Mateusz Burzyński

Could you tell me what problem are you facing?

We should "inherit" it: - [here](https://github.com/statelyai/xstate/blob/94c884103eb928938dbc16284822bc707f0df584/packages/core/src/spawn.ts#L83) - and [here](https://github.com/statelyai/xstate/blob/94c884103eb928938dbc16284822bc707f0df584/packages/core/src/actions/spawnChild.ts#L61) Alternatively, we could try to "inherit" it through the `system`.

This isn't fixable since this relationship is, indeed, circular and TypeScript has to resolve it eagerly. One side has to be typed here without the `typeof` query to avoid this...

To be fair, this isn't a v5 problem. This was always quirky/incoherent. Even if you were able to call `child.stop()` in `assign`... that wasn't the right thing to do. `assign`...

Ideally, you wouldn't need `isDatePickerActor` at all there. Its type should already be derived from whatever you have defined in the `setup`, see an example [here](https://www.typescriptlang.org/play?#code/JYWwDg9gTgLgBAbzgZwKYwK5jgXzgMyghDgCIAPZGAQxlVIG44B6ZuGATzFWQC44ArADphAJgBQoSLERwMaAIIBjGNFwEiJUgAFKNOsyipqKxizadufOABYhABiEBmceNTlp8JRAB2VOD6oAO7IAArU3FAAsiYAFsCBcAC8KOhYABQI4nDsXDz8CHjUyIjZOXDePnTkMAVl5ezAMAA2+SgwUAkA5gDaALoM9biDDagAbqhVfEM5AD6lDYu53PykACIAogAyGwAqG4wzi8AAJvxUnT5dI0u4R-NZt+WWqKsKa2uHTzkwTa3nHW6NyWOBuOAANOIcABKIRKIy0VAxJTxQKZMqVaq1WS-Fptfq4SE5BJNYDUZqrU6tUhE9qI6yPcpU16ICFlNkwwZuDzQLy+fwAIwgEAA1sjUahkqlMGB0T88gyiiVGTlMe5sSrnn82hduv1gcMhuNJjBpksHkcGi9Vpsdvsvt9TgDLtdLXdzQtvtayO9Pgalrj-u0Xf7yjh7ji8qsAMIACQ20YA0gB9ADyAFVdmYncHuoachzYfDjHRxQlUHKKr4sQVGnjrAS2cSfKTyZSTtTaVR6XUGsyCk3CVDoVzKv4wNQjFUy4kUmgZZWXoq4MVPar4s0TlPe4shaLVnuRQ6GoEQqtT8hj-mhzkTKooAyhof+IeZ6haTkL-wL+FIm-aYWcIIqWcTlpknKuO4nhVn48AbOQ1DgK0UrpNCyQAHyemO8A9MgPgRMgsQQDAfRSvIqDKPe6QTlOMBviOGL8vAh5SnhBFETAcIblukxCIeNysA0AB6AD8rg5EYmBQD4AQYM0zSDKC4hAA) > The union of...

I find it slightly weird that this would introduce a lack of symmetry between [the proposed `spawn` action](https://github.com/statelyai/xstate/pull/4329) - that action currently *doesn't* support spawning of multiple actors at the...

v4 didn't allow dynamic IDs on `invoke` but I don't see a reason why we shouldn't allow this in v5. `spawnChild` already supports this and both use the same implementation...

This turned out to be slightly more complicated than I anticipated since we were relying on those particular IDs to be static. This bit of the implementation is funky -...

In v5 we avoid creating redundant snapshot values when there is no change in the snapshot's value as a whole. A side-effect action fired in response to an event doesn't...

> I think a new beta release of xstate inspect is required. Yes, this. v5 is a work in progress though and the tooling around it will be released once...