Spines currently let you put a point on each cell in the pixel grid. Allowing them arbitrary positions in floating-point space like a traditional "vector" layer is not the intention, but for some tasks they might be needed.
Consider a 2-point Spine designed to be used to indicate the location of a mirror axis. If you're on a 32x32 grid and you want to mirror it in the middle, then you need the points between. This would mean that spine points can be on any grid center, edge, or corner.
It creates some rendering questions if that Spine is reused to draw a line 1 pixel in width, but that is the business of the Shapers to have settings to resolve. It seems a better solution than telling the mirror Shaper in its properties to "lean left" or "lean right". :-/
Spines currently let you put a point on each cell in the pixel grid. Allowing them arbitrary positions in floating-point space like a traditional "vector" layer is not the intention, but for some tasks they might be needed.
Consider a 2-point Spine designed to be used to indicate the location of a mirror axis. If you're on a 32x32 grid and you want to mirror it in the middle, then you need the points between. This would mean that spine points can be on any grid center, edge, or corner.
It creates some rendering questions if that Spine is reused to draw a line 1 pixel in width, but that is the business of the Shapers to have settings to resolve. It seems a better solution than telling the mirror Shaper in its properties to "lean left" or "lean right". :-/