Time Nick Message 00:47 [MatrxMT] am I the only one having issues with GH since they updated the website? Like, pages endlessly loading from time to time 00:49 [MatrxMT] also great, issues don't render anymore sub-tasks (contrary to PRs), great job 01:14 MTDiscord yes 07:50 MTDiscord Nope, similar experience here 13:04 sfan5 works for me 13:47 MTDiscord Zughy: I've been having similar issues 14:45 rubenwardy Idea: if the server told the client which files were available from remote media, the client could just request them without needing to POST index.mth. This would allow using static hosting for remote media. The server could build this list by doing a GET request of an index file or could HEAD/GET all the media files 14:52 sfan5 solving the wrong thing IMO 14:52 sfan5 just have the client do GET to index.mth instead? 18:00 sfan5 merging #15697, #15715, #15714, #15713, #15704, #15718 in 10m or so 18:00 ShadowBot https://github.com/minetest/minetest/issues/15697 -- Restrict relative mouse mode to Wayland users by sfan5 18:00 ShadowBot https://github.com/minetest/minetest/issues/15715 -- Lua get_biome_data: calc heat and humidity only once by kno10 18:00 ShadowBot https://github.com/minetest/minetest/issues/15714 -- Move EnumString to separate file and add enum_to_string by cx384 18:00 ShadowBot https://github.com/minetest/minetest/issues/15713 -- Fix documentation for the official Docker image after project rename by AFCMS 18:00 ShadowBot https://github.com/minetest/minetest/issues/15704 -- Disable SDL2 for 5.11.0 by sfan5 18:00 ShadowBot https://github.com/minetest/minetest/issues/15718 -- Rename dev.minetest.net to dev.luanti.org by mark-wiemer 18:12 [MatrxMT] :o 18:13 [MatrxMT] the merge train has arrived 18:19 MTDiscord Thanks for the ping on GitHub about the benchmarks not getting reported. Lars, have you looked into it at all or not yet? I can look into it next Saturday and work on a fix. 18:20 MTDiscord My semester starts tomorrow, and I don't want to get behind on classes right off the bat. 18:20 sfan5 I only meant "how do we make --run-benchmarks skip the tests" actually 18:23 MTDiscord The short answer is "I don't know". But my hunch is to set some configuration parameter in the run call provided for this thing. I think I've run Catch2 benchmarks without the tests before using CTest's CLI, which means Catch2 must have a CLI option for it. If it has a CLI option then that can be set using the API as well. 18:24 MTDiscord I don't recall what that option is or the details of the API to set configuration parameters. 18:27 MTDiscord I can do some research this afternoon maybe. It should not take that long, and I want to propose something actually useful and actionable in the issue, not just a hunch. 18:45 sfan5 >note: candidate function not viable: cannot convert argument of incomplete type 'GenericCAO *' to 'ActiveObject *' for 9th argument 18:45 sfan5 what 18:48 MTDiscord Someone changed some imports a bit ago to include less Irrlicht bloat. I'm putting my bet on that commit haha. 19:47 sfan5 i tried to optimize collision code and made it 10 times slower, oops 19:56 sfan5 maybe it is more efficient if you're going 500m/s 20:17 celeron55 changelog: increased default player speed to 500m/s to take advantage of new efficient collision code 21:14 MTDiscord I had looked into it very briefly to decide it's not trivial 21:14 MTDiscord Ugh. 21:15 MTDiscord Well, programming would be boring if there weren't problems like this. 21:15 sfan5 i got some improvements out of simple changes :) 21:16 MTDiscord I think this needs two paths 21:16 MTDiscord one for the typical small area and one for very fast objects 21:23 sfan5 see my comment in #15227 21:23 ShadowBot https://github.com/minetest/minetest/issues/15227 -- Node - object collision performance is abysmal for very fast entities