Is it possible to do a visual transform without triggering a recomposition? E.g. say I want to use `...

# compose

[Brian G]

Is it possible to do a visual transform without triggering a recomposition? E.g. say I want to use Box(Modifier.offset(x=x, y=y)) { Content() } but update x and y very frequently, while the contents of the box never changes. Right now I see the content get recomposed often, any way to avoid that? Is that in "future optimization" territory?

[Alexandre Elias [G]]

yes, the variant of offset that takes a lambda, like Box(Modifier.offset { IntOffset(x, y) }) exists specifically to avoid unnecessary recompositions

[Brian G]

Awesome! I actually saw that in autocomplete and was like "why would this take a lambda? that's weird" 🤦‍♂️