Sync changes from upstream repository

Co-authored-by: Andrew Le Bihan <andy@mcneel.com>
Co-authored-by: Dale Fugier <dale@mcneel.com>
Co-authored-by: Dale Lear <dalelear@mcneel.com>
Co-authored-by: David Eränen <deranen@gmail.com>
Co-authored-by: Greg Arden <greg@mcneel.com>
Co-authored-by: Jussi Aaltonen <jussi@mcneel.com>
Co-authored-by: Lowell <lowell@mcneel.com>
Co-authored-by: Mikko Oksanen <mikko@mcneel.com>
Co-authored-by: Pierre Cuvilliers <pierre@mcneel.com>
Co-authored-by: Steve Baer <steve@mcneel.com>
Co-authored-by: chuck <chuck@mcneel.com>
Co-authored-by: piac <giulio@mcneel.com>
Co-authored-by: wfcook <wfcook@mcneel.com>
This commit is contained in:
Bozo The Builder
2021-10-14 07:57:14 -07:00
parent 2b96cf3142
commit 95dde7ec84
50 changed files with 21418 additions and 882 deletions

View File

@@ -168,7 +168,7 @@ bool ON_PassesNanTest()
const double zero = 0.0;
const double one = 1.0;
// nan != * and * != nan should alwasy be true
// nan != * and * != nan should always be true
const bool b_NE_test
= nan1 != nan1
&& nan1 != nan2
@@ -2566,7 +2566,7 @@ double ON_SolveNxN(bool bFullPivot, bool bNormalize, int n, double* M[], double
// The Xdex_buffer[] hoo haa is here to avoid an potentially time
// consuming call to heap services when the matrix is small.
// When n > 64 the numerical portion of the computation is
// long enough that the time to call onmalloc() is negligable.
// long enough that the time to call onmalloc() is negligible.
// (When n > 10-ish, this calculation is likely to return junk
// unless you have a special case matrix, in which case this
// function will be much slower than one that is designed to