Henning Koch

Results 67 comments of Henning Koch

Closing this as we couldn't achieve reproduction in over a year.

The use case described in this use is solved by Unpoly 3.5.0. See [Handling changes in frontend code](https://unpoly.com/handling-asset-changes).

This can also be reproduced in Unpoly 3.0.0-rc2: https://maze-cyclic-pen.glitch.me/ Note that putting a secondary submit button like "Cancel" first in the DOM is also problematic for keyboard users. Pressing return...

> However, one step of this wizard has submit buttons earlier in the form for adding additional items to an expanding list, so I'm not sure how I'd work around...

I like it. It could be implemented by parsing and merging an `{ url }` option [here](https://github.com/unpoly/unpoly/blob/040a4bfc66437c81eabd567cdf31f4f737c37645/src/unpoly/classes/form_validator.js#L156C1-L161). Do you think `[up-validate-url]` or `[up-validate-action]` is the more intuitive name here? Arguments...

> -> Should up-validate be implicitly enabled? No. It's a modifying attribute of `[up-validate]`. > How should the behavior be on nested up-validate-url? We cannot easily allow nesting `[up-validate-url]` as...

Closing as duplicate of #513.

I can share some pains, design decisions and experiences we had with relative references over the years. This is from my work with [Unpoly](https://unpoly.com/), an HTML extension where [links update...

If you want to build this: The release process is automated in the [`Rakefile`](https://github.com/unpoly/unpoly/blob/master/Rakefile). The task that makes a release is `release:process`.

We generally don't support compilation for temporary state classes like `.up-loading` or `.up-currrent`. A compiler/destructor would also not be the best tool here, as there are cases where `.up-loading` is...