type this 01:19 rubenwardy Also rpi 02:52 MTDiscord Is there a way for this happening every ~2-5 minutes to cause a server (in this case CTF) to completely lag out? https://gcdnb.pbrd.co/images/5l3YXFzvAjP3.png?o=1 02:53 MTDiscord Screenshotted this particular peer id for luls 03:00 MTDiscord It's happening only once and still causing issues? 03:07 MTDiscord Once every 10 minutes, exactly, actually 03:08 MTDiscord Is there a misbehaving serverlist pinger out there that's not closing the sessions properly or something? 03:08 MTDiscord I'm trying to figure out how someone is attacking the server, from what they told me, which can't be fully trusted, it's from quitting immediately after joining the game, possibly with a modified client 03:08 MTDiscord Weirdly it looks like it IS trying to close the session properly... 03:10 MTDiscord By "completely lag out" do you mean the step size goes super high (like there's a lot of CPU usage or database contention or something) or do you mean it totally freezes up for a while (like the main thread itself is being blocked)? 03:11 MTDiscord I wonder if it would be useful to get a raw packet capture 03:12 MTDiscord I've only been around when it happened once. The IRC/Discord relay might have still worked, but the whole game froze up and players were unable to do anything 03:12 MTDiscord Did it self-resolve, or did it need to be killed externally? 03:13 MTDiscord It self-resolves if the attacker stops 03:30 MTDiscord It doesn't even trigger on_prejoinplayer 03:33 MTDiscord I see why two "Handling peer change" are shown, that's the join and leave being processed 04:46 MTDiscord This would resolve every CVE 06:58 MTDiscord If I overwrite ClientInterface::CreateClient() with https://pastebin.com/cSKTpL9a (code for getting address copied from src/network/serverpackethandler.cpp: Server::handleCommand_Init()) then I get error with each attack, for normal logins I get the peer address 06:59 MTDiscord Server::handlePeerChanges() calls CreateClient() for a change of type PEER_ADDED 09:52 sfan5 creating and deleting a peer immediately is not unusual, needs more info 16:02 sfan5 pushing http://sprunge.us/b3geiT?diff in a while 16:03 sfan5 (the instructions don't really make sense with a docker image in the registry and the relevant parts can be readded once we have one again) 16:03 sfan5 without a* 16:04 rubenwardy Make sense 20:46 sfan5 merging #14206, #14195, #14142, #14109, #14184 in 10m 20:46 ShadowBot https://github.com/minetest/minetest/issues/14206 -- Don't apply gui_scaling & DPI twice to table[] / textlist[] scrollbar by grorp 20:46 ShadowBot https://github.com/minetest/minetest/issues/14195 -- Use AL_SOFT_direct_channels_remix extension for non-positional stereo sounds by Desour 20:46 ShadowBot https://github.com/minetest/minetest/issues/14142 -- Legible Lua profiler by fluxionary 20:46 ShadowBot https://github.com/minetest/minetest/issues/14109 -- [NOSQUASH] Fix tonemapping and apply saturation even if tonemapping is disabled by rollerozxa 20:46 ShadowBot https://github.com/minetest/minetest/issues/14184 -- Replace clientmap's MeshBufListList with a hashmap by Desour 20:48 Desour 14195 technically still needs a 2nd approval. but Krock has marked his code comments as resolved. so I hereby self-approve 21:57 MTDiscord Question for core devs: if a professional VR dev team developed VR compatibility for minetest, and made a PR, would that be taken seriously/reviewed/merged in atimely manner, or would it likely sit since it is not on the roadmap? 22:00 celeron55 if it's neatly made, then i suppose why not. the biggest question is probably how it will be maintained, as i don't expect any core devs to have any VR devices 22:01 celeron55 "professional VR dev team" sounds like an awful lot of code 22:02 MTDiscord Specifically, team beef studios, who are known for porting many foss games to VR 22:03 rubenwardy Your best chance is to hire a dev, shame x2024 has gone 22:03 celeron55 historically different rendering modes were taken into the engine quite early. in that sense VR support wouldn't be stepping out of line 22:05 MTDiscord This is specifically asking, "if Team Beef develops the VR port, can they expect to merge the changes upstream or have to maintain a fork?" 22:15 celeron55 it's subject to the same process as everything. currently graphics things are on the roadmap which helps. but it woould be subject to review like anything 22:15 celeron55 -o 22:15 celeron55 it's always appreciated when people try to work with upstream instead of doing a fork. forks tend to get abandoned eventually and it's sad for everyone 23:26 rubenwardy It's a cool thing to support, I don't own any vr hardware to test it though