Recently helped someone with a Spark project and encountered “reactive JS” for the first and hopefully last time. Never minded using JS but happy to admit that it’s a dumpster fire design-wise…one that Spark engineers apparently decided didn’t have quite enough petrol on it.

Rxjs is pretty common in some environments and it’s not so crazy, the only thing is that automated testing was written by someone very drunk

@piratefox I don't know that it's exactly the same as RxJS (could be, just reading up now), they also call it Declarative JS in the docs - with this you can't, or can very rarely, directly set a property eg position.x, you subscribe position to a signal and have this control it independently. Trouble is, the project required lots of direct setting of values, which made doing this an absolute round-the-houses.

I do see some appeal in the idea in certain cases, perhaps there was a neater solution than the one we found - and I get that Spark is a platform designed to lock non-coders into the Meta ecosystem as readily as coders, this is perhaps a lot easier to grasp or copy/paste than lower-level logic - but it felt like an Apple-style "you ARE doing it this way" moment to me.