Add transform_marker attribute to meshscatter

[ffreyer]

Description

Fixes #3323 (with changed default)

This pr adds the transform_marker attribute to meshscatter, which controls whether model applies to the scattered mesh or not. I.e. with transform_marker = false the marker used with meshscatter does not get affected by model, only the given positions do.

Example:

f = Figure(size = (300, 500))
a1,p1 = meshscatter(f[1, 1], Rect3f(Point3f(0), Vec3f(0.3)), transform_marker = true)
a2,p2 = meshscatter(f[2, 1], Rect3f(Point3f(0), Vec3f(0.3)), transform_marker = false)
scale!.((p1, p2), 5,5,5)
f

The current master behavior is transform_marker = true. I changed this to false to fix #3323. I also think it generally makes more sense, because markersize, rotation and the position argument of meshscatter already do what model does, on a per marker basis.

Example from Issue (only changed color):

Other changes

This pr now also includes some fixes to the meshscatter transformation/projection pipeline:

• fixes order of float32convert and model in CairoMakie projection functions. This should be float32convert * model * transform_func(positions) (this also affects other primitives)
• CairoMakie's draw_mesh3D was always applying transform_func to vertices, which does not match the GLMakie behavior for voxels and meshscatter
• CairoMakie's draw_mesh3D did not work without normals
• In GLMakie and WGLMakie the mesh from meshscatter did not consider float32converts at all. That meant that the mesh will be incorrectly scaled if f32c includes scaling.

TODO:

• test transformation pipeline for meshscatter more thoroughly (model x transform_marker x transform func x float32convert)

Type of change

• Breaking change (changes default behavior of meshscatter)

Checklist

• Added an entry in CHANGELOG.md (for new features and breaking changes)
• Added or changed relevant sections in the documentation
• Added unit tests for new algorithms, conversion methods, etc.
• Added reference image tests for new plotting functions, recipes, visual options, etc.

[MakieBot]

Benchmark Results

SHA: 00af282008ee877613bb67a1a622ff6b570010b2

Warning

These results are subject to substantial noise because GitHub's CI runs on shared machines that are not ideally suited for benchmarking.