Li-Lun Lin

That should be possible IMO. I'll try working on it later today. > avoid potential breaking changes when bumping the versions of the Rust toolchains If this is the only...

Hello @hydai, I've been researching release pipelines recently and came across some interesting tools that I'd like to discuss with you: - [`knope`](https://github.com/knope-dev/knope): This tool has fewer users and stars,...

Your points are totally valid, and I’m good with either approach. I suggested knope because its pattern is widely used in the JS ecosystem—or rather, it's bringing a good pattern,...

Let me make sure I understand correctly: we can have a function that takes in the JSON (likely parsed as structs) and returns the desired URL/hash for downloading the binaries....

@hydai I’ve been investigating OS/architecture detection and was wondering if runtime-based detection is necessary. Since users won’t be running `wasmedgeup` on an OS/architecture different from their host machine, I’d argue...

I'm curious—do you remember the specific case? I feel like something like `args.os.unwrap(compile_target_os())` should handle this correctly, though.

@hydai For now, they don't support customization for the commit message used by knope-bot yet. Maybe we could ignore the commit made by knope-bot in commitlint as a workaround.

I'm not sure what the bot is doing exactly, if it's performing rebases then rewriting the commit message by ourselves would not have a desired outcome.

> Never mind, let's bypass the commitlint for now as a workaround. Got it, I'll work on this. > Would you like to trigger the release process again? Sure! That...

Oh, I see. But TypeScript language server does not error if I use the extensionless module specifier either. Why is that the case?