Diego Guidi

my2cents: any improvement is weel accepted, of course. but as far as I understand, you are planning to extend a library, but the library itself can not be supported by...

> it would still make sense to deviate why not simply an overload like: `AsText(bool omitZM = true)` so consistency with JTS is preserved

Not really. I see two main issues: 1. as a user, I'd expect that **a ZM geometry.AsText writes ZM values** as default behavior, as @bricelam said 1. as a JTS/NTS...

I don't think the two operations can be comparable, but I may be wrong. TestBuffer creates the edge/node graph for the entire geometry, actually twice because the buffer fails with...

What I mean to say: the algorithm does an entirely different thing for multi geometries; it builds the entire nodes/edge graph and then compute the buffer. If your workaround can...

as a general suggestion, use always the geometry factory to build geometries, and not the constructors

I quickly checked and the code you suggested isn't called at all using a unit test against develop.

> Thanks for suggestion, what is the advantage of doing that? see the [Getting Started](https://github.com/NetTopologySuite/NetTopologySuite/wiki/GettingStarted) page

> EnforceGabriel() is calling _splitFinder.FindSplitPoint(seg, encroachPt) that is an instance of `NonEncroachingSplitPointFinder`... see [this line](https://github.com/NetTopologySuite/NetTopologySuite/blob/develop/src/NetTopologySuite/Triangulate/ConformingDelaunayTriangulator.cs#L70). howewer, the problem still persist. this is the code I used ``` using System; using...

looks that all `LineSegment` computations [does not support Z values](https://github.com/NetTopologySuite/NetTopologySuite/blob/develop/src/NetTopologySuite/Geometries/LineSegment.cs#L270) I just fixed **this specific test** with this changes in the `LineSegment` class: ``` public Coordinate PointAlong(double segmentLengthFraction) { double...