(2020-08-17, 20:06)Roland Melkert Wrote: You could make a template for this in the current 1.6 version I think, you'll need.
A static 'knot.dat' for use as a cap.
A single twist of the 2 strings spread over 4 or more .dat's (so it is more bendable).
The rolled portion could also be a static cap.
Whit this you could make a deform (HQ) template which, I hope, would approach the result you've got doing it the hard way. But the way you have it now will probably be the higher quality version.
If the large file size is an issue you can also use the 4 segments to do LQ plotting (overlapping semengts) but I think the roll of the 2 strings will make that very noticeable.
I am very pleased Roland, that you comment on this personally.
As you can see for yourself, if you call up the following PT link,
https://ldraw.org/cgi-bin/ptscan.cgi?q=5...ope=header
the segment sub-part (56823s02.dat) consists of 3, not two plys , and these are only indicated in their shape.
Of course, I also had the considerations about the division, as you are just describing it - but the core problem of achieving a continuous and clean twist, with a simultaneous directional course throughout, fails with the current possibilities, with our standard resources.
At least for me, which is certainly due to the fact that I am still quite virginal with your LDCAD software, because I have only been using it for a relatively short time.
(BTW - LDCAD is world class - Chapeau!)
The topic - just discovered to my surprise - was already there: "progressive roll" & "twisted strings"
https://forums.ldraw.org/thread-24138.html
With C4d I only have the possibility to get to ldraw via stl2dat and to transfer the basic elements individually with my python script. A specific subpart recognition of the LDraw library is not possible (which is not entirely true, stl2dat has a cylinder recognition, but only works if you have used solidworks)
I allowed myself the fun of doing the whole thing in C4D with a simple square as the base segment - then "stl2dat create all edges" .... and last but not least, used rectifier to display the (1 + 1799) cloned squares as recte3 objects.
The object names changed to 56823s02 ... Et voilá!
But ... actually not a stupid idea ... only the alignment was not clean - very arbitrary .. in parts even chaotic.
If you like, have a look at the Help! section in 56823s02.dat - all (if I haven't forgotten anything) essential information is listed there in detail.
Do you happen to have C4D available? - then I could send you the file for deeper insights.