Fix dupe issue
diff --git a/extensions/NV/NV_blend_equation_advanced.txt b/extensions/NV/NV_blend_equation_advanced.txt index 25c398c..bfe1115 100644 --- a/extensions/NV/NV_blend_equation_advanced.txt +++ b/extensions/NV/NV_blend_equation_advanced.txt
@@ -1676,19 +1676,7 @@ include a special case for Cb==0 while others do not. We have added a special case there as well. - (35) How should ClipColor() behave if the inputs cause a divide-by-zero, - which may happen when all color components are the same value and - outside the range [0..1]? - - For example, if the color parameter to ClipColor() has all color - components as equal and above 1.0 then maxcol==lum and the term: - color = lum + ((color-lum)*(1-lum)) / (maxcol-lum); - will be a divide-by-zero. - - RESOLVED: The HSL blend equations are not well defined when used with - ill-conditioned inputs. The inputs must be in the range [0..1]. - - (36) For "HSL" blend equations, the blend equation involves a clipping + (35) For "HSL" blend equations, the blend equation involves a clipping step where colors may be "clipped" if the blend would produce components are outside the range [0,1]. Are there inputs where this blend could produce ill-defined or nonsensical results? @@ -1713,7 +1701,7 @@ 10 02/14/18 pdaniell Fix ClipColor() equation where in the "if (maxcol > 1.0)" body the "(color-lum)*lum" term should have been "(color-lum)*(1-lum)". - Also add new issue 36 for the case where the + Also add new issue 35 for the case where the inputs to SetLum() are outside the range [0..1] and could cause a divide-by-zero in ClipColor().