It may be reasonable to extrude by rotation about the Z axis, but it is not at all obvious. Rotation about the Y axis would be more obvious and just as useful. I think the documentation should explicitly state that the 2D object must be in the XY plane and gets rotated by 90 degrees into the XZ plane before being extruded. Thus, positive Y values become positive Z values in the extrusion. As I have only just started using OpenSCAD, I thought I'd ask in the forum before offering to edit the documentation. -- View this message in context: http://forum.openscad.org/rotate-extrude-issue-tp7442p14807.html Sent from the OpenSCAD mailing list archive at Nabble.com.

>I think the documentation should explicitly state that the 2D object must bein the XY plane All 2D objects are in the XY, how could they be otherwise, there is no Z in 2D? On 28 November 2015 at 20:46, ChrisLT <clusbytaylor@gmail.com> wrote: > It may be reasonable to extrude by rotation about the Z axis, but it is not > at all obvious. > Rotation about the Y axis would be more obvious and just as useful. > I think the documentation should explicitly state that the 2D object must > be > in the XY plane > and gets rotated by 90 degrees into the XZ plane before being extruded. > Thus, positive Y values > become positive Z values in the extrusion. > As I have only just started using OpenSCAD, I thought I'd ask in the forum > before offering to edit > the documentation. > > > > -- > View this message in context: > http://forum.openscad.org/rotate-extrude-issue-tp7442p14807.html > Sent from the OpenSCAD mailing list archive at Nabble.com. > > _______________________________________________ > OpenSCAD mailing list > Discuss@lists.openscad.org > http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org >

Yes, all 2D objects are created in the XY plane, but can be rotated into any other plane. If I have an object in the XY plane I would expect to rotate it about the Y axis, which is clearly its axis of symmetry, not for it to be flipped first (or, equivalently, last) about the X axis into the XZ plane. I am merely saying that I found this counter-intuitive, and feel that it should be explicitly documented. The documentation on rotating a polygon does go some way towards saying this, if you read it carefully, but it should, in my view, be more clearly stated earlier. If you are very familiar with OpenSCAD you may not realise that this looks strange to a newcomer, but I assure you it does. -- View this message in context: http://forum.openscad.org/rotate-extrude-issue-tp7442p14812.html Sent from the OpenSCAD mailing list archive at Nabble.com.

I am not arguing against clearer documentation but trying to explain why it is as it is. Once you rotate extrude an object it has one special axis (the axis of rotation). In 3D Z is the special axis so it makes sense to me to orientate it upwards, making it the same as a cylinder. Although it may seem you can rotate 2D objects around any axis it doesn't make any sense because they remain 2D objects, with no Z dimension. The only way to make them 3D objects is to extrude them. However I am surprised by this: - rotate([45, 0, 0]) circle(100); Shows a tilted circle with F5 and a flat ellipse with F6. Both seem to be bugs to me. X, Y rotations should be ignored when applied to a 2D object. It seems that OpenCSG shows the rotated 2D object in 3D space but CGAL does an implicit projection to 2D. linear_extrude(height = 100) rotate([45, 0, 0]) circle(100); Makes a cylinder with an elliptical cross section in both F5 and F6, so again an implicit projection before the extrude. On 29 November 2015 at 11:28, ChrisLT <clusbytaylor@gmail.com> wrote: > Yes, all 2D objects are created in the XY plane, but can be rotated into > any > other plane. If I have an object in the XY plane I would expect to rotate > it > about the Y axis, which is clearly its axis of symmetry, not for it to be > flipped first (or, equivalently, last) about the X axis into the XZ plane. > I am merely saying that I found this counter-intuitive, and feel that it > should be explicitly documented. The documentation on rotating a polygon > does go some way towards saying this, if you read it carefully, but it > should, in my view, be more clearly stated earlier. > If you are very familiar with OpenSCAD you may not realise that this looks > strange to a newcomer, but I assure you it does. > > > > -- > View this message in context: > http://forum.openscad.org/rotate-extrude-issue-tp7442p14812.html > Sent from the OpenSCAD mailing list archive at Nabble.com. > > _______________________________________________ > OpenSCAD mailing list > Discuss@lists.openscad.org > http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org >

Noted! Anyone is welcome to send complaints re: docs to the topic: "Improving the doc" it will get priority treatment (by me, probably). ChrisLT wrote > I am merely saying that I found this counter-intuitive, and feel that it > should be explicitly documented. -- View this message in context: http://forum.openscad.org/rotate-extrude-issue-tp7442p14830.html Sent from the OpenSCAD mailing list archive at Nabble.com.