Working with Parity to align RMRK pallet as close as possible

[Yuripetusko]

There are three main features of RMRK pallet that unique v1 doesn't have.

1. Resources
2. Nesting
3. Equip logic

We've been chatting with @jsidorenko from Parity for ways to add these features to uniques v2 to support RMRK functionality, while keeping uniques as generic as possible

Resources

Since resources in RMRK standard cannot be simple uint/string values, we can save them as Structs. i.e. ResourceStruct{id: 1, part_id: 1, base_id: 1, metadata: 'ipfs://...', collection_whitelist: [0, 2, 42]}

Then when a pallet validation or JS need to get them back it can be done by calling decode(value, ResourceStruct)

This way we can save resources in existing uniques Attributes storage without the need for any new storage

Nesting

This is being discussed and worked on, the main bottleneck is expensive traversing down the tree during certain operation like burn, we can perhaps set a limit to recursion and throw if it's too deep, to force user to un-bundle the nesting before performing this. But this is still TBD

Equip

Given the Resource struct in Attribute storage, the only thing left for equip method to correctly validate is to decode the Resource struct of a child NFT that is being equipped, and check if parent NFT has compatible resource struct of type of Base with matching base_id and slot_id and collection_whitelist

Let me know if I am missing anything from RMRK spec ?