# Regression test: a picture cache slice whose raster space sits far from the
# origin (a large scroll or pinch-zoom offset) cannot round-trip its culling
# rect through the f32 scale+offset exactly, and the resulting sub-pixel loss
# must not trip the "culling rect loses part of the screen" assertion
# (bug 2073850). The slice itself is scrolled out of view, so the visible rect
# is there to keep the reftest non-blank.
---
root:
  items:
    - type: reference-frame
      id: 4
      bounds: [0, 5000001.3, 400, 300]
      is-2d: true
      transform: scale(1.125)
      items:
        - type: scroll-frame
          id: 5
          bounds: [0, 0, 400, 300]
          content-size: [400, 600]
          items:
            - type: rect
              bounds: [0, 0, 400, 300]
              color: green
    - type: rect
      bounds: [0, 0, 100, 100]
      color: red