Time Nick Message 19:01 Krock will push https://krock-works.uk.to/u/patches/0001-guiConfirmRegistration-Fix-hidden-error-message.patch in 15 minutes 19:01 Krock fix for the error message on mismatch 19:13 p_gimeno gotta love those magic numbers ^.^ 19:22 rubenwardy I want to add auto-layouting containers to formspecs 19:22 rubenwardy to do away with this 19:22 rubenwardy stack_container[0,0;3,3;vertical] container[0,0;3,3] label[0,0;Hello!] container_end[] stack_container_end[] 19:23 rubenwardy the positions of children in auto-layouting containers would be offsets or ignored 19:23 rubenwardy maybe 19:23 rubenwardy for something like a grid layouter, you'd have to require all children to be cell_containers 19:23 rubenwardy to provide meta 19:24 rubenwardy maybe 19:26 rubenwardy would be an idea to replace the serialisation format before then 19:26 rubenwardy I'm getting warmer to the idea of embedding the entire Irrlicht GUI code into our code, and maintaining that ourselves 19:28 rubenwardy Do we want to do 5.1.1? 19:28 rubenwardy There's been some important bug fixes, iirc 19:29 sfan5 why not just 5.2? 19:29 rubenwardy I wish our release precedure was more automated, especially making things like Windows and Android builds 19:29 rubenwardy the benefit of releasing patch releases is that it increases stability without adding features which decrease it 19:30 rubenwardy I guess we could release minor versions more 19:30 rubenwardy with patch releases, there's also less need for a full week freeze 19:31 sfan5 with the downside of having to spend time on backporting changes, managing different releases and stuff 19:31 rubenwardy yeah 19:31 sfan5 IMO patch releases should be reserved for when things are really broken (see 5.0.1) 19:32 sfan5 and if we feel there are important fixes we can just a release out of schedule 19:32 rubenwardy ok 19:33 rubenwardy that's fine be me 19:34 sfan5 another reason not to wait until march/april with 5.2 is to get it into ubuntu 20.04 19:34 sfan5 but that's not really important 19:37 sfan5 anyway just to state it explicitly: I would support doing a 5.1.1 / 5.2 release 19:42 Krock pushing patch... 19:44 Krock +1 on "really broken" releases 19:45 Krock I personally wouldn't bother backporting bugfixes if they're not somewhat necessary 19:46 Krock January/February for 5.2.0 would be great. I think master is already quite stable 20:01 rubenwardy merging #9090 in 5 20:01 ShadowBot https://github.com/minetest/minetest/issues/9090 -- Refactor to centralize GUIButton styling/rendering code by Df458 20:06 rubenwardy done 21:05 rubenwardy Pushing trivial bug fix in 20 21:05 rubenwardy !title https://github.com/rubenwardy/minetest/commit/8e63d22d6472b61477f804dc5b00705650f90409 21:05 ShadowBot rubenwardy: Fix spaces breaking formspec_version[] tag · rubenwardy/minetest@8e63d22 · GitHub 21:32 rubenwardy oops 22:12 rubenwardy #9184 22:12 ShadowBot https://github.com/minetest/minetest/issues/9184 -- WIP: Add Lua builtin unit tests by rubenwardy 22:13 rubenwardy no objections I trust? 22:19 rubenwardy also: 22:19 rubenwardy #9185 22:19 ShadowBot https://github.com/minetest/minetest/issues/9185 -- Ignore build/, except build/android by rubenwardy 22:20 sfan5 using build as cmake build dir was already ruined when build/android was created 22:21 rubenwardy yeah 22:21 rubenwardy my editor has support for cmake, but it appears that the setting for build dir is global 22:22 rubenwardy through a plugin* 22:54 p_gimeno I use a compile/ dir because of the things in build/ 22:54 rubenwardy good idea 22:55 rubenwardy that isn't git ignored though 22:59 rubenwardy oh great, non-determinist behaviour in core.privs_to_string 23:08 paramat i agree with this: "patch releases should be reserved for when things are really broken", and prefer we do not do a patch release. we could release minor versions more often, so 5.2.0 early next year seems good 23:08 rubenwardy ok, glad that's decided 23:29 rubenwardy #9184 is ready for review 23:29 ShadowBot https://github.com/minetest/minetest/issues/9184 -- Add Lua builtin unit tests by rubenwardy 23:30 rubenwardy it only adds tests for vector and string.split, and moves existing tests to the framework