@Tyriar Would it be worth to further track down the reason for the lines numbers provided being off?

I'd suspect a race condition from resizing as well, but doesnt that mean that by opposite resizing the numbers are also off? This ofc would not run into an NPE, but still apply wrong numbers for whatever reason _updateModel was called, prolly missing model updates (eg. from resizing - skipping updates on higher/further down lines).

I don't think the opposite is a concern as a full screen update would have been requested after the resize, I guess we could remove this workaround and then validate the end value before using it in render service or debouncer?

Reopened the issue to do that, that way we would still get the exception if something else is happening