Update to readme to fix #14

README.jmd

@@ -29,7 +29,7 @@ rdata = rand(StableRNG(1), length(triangles(t)))
 # work over all triangles
 for (i,tri) in enumerate(triangles(t))
   tripts = pts[collect(tri)]
-  poly!(ax , tripts, color=[norm(mean(tripts))*(1+0.5*rdata[i])],
+  poly!(ax, tripts, color=norm(mean(tripts))*(1+0.5*rdata[i])*[1,1,1],
     colorrange=(0,3), 
     colormap=map(c->RGBAf(c.r,c.g,c.b,0.9), Makie.to_colormap(:matter)))
 end
@@ -149,12 +149,15 @@ of floating point computations.
 a good deal of this functionality is provided by the package `VoronoiCells.jl`. 
 - The `DelaunayTriangulation.jl` package that 
 uses [`ExactPredicates.jl`](https://github.com/lairez/ExactPredicates.jl) to implement various 
-computational geometry tests. 
+computational geometry tests. This package is under active development and implements a number 
+useful constrained Delaunay triangulations. 
 
-In comparison, the `Delaunator.jl` package seeks to mirror the javascript Delauantor packages that give good
+In comparison, the `Delaunator.jl` package seeks to mirror the javascript d3-delaunay codes that give good
 enough triangulations for many pixel-level graphics applications and are fast for 2d problems, rather than those that 
 might be suitable for those with computational geometry applications that need better guarantees
-(although we do hope to improve this in the future). 
+(although we do hope to improve this in the future). The underlying Delaunator javascript library does use the
+accurate primitives, although these are relaxed slightly in the d3-delaunay usage -- especially in terms of the
+Voronoi cells. 
 
 > This readme is auto-generated by weave from `README.jmd`
 ```julia; eval=false, echo=false 

README.md

@@ -12,7 +12,7 @@ A port of [Mapbox's Delaunator](https://github.com/mapbox/delaunator) to Julia.
 
 The Delaunator algorithm computes a simple 2d triangulation of an arbitrary set of points in the plane _quickly_. This package provides a Julia implementation of the algorithm along with a number of supporting routines that operate on the Delaunator data structures. 
 
-On a 2020 M1 Macbook Air, the Deluantor.jl package will triangulate a million points in about 500ms including computing circumcenters for the triangles for the closest site diagram. 
+On a 2020 M1 Macbook Air, the `Delaunator.jl` package will triangulate a million points in about 500ms, including computing circumcenters for the triangles for the closest site diagram. 
 
 ```julia
 using GeometryBasics, Delaunator, StableRNGs, CairoMakie, Statistics, LinearAlgebra
@@ -29,7 +29,7 @@ rdata = rand(StableRNG(1), length(triangles(t)))
 # work over all triangles
 for (i,tri) in enumerate(triangles(t))
   tripts = pts[collect(tri)]
-  poly!(ax , tripts, color=[norm(mean(tripts))*(1+0.5*rdata[i])],
+  poly!(ax, tripts, color=norm(mean(tripts))*(1+0.5*rdata[i])*[1,1,1],
     colorrange=(0,3), 
     colormap=map(c->RGBAf(c.r,c.g,c.b,0.9), Makie.to_colormap(:matter)))
 end
@@ -197,11 +197,14 @@ of floating point computations.
 a good deal of this functionality is provided by the package `VoronoiCells.jl`. 
 - The `DelaunayTriangulation.jl` package that 
 uses [`ExactPredicates.jl`](https://github.com/lairez/ExactPredicates.jl) to implement various 
-computational geometry tests. 
+computational geometry tests. This package is under active development and implements a number 
+useful constrained Delaunay triangulations. 
 
-In comparison, the `Delaunator.jl` package seeks to mirror the javascript Delauantor packages that give good
+In comparison, the `Delaunator.jl` package seeks to mirror the javascript d3-delaunay codes that give good
 enough triangulations for many pixel-level graphics applications and are fast for 2d problems, rather than those that 
 might be suitable for those with computational geometry applications that need better guarantees
-(although we do hope to improve this in the future). 
+(although we do hope to improve this in the future). The underlying Delaunator javascript library does use the
+accurate primitives, although these are relaxed slightly in the d3-delaunay usage -- especially in terms of the
+Voronoi cells. 
 
 > This readme is auto-generated by weave from `README.jmd`