Roadmap for the next versions

[wuda-io]

Hello guys,
following changes are planned in the future:

Drop TypeScript and use JSDocs

Drop TypeScript and use JSDocs instead, due to saving dependencies and
intermediade transpilation step and also good support in VSCode. Also we have already many comments in the code,
which later can also be utilized to generate the documentation.

Rewrite in ESModules

Rewrite to ES Modules to use the library on node and in the browser. So we can import only components we need via import(), in the browser and also in node. To support require() or CommonJS Module we need to transpile from ES Modules to CJS which can be done by Rollup. Check out: https://stackoverflow.com/a/75348391/8830502

Dynamic Component creation and nesting components

Add a static create() function for each component which can be used to dynamically create Components on the server and in the browser. The create function has to return a html string. Dynamic events can be handled then on top of the htmlstring in the client/browser side. Also this can be used to make configurators for each component and documentation.

Create more Components and also build components upon components. This also requires dynamically creating components.

Drop SCSS and use CSS

Drop SCSS and use CSS directly. (Probably not that soon)

What do you think about this changes?

[danice]

About the JSDocs:
I don't like it. In fact I found it awful: too much verbose, less clear and much harder to write than "classic" TS (because TS considers JSDocs as an alternative annotation format). I don't think the advantages justify the work to change and maintain this.
For documentation I suggest checking TSDoc.

About "Rewrite in ESModules":
I really don't understand this. We are already using ES modules since the move to TS I did more than a year ago.

Drop SCSS and use CSS:
Maybe, but this needs more explanation. I suppose you refer to that new CSS features make scss not necessary, right? (or are you thinking in using tools like PostCSS?)

---

I should add that I consider materializecss a collaborative project and I don't think is correct to just say "following changes are planned in the future"... with a previous discussion with the community, more necessary in this case as these affects to the core of the project.

[wuda-io]

I agree with you. Maybe my sentence was a bit directly formulated. I dont want to make crazy changes, but
I suggested them and wanted to discuss. Thanks for your reply. The problem I had was the following:

I tried to import materialize into node as ESM and it was not possible.
It is cool that we can use Materialize in the browser, but I also want to use it in the backend to generate code examples.
For example I want to dynamically create a Modal Window in the browser and open it.

So I found a solution with minimal changes to run in browser AND on node. I will make a PR in the upcoming days.
I also want to get rid of Grunt and Webpack in the Core-Library and we can use Rollup which works really well and is
also used under the hood of Vite.

[IvanPostu]

My suggestion would be to completely rename the project differently from the original one because, at the moment, it is difficult to search for information, and navigating the internet creates confusion because some resources point to the current project while others point to the old one.

Here is an example:

I also bet that seeking solutions on Stack Overflow can be tricky as well.