mån 2022-12-19 klockan 13:04 +0000 skrev andy pugh: > On Mon, 19 Dec 2022 at 11:27, <l...@pibf.de> wrote: > > > I do not know what seicanon.cc is for! > > saicanon is related to the "stand alone interpreter" which I think > is, > in turn, related to the G-code preview and (maybe) some of the tests. > https://linuxcnc.org/docs/stable/html/code/rs274.html > (The manpage above tells you very little) > > > So first arc() has to be expanded to UVW planes? > > Ideally, yes. Though it isn't clear how you would do that if you were > also trying to cut an arc at the other end of a foam cutter in the XY > plane. > > The issue with arcs is that it isn't clear how to define the centre > point for UVW arcs. > > This is moving well away from your original pull request, so don't > consider the following to be a wish-list for that. But I think it is > an interesting point of discussion. > > For foam cutters it would be really nice to be able to cut > simultaneous arcs and NURBS / Splines on each end of the machine. But > it is difficult for a number of reasons: > 1) How do you define the arc centres for UVW arcs? (especially if you > are _simultaneously_ doing an XYZ arc.) Three points with three in the (x,y,z) space will define an arc in three dimensions. Then there is three degrees of freedom for the tool direction at each point or two for a rotating/"round" tool if I think correct.
> 2) Synchronisation: Arcs are not too hard, each end of the wire gets > to the end of the arc at the same time. It's not so clear with > NURBS/Splines. Do you aim to hit control points simultaneously? Or > finish the entire curve simultaneously? > 3) Plane definition. Currently we have G17,18,19,17.1,18.1,19.1. > > https://linuxcnc.org/docs/stable/html/gcode/g-code.html#gcode:g17-g19.1 > We would need to find a way to define "simultaneous planes" or > possibly just allow UV arcs if the XY plane is selected (and run them > simultaneously). (possibly it makes sense for compatibility to make > this the default G17.1 mode, as this doesn't alter anything that > currently works) Should be possible to define a "flat" plane in a three dimensional space with three vectors. Nicklas Karlsson _______________________________________________ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
