It looks like `BytesPerRow` (the stride) has some kind of alignment requirement (testing on M1), that doesn't let us just use the width. So for this to work, softbuffer needs to expose the stride, and the user needs to deal with that...

This doesn't seem to be showing the right thing even when width matches stride. Not sure why.

Division is more expensive than multiplication, and this doesn't seem to
be optimize away, so iterating over `(0..height)` and `(0..width)`
provides better performance.

It's good to have the example show the best way to write code using
softbuffer, and this helps when using examples as a benchmark. This will
also be needed if Softbuffer supports formats where `stride` isn't just
`width * pixel_size`, and needs alignment.