An example implementation of a Catalog dapp.
Jakub Mitoraj 504f36428c TUI support for new Content creation | 19 uur geleden | |
---|---|---|
assets | 1 maand geleden | |
src | 19 uur geleden | |
Cargo.lock | 2 weken geleden | |
Cargo.toml | 1 maand geleden | |
LICENSE | 3 maanden geleden | |
README.md | 3 maanden geleden |
a distributed peer to peer application implementing built-in Catalog app-type of gnome protocol.
This application is intended to provide fast and easy access to other applications and resources created on top of gnome protocol that uses swarm-consensus mechanism for ensuring synchronization of data across the Internet.
Every user will be given a village to maintain. His village will consist of a street with his own home and other houses that are owned by users' friends and others. User will be able to enter selected home as a guest and see what's inside. User will also be allowed to visit another user's village in order to browse it's contents. Maybe we can have a house with regular front door and back door used as a teleport. Inside his house a user will define rooms with items inside of them. Access to each room (and house, and not village!) will be defined by the user. A room will contain data or links to other user's data. Others will be allowed to open/download data behind those links. In front of every house there will be a post box where any user will be able to put an encrypted message to owner of given house. A user can create more streets if he wants to. On those streets he can put other houses or different kinds of buildings representing different types of p2p applications that he and visitors will be able to enter and use.
Every village will have it's data stored in a p2p fashion by using a swarm. If a user adds a new friend, then he will also sync that swarm onto his hard drive. User will be able to configure what data should get synchronized by default, or for particular friend. If someone stores a lot of files and our memory space is limited we can sync only parts of our friends swarmed data.
This is a very high level overview and it will take a lot of time to crystalize.
In order to create a village, or any other implementation of Catalog app type we need to define data types within Catalog. Then village app can interpret these types according to it's environment of streets, houses, rooms etc. A house/building is a representation of a link to another application. A link consists of SwarmID (unique FounderId and a SwarmName), and ContentID. ContentID is just a number used to identify some particular data within a swarm. ContentID should map to ContentHeader and DataChunks. ContentHeader should have DataType field, name and ContentIndex. ContentIndex should be a binary tree of hashes of sub-hashes of Leafs that are 1024 bytes long data chunks. DataChunks, when synced, should contain all the data chunks of given ContentID. We should always sync all ContentHeaders of all ContentIDs for a given swarm, but we may choose to ignore syncing DataChunks for some or all ContentIDs. ContentIndex is a tool to reconstruct data from pieces in correct order. A link should be up to 64 bytes long, so 1 byte for DataType, 8 bytes for FounderId, 4 bytes for ContentID, and 51 bytes for SwarmName (an UTF-8 String of limited to 51 bytes size). A house is a link to other user's home/village, whereas other types of building link to other types of application, like a Forum app for example. A room within a house, together with floors is used for ordering data. Inside a room different objects can be used to represent different kinds of files. For example a square can represent a picture, just like icons do on a computer system. Overall this can be remotely representing folders/files on computer's disk. A user can also use a room as a way to organize information on some specific topic with interlinked notes containing optional external links (sort of Zettelkasten method). A house can have up to 32 floors with up to 8 rooms on each floor (u8 value 5|3 split). A village can cosist of up to 256 streets. Each street is a representation of a label/tag that given user is interested in. There can be duplicates of houses on multiple different streets, but not on the same street. On every street there can be a Broadcast channel set up used for providing unsynced data for everyone interested in whatever content it offers. Later there can also be Multicast channels with access limited to selected visitors.