5-8cyls

[Philippe Hurbain]

For the arm with hand (109866), I have a need for a 5-8cyls primitive (attached). This one renders fine in LDView with primitive substitution. Opinions, objections? Other way to do that with existing prims? 
This is directly related to this discussion:
https://forums.ldraw.org/thread-23917-po...l#pid36070

[Gerald Lasser]

I do not see an issue, I made 11-48clys for the new panel.

a standalone of this would be interesting. i.e. minus the 2-4cyls, how to cal it and how to define it

[Magnus Forsberg]

For the arm with hand (109866), I have a need for a 5-8cyls primitive (attached). This one renders fine in LDView with primitive substitution. Opinions, objections? Other way to do that with existing prims? 
This is directly related to this discussion:
https://forums.ldraw.org/thread-23917-po...l#pid36070

I don't like 5-8, since it is equal to a 2-4 + a 1-8, but in this case I see the benefits.
But a half slope, like the one discussed in that thread, is better.

We need a way of naming them.

[Philippe Hurbain]

I don't like 5-8, since it is equal to a 2-4 + a 1-8, but in this case I see the benefits.
But a half slope, like the one discussed in that thread, is better.

We need a way of naming them.

Not sure the half slope is a solution here. To be scalable it needs to eliminate the cylinder padding. But then this padding introduces a t-junction with neighbor 2-4cyls.
For now I'll use the 5-8, if we find a better solution I'll change.

[Peter Blomberg]

Several duplo parts have bottom cylinders (underside studs) meeting a slope that is not going thru the axis. They would all benefit from more flexibility regarding the n-fcyls.

Regarding the proposed 3-8cyls;
To describe the two segments exceeding the 2-4cyls, one could use 1-8ncyls or 1-8ccyls for 'negative' or 'complement'. The negative would be consistent with tndis and less prone to spelling mistakes.

As for cylinders cut at any other point than the axis or the cylinder edge, one could use the inverse of the fraction in the numbering.
cyls would have an implicit 1, so cut at the edge.
cyls2 would be cut at half the diameter, i.e. at the axis.
cyls3 would be cut at one third of the diameter, i.e. one third off the high edge.
cyls1.5 would be cut at two third diameters from the high edge, i.e. one third from the low edge.
cyls1.25 would be cut 0.8 diameters from the high edge, i.e. 0.2 diameters from the low edge.

This would also mean that the 2-4cylj1x1 primitive would be equal to two mirroring 1-4ncyls back to back.

If the point in the file name can be solved, then it becomes possible to define rings with other than integer relations.