JaleelB
[question]: Why release as a package?
Question description
The reason behind shadcn is to have ownership of the code and make all components lives in your project, so just wondering the decision to make this as a package instead of copy/paste component? also need to be highlighted that this package uses additional dependencies like react-easy-sort.
It will also have a duplicate components code since it also bundles the components like button, popover etc... another thing is it breaks the pattern of composable components because of the abstraction, thats why we have this css props all over.
We can have an inspiration from shadcn vue tags-input
It could be nice to have a cli command instead, to properly align to the theming from shadcn specially when not using cssVariables
Thank you for your thoughtful questions about my tag input component. You've raised some valid points that I've been considering:
- While shadcn's copy/paste approach works well for simpler components, my tag input has grown significantly in complexity and features. Packaging it allows for easier maintenance and updates across projects. This decision was made quite a while before the new shadcn cli
- I plan to remove the react-easy-sort dependency with my own implementation in the near future to reduce external dependencies.
- I acknowledge the concern about bundled component duplication. With the new shadcn cli, i plan on configuring an option to allow others who want control of the code to just be able to copy it over via the cli
- I'm revisiting the component's API to see if I can make it more composable while maintaining its advanced features.
Improving Emblor has been on my roadmap for a while. I've been planning a comprehensive overhaul of the component. Here's what I'm considering:
- A complete redesign of the API and component structure to make it more modular and composable.
- Potentially making the component unstyled, which would align more closely with the shadcn philosophy and increase flexibility for users.
- Reevaluating all current features to ensure they're necessary and considering a possible core + plugin architecture for advanced features.
- Improving performance and optimizing for large datasets.
The project started out as something small and grew over time as I added more features, and I pretty much slapped features here and there. Now, I'm taking the time to step back and rethink the entire approach. This redesign will aim to address the concerns you and others have raised, while maintaining and improving upon it
I appreciate your feedback, as it helps validate the need for these changes. As I work on this overhaul, I'll be sure to keep the community updated on the progress and may reach out for additional input.
If you have any specific feature requests or suggestions for this redesign, I'd be glad to hear them. Your insights could be valuable in shaping the future of Emblor.
I feel similar here. ive gone and kind of stolen the code out to put in my shadcn components out of need, but I do want to use the new stuff properly when it is released.
@JaleelB Do you have any idea when you will be working on v2 or will it be ready when its ready 🙃 .
I feel similar here. ive gone and kind of stolen the code out to put in my shadcn components out of need, but I do want to use the new stuff properly when it is released.
@JaleelB Do you have any idea when you will be working on v2 or will it be ready when its ready 🙃 .
Hey @pcfreak30 I am currently working on v2. Planning to release a beta or full v2 by end of month
I feel similar here. ive gone and kind of stolen the code out to put in my shadcn components out of need, but I do want to use the new stuff properly when it is released. @JaleelB Do you have any idea when you will be working on v2 or will it be ready when its ready 🙃 .
Hey @pcfreak30 I am currently working on v2. Planning to release a beta or full v2 by end of month
Checking in 🙃