🚨 Potential upcoming breaking changes - v0.5.0

[zadiq]

Hi all,

I am currently working on some exciting updates in the background for v0.5.0. It is mostly around reactivity and HTML-over the wire.

My aim for labb is to provide a end-end fullstack library for developing fully dynamic webapps without the complexity of JS frameworks or complex infrastructure and BE setup whilst maintaining a lean JS footprint. I believe this can be achieved with simple primitives that don't break the flow Django developers are familiar with - no weird Python in HTML, HTML in Python, websocket requirements etc.

The updates would potentially be affecting the JS library and syntax currently used for reactivity in labb components. We could be moving away from AlpineJS to something I believe is better and inline with my goals for labb (hint -> #63). Hence, the main impact of the release would be on the usage of .x variant of labb components.

I apologies for any inconvenience this might cause. Please continue using labb components, but do limit any usage of .x variants as it could be dropped - you wouldn't need to use .x to perform reactive actions anymore. If you are not using .x in your codebase, no need to worry. Other features and API of labb should stay untouched. Reactivity would continue to be an optional feature, and labb core components will always be 0 JS by default. Any release would come with a comprehensive migration guide.

I will share more updates (including detailed reasons for the migration from AlpineJS) closer to an alpha release. Please consider subscribing to labb's newsletter to receive news and updates like this - max of 2 newsletters per month.

[LeoWelter]

consider, if it is datastar you're talking about, consider this: https://finance.biggo.com/news/202510101932_Datastar-Pro-version-sparks-developer-backlash
Alpine.js has a bigger audience than datastar, impacting adoptation. For my project members it would be a no-go

[zadiq]

Hi @LeoWelter, thanks for raising this. I had previously come across similar articles when looking into datastar, but I decided to look at things more objectively.

After looking at what the package has to offer in comparison to Alpine (plus HTMX combo), I decided it was the best fit for fulfilling my objectives in making labb a full stack reactive library (with minimal JS footprint).

I hear your point on Alpine ecosystem being larger, but the datastar's syntaxes aren't that complex tbh. I got the gist within a day, and labb's integration with datastar is abstracted so developers don't need to learn those syntaxes to get started.

Interestingly, I was easily able to replicate any of datastar's paywalled features that I needed for labb without any issues.

If the lack of AlpineJS is a deal breaker for you, I suggest:

• trying out Django cotton's official UI library - built with tailwindcss and Alpine
• or use labb components (comes with a lot of CSS based interactivity), and build any advanced interactive components with the library of your choice. labb core components will always be zero JS by default. The upcoming integration is an opt-in for a fully truly reactive full stack experience.