rebuild components

[dakom]

@ismellike latest wasi-build is still failing

[ismellike]

adding note might be osx specific since this branch ran CI fine #1049

[ismellike]

Can you try just wasi-build again? I have no idea how it was working before, but I went through and tried to match as best as possible with improvements. I am seeing this note on cw optimizer https://github.com/CosmWasm/optimizer?tab=readme-ov-file#notice. Still the greater question is how amd64 passes CI but arm64 build doesn't

[ismellike]

A diff between the two wasms shows a difference in wit versions used ❓ ❓ no idea how that's possible since all wit files are tracked and mounted with wit_bindgen using them in our example components :

<   (import "wavs:types/chain@2.0.0" (instance (;0;) (type 0)))
---
>   (import "wavs:types/chain@1.3.0" (instance (;0;) (type 0)))

[ismellike]

can you give just wasi-build a run @dakom got a couple updates in on wasi-builder.

also fyi, needed to pin alloy-json-rpc

[dakom]

@ismellike

good news: it built and passed the e2e tests 🚀

bad news: it created a diff of components, point of this is to have reproducible builds so that should not happen... if it's unavoidable for some reason, is what it is, but not sure why there would be a diff

[ismellike]

@ismellike

good news: it built and passed the e2e tests 🚀

bad news: it created a diff of components, point of this is to have reproducible builds so that should not happen... if it's unavoidable for some reason, is what it is, but not sure why there would be a diff

amazing! @dakom can you try once more, and I think we call it there regardless? Maybe we open an issue over on the wasi-builder repo for determinism.

[dakom]

Still getting the diff, but it's building and passing tests, imho good to merge - agree we can move that issue to the builder :)