Polkadot v0.9.38 #1332
Polkadot v0.9.38 #1332
Conversation
|
@NunoAlexandre I can not reproduce the integration test failures locally. Could you try to? |
That's interesting, because I can indeed. What's your rust toolchain? It should be picking it up from the toolchain file but still worth checking! |
|
@NunoAlexandre @mustermeiszer I managed to reproduce some failures locally. Currently looking into them and will update as soon as I have more info. |
That's awesome, thanks! Since I often bump against those errors when updating our chain to a newer Polkadot version, I would very much appreciate if you could educate me on how to fix them so I can move that forward in the future, shall that happen again 👍 |
|
After doing some investigation into the test errors we figured out that the original error came from genesis builder, where a decoding error would happen when building the wasm, more info here - https://substrate.stackexchange.com/questions/7881/error-loading-of-original-wasm-failed Tested locally with |
NunoAlexandre
force-pushed
the
polkadot-v0.9.38
branch
from
8126e4d
to
f5505a5
Compare
| @@ -2,7 +2,7 @@ | |||
|
|
|||
| set -eux | |||
|
|
|||
| RUST_TOOLCHAIN="${RUST_TOOLCHAIN:-nightly-2022-11-14}" | |||
| RUST_TOOLCHAIN="${RUST_TOOLCHAIN:-nightly-2023-02-07}" | |||
There was a problem hiding this comment.
Do we know on which one parity is running polkadot and substrate?
There was a problem hiding this comment.
No, I just picked this since is the one for rust v1.69. I'll double-check what parity is using for their v0.9.38 polkadot/substrate.
There was a problem hiding this comment.
I don't think parity is using any specific versions given:
https://github.com/paritytech/substrate/blob/polkadot-v0.9.38/.maintain/init.sh#L8
I also checked their CI scripts and it seems that they are using whatever is stable/nightly when they build, however, someone might know more. Maybe @wischli can help? ^^
There was a problem hiding this comment.
Parity doesn't use a nightly (I believe we only need it for fudge these days?), so I wouldn't be surprised if they just track stable.
There was a problem hiding this comment.
The releases in the Polkadot repository expose minimal compatible Rust versions For v0.9.38 that is
- Rust Stable: rustc 1.66.1 (90743e729 2023-01-10)
- Rust Nightly: rustc 1.67.0-nightly (a00f8ba7f 2022-11-15)
For the latest release v0.9.42 we are on
- Rust Stable: rustc 1.68.2 (9eb3afe9e 2023-03-27)
- Rust Nightly: rustc 1.67.0-nightly (a00f8ba7f 2022-11-15)
For their CI, they seem to follow latest stable as stated by @branan, here's an excerpt:
[2023-05-30 22:10:38] $ cargo --version
[2023-05-30 22:10:38] cargo 1.69.0 (6e9a83356 2023-04-12)
However, in their Github actions, they are using hongfuzz fuzzer with latest minimal nightly.
There was a problem hiding this comment.
Parity doesn't use a nightly (I believe we only need it for fudge these days?), so I wouldn't be surprised if they just track stable.
Where does this assumption come from? We could remove the necessity in fudge then.
| "The function is registered but the type mismatches. Expected {}, found: {}", | ||
| expected, found |
There was a problem hiding this comment.
Saving code lines with the new rust feature 🤯
|
Thank you all so much ❤️ auto-merge is enabled 🚀 |
The update to Polkadot 0.9.38 emcompasses two major changes in the framework:
None of those changes ended up having a substantial impact in the codebase. Regarding XCM, I took the opportunity to simplify the Xcm config around
CurrencyIdConvertby using theOrmlAssetRegistryfor all the conversions, now that 1) we have CFG and AIR registered in Centrifuge's and Altair's registries (respectively) and 2) since thelocationof Tranche tokens is set toNoneand can't therefore be handled through XCM.To Do
ignoredfor now