refactor(fuel): configuring forc and fuel-core upgrades to the Pyth fuel contract #2858
Conversation
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
| @@ -374,7 +374,7 @@ fn update_fee(update_data: Vec<Bytes>) -> u64 { | |||
| total_fee(total_number_of_updates, storage.single_update_fee) | |||
| } | |||
|
|
|||
| #[storage(read, write), payable] | |||
| #[storage(read, write)] | |||
There was a problem hiding this comment.
One thing I wanted a second opinion on: one of the new issues that was introduced was that helper functions cannot be listed as "payable" anymore (only functions within the abi are allowed to have that tag). Leaving the tag in there stops the forc build. Removing the tag allows it to build, and all integration tests still pass. I still wanted to make sure that this is a valid removal (i.e. does the payable trait pass down from its calling function).
There was a problem hiding this comment.
Yeah i believe it is ok, i think Fuel has borrowed some semantics from EVM (just as it is a layer 2) and in EVM entrypoint functions that are not payable always check that msg.value == 0
| // let a: b256 = child_a.into(); | ||
| // let b: b256 = child_b.into(); |
| let a: b256 = b256::from_le_bytes(child_a.clone()); | ||
| let b: b256 = b256::from_le_bytes(child_b.clone()); |
There was a problem hiding this comment.
I'm so confused how little endian could work here. generally raw bytes operations translate to big endian on numbers. also in the docs i see the example is similar.
https://fuellabs.github.io/sway/master/std/primitive.b256.html#method.from_be_bytes
If I'm right. I'm wondering how the tests passed with this. if I'm wrong, i like to know how :?
There was a problem hiding this comment.
I think I misinterpreted the structure of the array, BE is definitely right. For some reason the tests don't change state even when I completely mess with the code. I'll try to rebuild things and see if that changes anything.
Summary
Swaylend Fuel clients asked us to refactor the fuel contract to work with newer versions of the forc and fuel-core libraries (0.68.9 and 0.43.2 respectively). Simply changing the versions resulted in some compile errors, so code changes had to be made.
Rationale
Client request for support for updated libraries.
How has this been tested?
I refactored the code until forc build worked and the existing integration tests through cargo test all passed.