Open madjin opened 1 year ago
Voxels has a wearables / layering system with JSON export that would be useful to investigate for interop purposes, basically works by parenting / offsetting the assets to bones.
JSON: https://gist.github.com/madjin/bc8ee04f228d1bf43f60ac9f98155720
Converted from proprietary format (.vox) to svox to glTF then encoded the metadata about each wearable into the glTF file itself using glTF-Metadata-CLI
Challenge: Different rigs / armature bone names, how will user + wearable know offset if name is different?
We can then test Voxel wearables on other avatars (currently not possible in Voxels but platform open to it)
See the finished script here: https://github.com/M3-org/cryptovoxels-wearables/blob/main/scripts/convert_json.sh
Interesting case study via Webaverse: https://hackmd.io/@platote/ryik4slc9#6-Adding-a-hat
Develop a tier/category system that groups the different metaverses based on similar characteristics. ie. mesh count, poly count, material count, etc.
VRChat offers good guidelines for avatar performance: https://docs.vrchat.com/docs/avatar-performance-ranking-system#quest-limits
It's influenced Hyperfy's avatar performance guidelines: https://docs.hyperfy.io/avatars#rank-system
Notes on layering system: https://github.com/gm3/boom-tools/wiki/API-Reference
We can use this open source character creator program to help document parts about the layering system, there are a good handful of open bounties for helping with such too: https://github.com/webaverse-studios/CharacterCreator
There's also many dresser prefabs / apps worth looking into on booth.pm, an indie marketplace many avatar creators use, such as Auto Dresser: https://booth.pm/ja/items/1300847