Vojtech Bubnik OK1IAK

For us at the PrusaSlicer team, the Z coordinate is often used to differentiate the source of segments / vertices. However I believe it did not work quite reliably, I only guess that the Z was not processed correctly by the code removing zero area loops. On Fri, Feb 25, 2022 at 11:21 PM Angus Johnson angusj@users.sourceforge.net wrote: As you've noticed, yes this feature isn't currently supported. However, depending on the level of interest, I could be persuaded to reimplement the Z coordinate. (And...

Angus, thanks for your lengthy response, however I think we do not understand each other. I think our misunderstanding is caused by how each of us think of the application of the clipping library, you think of rendering, while we tend to apply Clipper in CAD like scenarios. I will try to prepare some test cases if time permits to base our discussion on concrete examples. Vojtech On Wed, Oct 13, 2021 at 1:38 PM Angus Johnson angusj@users.sourceforge.net wrote: Hi again Vojtech. Now, for performance...

Angus, I am pointing to mathematical morphology operators used for example in computer vision. Mathematical morphology - Wikipedia https://en.wikipedia.org/wiki/Mathematical_morphology Opening (morphology) - Wikipedia https://en.wikipedia.org/wiki/Opening_%28morphology%29 Closing (morphology) - Wikipedia https://en.wikipedia.org/wiki/Closing_(morphology) Morphological closing operation is used for example for filetting Fillet (mechanics) - Wikipedia https://en.wikipedia.org/wiki/Fillet_(mechanics)...

Hello Angus. As you know, we are using your ClipperLib in PrusaSlicer for a lot of heavy lifting. I have to admit though that after 5 years on the project I am still not sure about the properties and limitations of your Clipper library. Maybe the documentations needs a bit of improvement? Our software slices an object into set of layers, each layer consisting of a nesting of polygons, which is idempotent to Clipper union with the even / odd rule (meaning the polygons are already unified, correctly...

I am using C++ clipper. C++ Clipper offers two variants of Execute, one accepting list of polygons as a return type, the other a Polytree type. On Mon, Jun 14, 2021 at 10:44 AM Mikael whizpah@users.sourceforge.net wrote: Hi, using the Golang or Delphi version? In the port, there is a method called "Execute2" which has the signature: func (c Clipper) Execute2(clipType ClipType, subjFillType, clipFillType PolyFillType) (polytree PolyTree, succeeded bool) Is it something similar you are referring to,...

I have also tested with the PolyTree as a result type and I don't see any hang up either. On Mon, Jun 14, 2021 at 8:16 AM Mikael whizpah@users.sourceforge.net wrote: Hi! Thank you so much for the time and effort investigating the issue I'm having. I'm glad to hear that the Delphi version of Clipper performs as expected. I'm in contact with the person who did the Go port (Chris Tessum), and going to try and look at his issue board or recent pull requests to see if there is a version of the Go port...

For what it's worth, we see some extreme slow downs with complex contours, but I don't remember seeing an endless loop. The Vatti algorithm implemented by Angus seems to be ok, but the succeeding reduction of edge overlaps may be quite slow. Also we have found out that building the polytree during complex operations could be extremely slow, so we are first calling an operation returning a set of polygons, and we are building a polytree from the reduced and cleaned-up set of polygons by another clipper...

Just test the Z-Fill callback first. I am not quite sure it is working correctly in all cases, I burned my fingers with it. On Thu, Jan 9, 2020 at 4:18 PM Jay Tee jay-tee@users.sourceforge.net wrote: Hi Chuck, you might be able to achieve your desired result with the help of Clipper's Z-Fill Callback http://www.angusj.com/delphi/clipper/documentation/Docs/Units/ClipperLib/Classes/Clipper/Properties/ZFillFunction.htm . Just assign an ID to each input polygon and store it in the Z component of all...