I think, moving forward, that there will be value in having the object collection distributed in distinct hierarchies with versioning and owners.
there should (imho) be CORE generic attribute types,
(bool/str/int/float/enum/address/lat-lng/street/city/state/country etc)
there should (imho) be CORE object types/templates
These should be curated/owned by anytype.
These should be distributed versioned.
There could be add-on collections (scientific | architectural | fashion | creatures | planets | etc)
There could be ‘manufacturer’ collections (brands/etc)
Each collection should have an owner, and a version, and be granularly importable…
this permits a distributed ownership model (Dell could own the object hierarchy of their computer types, for a silly example) , while also facilitating a mechanism both for declaring dependencies, and providing a mechanism to discover when expectations of the state of your dependencies has changed, such that the user can be informed that there may be some data manipulation necessary
it seem to me that without the ability to consistently import/export/sync/align objects/types/attributes and their properties, the sharing of datas is likely to be incredibly challenging between distinct islands of data… as the attributes I’ve created in my environment may or may not be compatible with the attributes you’ve created in yours
does this make sense? or does the idea need to be articulated better to be of use?