Reactivity Transform

[ghost]

You guys did a real mess with this feature. If it was meant to be experimental, then you shouldn't have include it in the core. You could've also made installing it with the core optional via a flag when installing as it's experimental.

[edison1105]

maybe you need this tool :https://github.com/edison1105/drop-reactivity-transform

[LinusBorg]

What would that changed for you? What is the issue you are facing now?

• If you want to migrate away from it because you used it expecting it to become stable, and now feel stuck - see edison's reply.
• If you want to keep using it after we remove it: https://vue-macros.dev/features/reactivity-transform.html

Both ways are documented in our docs.

You could've also made installing it with the core optional via a flag when installing as it's experimental.

I mean ... you actually do/did have to enable it explicitly with a config flag.

So I'm not sure what making it an optional install would have changed? What's the "mess" that this would have prevented? Seriously interested in understanding your issue.

[LinusBorg]

The fact that you get this feature enabled by default when starting a new project is what I'm complaining about.

To the best of my recollection, it was never "enabled by default", or was it?. You had to go into your vite config and enable it manually, with a flag. (If we had that flag enabled in early versions of create-vue, then that would have been an unfortunate choice indeed, but don't think we did?)

Then you also had to read the docs to understand how to use it, which clearly stressed its experimental nature.

Would a separate install have you kept from using it, where the manual config and "experimental!" warnings in the docs didn't provide a high enough barrier? Is that the point?

I mean if some feature was meant to be experimental, then it's not a good idea to have it get installed with the core by default, that should've been optional.

Well, it's all about trade-offs, as it always is. When a framework releases an experimental feature, the explicit goal is for people to try it out, so we can get hands-on feedback.

1. making that too complicated will result in fewer people trying it, and less feedback on the experiment as a result.
2. Making it too "easy" risks users using it in their production apps "by accident" and then being disappointed if it doesn't become a stable feature and is instead removed.

Using an opt-in config flag and docs that make the the experimental nature of a feature clear is a common middle ground many frameworks and libraries use for experimental features.

Having to go about using a tool just to migrate away from it is not practical.

Not sure how this relates to the original criticism. Wether or not this feature would have come in a separate install, isn't it better to have a tool to help with migrating, than not having that, now that we will drop it?

Or do you want to continue using it? just install the vue-macro plugin.

[LinusBorg]

Great that this misunderstanding could be resolved. Just a small suggestion: Next time, maybe reconsider before opening a discussion where the first words are "You guys did a real mess". That would be great.