Time Nick Message 00:13 MTDiscord that's not even that bad lol. the way C++ always expands global names for symbols, i've achieved a symbol that was an entire screenful just by nesting some lambdas ;) 00:13 MTDiscord and my code wasn't even that contrived, and rather reasonably structured; mostly everything followed naturally from doing the lazy "use a lambda for iteration" 08:30 Krock Will merge #16654 #16684 #16529 #16164 in 15 minutes 08:30 ShadowBot https://github.com/luanti-org/luanti/issues/16654 -- IrrlichtMt: Implement Page Up/Down navigation on edit boxes by SmallJoker 08:30 ShadowBot https://github.com/luanti-org/luanti/issues/16684 -- Fix minor typo by a-bad-dev 08:30 ShadowBot https://github.com/luanti-org/luanti/issues/16529 -- Option to make sky stars deterministic by fetsorn 08:30 ShadowBot https://github.com/luanti-org/luanti/issues/16164 -- Exit confirmation dialog on ESC in main menu by siliconsniffer 08:48 Krock done 17:01 Krock @luatic How do I best test #16655 ? I have no idea which models or animations I should compare. 17:01 ShadowBot https://github.com/luanti-org/luanti/issues/16655 -- Refactor skinned mesh weight data structure by appgurueu 17:01 Krock let alone - what's affected and what not. 17:17 MTDiscord Krock: This should only affect skinned meshes, so .x, .gltf or .b3d. This refactor mostly only touches weights used in skeletal animation. 17:18 MTDiscord Devtest contains some test models for all three: The lava flan is .x, the animated spider is .gltf, sam is .b3d. 17:19 MTDiscord The animations can be compared against those of current master, but for the most part it suffices to just look at them: If something is wrong, you'll usually notice. 17:23 Krock alright. I'll also do some testing on a server with animated mobs. let's hope those are 'affected' by this change too 17:24 MTDiscord Krock: anything skeletally animated will be affected 17:24 Krock what other animations are there? texture animations? 17:24 MTDiscord so every .b3d model for example, which is most animated models (.x is very rare, not sure how much adoption .gltf has so far) 17:25 MTDiscord Krock: rigid animation, basically a special case of skeletal animation with a mesh buffer attached to a particular bone 17:26 MTDiscord some .x models use it (but it seems to be rare), and .gltf can use it as well 17:27 MTDiscord rigid animation uses a different path that doesn't involve weights, it composes the transformation matrix with the bone transformation 20:16 MTDiscord random thought: semitransparency issues can be alleviated to some degree by rendering everything in two passes 20:16 MTDiscord first you render all the opaque fragments, then you render all the semitransparent fragments 20:17 MTDiscord the advantage of this is that the "opaque fragment becomes invisible if there's a semitransparent fragment in front of it" bug becomes impossible 20:17 MTDiscord the disadvantage is that you now do two passes, and it still won't fix semitransparent fragments becoming invisible behind other semitransparent faces