5 replies

1. tbeseda 21 min later

   @zachleat I'm interested what your thoughts are here?
   So far, I’m just not seeing it. I like events and I understand their problems. Signals will have its own crop of antipatterns.
   But this isn't a well-researched first impression. Maybe I'm not envisioning it right?

   1. Zach Leatherman 26 min later

      @tbeseda mostly I’m excited to have a standard feature to help implement Eleventy’s Computed Data feature.

      (and it looks like they’ve solicited feedback from a bunch of clientside framework folks too)

      What’s not to love?

      1. tbeseda 42 min later

         @zachleat I think the linked article has a better explainer than the spec (to be expected!)
         I see the appeal. I suppose I'm too conservative around adding new primitives when there's a similar one in the standard.
         I should probably just write Erlang lol

      2. Ben 48 min later

         @zachleat @tbeseda a huge part of the “value” for React is the idea of a frontend framework where changing data directly updates the UI.

         Among other things, this API seems like it would open up a very large number of doors for achieving that without React.

         It also really neatly pulls in functionality for things like observables.

         1. tbeseda 51 min later

            @sangster @zachleat great point. The ability to bind a piece of the DOM is an advantage over events where one would need to handle updating the content. I can get behind that.