This section will explain how to properly interact with the Wormhome Core Message Layer in an EVM ecosystem.
Messages in Wormhole take the form of a Verified Action Approval (VAA) and both terms can be used interchangably. The rest of this section will only use the term VAA.
Here is the interface for applications to interact with Wormhole's Core Contract to publish VAAs or verify and parse a received VAAs.
Instantiating the interface will depend on the contract address of your development ecosystem and blockchain.
Below is an example line of code to instantiate the interface for mainnet Ethereum:
address private wormhole_core_bridge_address = address(0x98f3c9e6E3fAce36bAAd05FE09d375Ef1464288B); IWormhole core_bridge = IWormhole(wormhole_core_bridge_address);
The Wormhole Core Layer has two important interactions -- (1) emit VAAs, and (2) parse and verify VAAs that originated from other chains.
There are two forms of VAAs that can be emitted:
- Single VAA: all messages will be emitted in this format
- Batch VAA: messages that are generated from the same transaction will be emitted in this format. This feature was developed to provide an easier paradigm for composability and better gas efficiency for more involved cross-chain activity.
To emit a VAA, always use
publishMessage which takes in the following arguments:
nonce(uint32): a number assigned to each message
nonceprovides a mechanism by which to group messages together within a Batch VAA. How the
nonceis used is described below.
Consistency(uint8): the number of blocks that Guardians will wait before signing a message
- Each blockchain has different finality periods based on the consensus mechanism. In general, higher consistency values provides more security against blockchain reorgs. Here are the consistency levels by blockchain that are used by the xAsset layer to have a high level of guarantee against reorgs.
Payload(bytes): raw bytes to emit
- It is up to the emitting contract to properly define this arbitrary set of bytes.
publishMessage will output a
sequence (uint64) that is used in conjunction with
emitterAddress to retrive the generated VAA from the Guardian Network.
How Batch VAAs are generated
There are two mechanisms that allow messages to be Batched together that represent a base and more advanced level of composability.
- All messages originating from the same transaction will be batched together.
- Messages that originate from the same transaction and are assigned the same nonce are additionally batched together.
Note: Single VAAs will always be emitted for each message within a transaction, regardless of if a message is included in a batch or not.
Here is an example of how messages generated from the same transaction may be batched together:
A transaction X that generates 6 messages [A, B, C, D, E, F] that are assigned
nonce[1, 2, 2, 3, 3, 4] respectively will generate the following VAAs:
- (1) full transaction batch VAA
- [A, B, C, D, E, F]
- (2) smaller batch VAA
- [B, C]
- [D, E]
- (6) single VAA
Parsing and Verifying a VAA will depend on the type of VAA that your application expects: a Single VAA or a Batch VAA.
For either VAA type, remember to collect gas fees associated with submitting them on-chain after all VAAs have been verified.
To properly parse and verify a single VAA, always use
parseAndVerifyVM which takes in one argument:
encodedVM (bytes). This function will return three arguments:
vm(VM): Structured data that reflects the content of the VAA.
- A breakdown of this message format is described in the VAA section. Aside from the header information, which can be considered 'trusted', it is up to the recipient contract to properly parse the remaining payload, as this contains the verbatim message sent from the emitting contract.
valid(bool): Boolean that reflects whether or not the VAA was properly signed by the Guardian Network
reason(string): Explanatory error message if a VAA is invalid, or an empty string if it is valid.
To properly parse and verify a batch VAA, always use
parseAndVerifyBatchVM which takes in two arguments:
encodedVM (bytes) and
In most scenarios, you'll want to set
cache equal to true.
This will return a VM2 object, containing all the 'headless' VAAs contained inside the batch VAA. These headless VAAs can be verified by
parseAndVerifyVM, which means that modules which verify messages in an xDapp can be agnostic as to whether a message came from a batch VAA or a single VAA.
The Best Practices section goes into more depth of how to interact with the coreLayer.