One of the pieces of cruft we hadn't cleaned up from the Arbitrum release was that we hadn't actually released imwss on GitHub, which meant that our versioning check was doing so against hmwss. The two main consequences of that were:
Version bumping scripts were being overly generous (i.e. changes since imwss were passed if the version was bumped in hmwss
With the release, our checks broke because the latest release has now moved from truffle to hardhat
As a result of 1, we missed a required version bump in #1259, but it doesn't appear any others.
This PR intends to fix these issues. Arguably the most meaningful change is an indirect result of the release though, which is how I'm treating upgrades in tests. For historical behaviour, I'm assuming that doing an upgrade from e.g. 12 to 13 to 16 is sufficient to test, rather than having to deploy 14 and 15 as well. This seems to work for all of our tests so far; it will need to be revisited if that assumption no longer holds.
One of the pieces of cruft we hadn't cleaned up from the Arbitrum release was that we hadn't actually released
imwsson GitHub, which meant that our versioning check was doing so againsthmwss. The two main consequences of that were:imwsswere passed if the version was bumped inhmwsstruffletohardhatAs a result of 1, we missed a required version bump in #1259, but it doesn't appear any others.
This PR intends to fix these issues. Arguably the most meaningful change is an indirect result of the release though, which is how I'm treating upgrades in tests. For historical behaviour, I'm assuming that doing an upgrade from e.g. 12 to 13 to 16 is sufficient to test, rather than having to deploy 14 and 15 as well. This seems to work for all of our tests so far; it will need to be revisited if that assumption no longer holds.