Time Nick Message 00:35 Wuzzy Just to be clear, when you say "rebase needed" you don't neccessarily also imply a squash, right? 00:35 Wuzzy i have the feeling i asked this before ... but my memory is broken 00:42 MTDiscord anyone got an idea on allowing CSMs to create CSM specific entities 00:43 MTDiscord i have ideas related to wields 00:54 ashtrayoz Is it still practical to cross-compile the minetest engine from linux (preferably Debian/Devuan) ? 00:57 ashtrayoz Targets would include MsWin64 and maybe MsWin32. 01:33 Wuzzy I finally figured out what was wrong with clientside strings in my Builtin translate PR. And I fixed it accordingly. #10693 01:33 ShadowBot https://github.com/minetest/minetest/issues/10693 -- Builtin translate (2nd attempt) by Wuzzy2 02:02 pgimeno Wuzzy: I can't talk for the core team, but a rebase is substantially different from a squash. You can squash without rebasing and rebase without squashing, and what is often requested is a rebase, not a squash. 02:02 Wuzzy yea,, i suspected sth like that 06:41 ashtrayoz Ah, got it. Not enough to use update-alternatives, have to modify the toolchain file in util/buildbot as well. 11:26 sfan5 ashtrayoz: the official windows builds are cross-compiled 11:27 sfan5 ubuntu has some weird choice that the compiler we actually need is called not x86_64-w64-mingw32-gcc but x86_64-w64-mingw32-gcc-posix 13:32 Krock What's the non-deprecated way to retrieve an entity name? "get_entity_name" is deprecated for >9 years but no alternative is documented 13:32 Krock there's the "name" field within the entity table but I don't know whether that's supposed to be used 14:05 sfan5 I don't know either 14:05 sfan5 get_entity_name() seems like a good idea if once day you're going to have more than just luaentities and players 14:06 sfan5 though usually when that is used it's only used on luaentities anyway 14:10 MTDiscord the name field is supposed to be used 17:31 pgimeno judging by these two commits, it seems to be deprecated in favour of the .name field: https://github.com/minetest/minetest/commits/e82a76e43848eae556843709b66b6c8703354931 23:04 ashtrayoz sfan4: Yes, I thought update-alternatives --config x86_64-w64-mingw32-g++ and pointing to posix would fix that, but it didn't. 23:04 ashtrayoz So I SET(CMAKE_C_COMPILER x86_64-w64-mingw32-gcc-posix) and SET(CMAKE_CXX_COMPILER x86_64-w64-mingw32-g++-posix) 23:06 ashtrayoz It now compiles, but I am getting a missing entry point at runtime: _ZNSt7__cxx1118basic_stringstreamIcSt11char_traitsIcESaIcEEC1Ev in libstdc++-6.dll 23:13 ashtrayoz I am guessing DLL-hell on the windows box, the dll file shipped with the compiler shows _traitsIcESaIcEEC1EOS4_, not _traitsIcESaIcEEC1Ev 23:14 sfan5 where did you get your libstdc++-6.dll from? 23:17 ashtrayoz aptitude 23:17 ashtrayoz Copied from /usr/lib/gcc/x86_64-w64-mingw32/8.3-posix/libstdc++-6.dll 23:18 ashtrayoz (oops - I demoted you to sfan4 - Sorry 'bout that Chief!) 23:22 sfan5 that's the correct place in principle 23:22 sfan5 however, the buildbot uses libraries built by me with a newer compiler version 23:22 sfan5 and that ends up not working 23:22 ashtrayoz :-( 23:23 sfan5 https://github.com/minetest/minetest/blob/225e69063fa0c3dc96f960aa79dfc568fe3d89a8/.gitlab-ci.yml#L177-L178 23:23 sfan5 you can download and use the same mingw as the CI does, that one works