10/13

Image Added

Image Added

Image Added

Image Added

DImage Added

Image Added

evImage Added

Image Added

Image Added

Image Added

Image Added

Image Added

Image Added

Image Added

Image Added

Image Added

Image Added

elopment

...

Needs

Setup in Local:

Please see the following document for local setup: 

...

BROADCAST DATA a message 

MESSAGES: WHO SENT

• Sends a message visible to all parties in the network
  • The message describes who sent it, and exactly what data was sent
• A message has one or more attached pieces of business data
  • Can be sent in-line, uploaded in advance, or received from other parties
  • Can include smaller JSON payloads suitable for database storage
    • These can be verified against a datatype
  • Can include references to large (multi-megabyte/gigabyte) BLOB data
• Sequenced via the blockchain
  • The blockchain does not contain any data, just a hash pin
• Batched for efficiency
  • One batch can pin hundreds of message broadcasts
  • The whole batch is written to the shared storage

Privately send data

• Sends a message to a restricted set of parties
  • The message describes who sent it, to whom, and exactly what data was sent
• A message has one or more attached pieces of business data
  • Can be sent in-line, uploaded in advanced, or received from other parties
  • Can include smaller JSON payloads suitable for database storage
    • These can be verified against a datatype
  • Can include references to large (multi megabyte/gigabyte) BLOB data
• A group specifies who has visibility to the data
  • The author must be included in the group - auto-added if omitted
  • Can be specified in-line in the message by listing recipients directly
  • Can be referred to by hash
• Private sends are optionally sequenced via pinning to the blockchain
  • If the send is pinned:
    • The blockchain does not contain any data, just a hash pin
      • Even the ordering context (topic) is obscured in the on-chain data
      • This is true regardless of whether a restricted set of participants are maintaining the ledger, such as in the case of a Fabric Channel.
    • The message should not be considered confirmed (even by the sender) until it has been sequenced via the blockchain and a message_confirmed event occurs
    • Batched for efficiency
      • One batch can pin hundreds of private message sends
      • The batch flows privately off-chain from the sender to each recipient
  • If the send is unpinned:
    • No data is written to the blockchain at all
    • The message is marked confirmed immediately
      • The sender receives a message_confirmed event immediately
    • The other parties in the group get message_confirmed events as soon as the data arrives

Listen for events 

• Probably the most important aspect of FireFly is that it is an event-driven programming model.
• Parties interact by sending messages and transactions to each other, on and off chain. Once aggregated and confirmed those events drive processing in the other party.
• This allows the orchestration of complex multi-party system applications and business processes.
• FireFly provides each party with their own private history, that includes all exchanges outbound and inbound performed through the node into the multi-party system. That includes blockchain backed transactions, as well as completely off-chain message exchanges.
• The event transports are pluggable. The core transports are WebSockets and Webhooks. We focus on WebSockets in this getting started guide.

• Check out the Request/Reply section for more information on Webhooks

Define a datatype

• As your use case matures, it is important to agree formal datatypes between the parties. These canonical datatypes need to be defined and versioned, so that each member can extract and transform data from their internal systems into this datatype.
• Datatypes are broadcast to the network so everybody refers to the same JSON schema when validating their data. The broadcast must complete before a datatype can be used by an application to upload/broadcast/send data. The same system of broadcast within FireFly is used to broadcast definitions of datatypes, as is used to broadcast the data itself.

Responsibilities & Pluggable Elements

More importantly, what the split of responsibilities is between Connectors and Infrastructure Runtimes.

...