Why built Unpoly2 on top of htmx?

[Thomas Güttler]

I think building Unpoly2 on top of htmx would

reduce the time and energy you need to support Unpoly2.

What do you think about this?

Regards,

  Thomas Güttler

[szabesz]

Joining forces would make sense, I think. As for the technical aspects of the idea, maybe the authors could at least discuss it, which would be great.

[adam.me...@gmail.com]

I'm not sure I agree. The overlap of htmx and Unpoly is only superficial; a couple declarative statements of interactivity in the DOM. I'm not Henning, but my guess is that specific functionality only encompasses a small portion of the time consumed for Unpoly.

The real complexity is probably the complex interactions that v2 supports. And being mindful of project history, v1 has been chugging away perfecting with few commits against the project as of late. Possibly a testimony to how it's been architected and what level of maintenance it's required.

Also, htmx is predated by Intercooler; both Unpoly and Intercooler are similar in age (started within ~ 1 year), so if Unpoly had chosen Intercooler back in 2014, there's a good chance we'd still be using Intercooler and have a jQuery dependency. Not being dependent on any library provides flexibility should direction change course (and let's be honest, it's probably guaranteed in the frontend ecosystem, ala the decline of jQuery usage).

And with all that said, Unpoly 2 is complete. At least in its current iteration, minus docs. Rewriting would be pointless at this time.

Adam