Good catch. So I think I've maybe fixed this by changing the way we count old kernels, to being "version" based. It'll need some testing.
Good catch. So I think I've maybe fixed this by changing the way we count old kernels, to being "version" based. It'll need some testing.