Conversation
clangenb
added
A0-core
|
do we need to upgrade integritee-node to make integration tests work? |
| pub type IntegriteeUncheckedExtrinsic<Call> = | ||
| UncheckedExtrinsicV4<Address, Call, PairSignature, IntegriteeSignedExtra>; | ||
|
|
||
| /// Signature type of the [UncheckedExtrinsicV4]. | ||
| pub type Signature<SignedExtra> = Option<(Address, PairSignature, SignedExtra)>; | ||
| UncheckedExtrinsic<Address, Call, PairSignature, IntegriteeSignedExtra>; | ||
|
|
There was a problem hiding this comment.
This was the tricky part as to why the node failed to interpret the extrinsic correctly. The signature used by the api client is now a plain MultiSignature
cli/Cargo.toml
Outdated
| substrate-api-client = { default-features = false, features = ["std", "sync-api"], git = "https://github.com/encointer/substrate-api-client.git", branch = "v0.9.42-tag-v0.14.0-retracted-check-metadata-hash" } | ||
| substrate-client-keystore = { git = "https://github.com/encointer/substrate-api-client.git", branch = "v0.9.42-tag-v0.14.0-retracted-check-metadata-hash" } | ||
| substrate-api-client = { default-features = false, features = ["std", "sync-api"], git = "https://github.com/encointer/substrate-api-client.git", branch = "cl/extrinsic-v5-v0.9.42" } | ||
| substrate-client-keystore = { git = "https://github.com/encointer/substrate-api-client.git", branch = "cl/extrinsic-v5-v0.9.42" } |
There was a problem hiding this comment.
I still have to think of a proper branch name for this.
There was a problem hiding this comment.
Added a new branch with the rationale that we have too many patches to name them all in the branch name. I added the listed patches in the Readme of this branch.
v0.9.42-tag-v0.14.0-integritee-patch
However, it lives in the encointer fork. I am not sure if we want to have our own fork of it.
brenzi
left a comment
brenzi
left a comment
There was a problem hiding this comment.
LGTM, but I'd like to trigger a -dev release and test this on Paseo before merging
There was a problem hiding this comment.
I tested the following steps with 0.17.0-dev1
- deploy for asset-hub-paseo Incognitee testnet (running against stable2506-2 nodes)
- using try.incognitee.io: shield & unshield works
- using PJS, sending funds to shard vault yields in shielding (meaning the worker understood the extrinsic)
the errors about decoding extrinsics no longer show in the logs
| #[test] | ||
| fn can_parse_asset_hub_transaction_v4() { | ||
| use itp_utils::FromHexPrefixed; | ||
| use substrate_api_client::ac_primitives::extrinsics::UncheckedExtrinsic as XTV5; | ||
|
|
||
| // We only care about the preamble, so we use the `()` to stop decoding after the preamble. | ||
| let ex_v5: XTV5<Address, (), ParentchainSignature, ParentchainTxExtension> = XTV5::from_hex("0x4d0284007c77c5f95a72c19e051233b3b1e01df74ec13cb161b109fc5f16ce1c65316f2401f63f850629722493607fb04dfa64d9bd62c454ea325b96d71938d0a52f513c4b35681b9c323b9a09b9f7227e10a5707624afbeb2232759505119da68fa6de08b2400040000000a0300b1df638e78cd896db34d8e62722fd755360d61dac2f3b048e24cd26a5ace35690b008cb6611e01").unwrap(); | ||
|
|
||
| match &ex_v5.preamble { | ||
| // If the preamble is signed, it means that an XT V4 was sent | ||
| Preamble::Signed(..) => {}, | ||
| Preamble::General(..) => panic!("unexpected preamble belonging to XT V5"), | ||
| other => panic!("unexpected preamble: {:?}", other), | ||
| }; | ||
| } |
There was a problem hiding this comment.
This is how you can identify to which tx version an extrinsic belongs.
2 participants
This PR mostly consists of an api-client update to get the backwards compatible support for the Extrinsic V5. Closes #1723.
Tests