2022031 Release, Linux AppImage; Ubuntu 20.04.3 LTS/18.04.6 LTS
The idea is that the ellipse should end at the point named A16 originating from A15. In the 20220110 release it does so.
Here are my files: cloak.1.2.val (15.1 KB) Cloak.vit (1.1 KB)
Hmmm… seems like I might’ve fixed one issue (ellipical arcs not mirroring) and regressed to a new issue? I’ll take a look as soon as I can.
2 Likes
Verified… sort of. The arc is still ending at A16… sort of, but it’s looping at the end for some reason. if you zoom in on the piece you can see the loop:
1 Like
Yes, it’s very strange. The ellipse should end at A16.
And anyway, on the pattern piece, it shouldn’t take the portion above A16. The curve/arc should end at the node selected unless it is selected before and after the node:
I’ll figure it out… the error looks to be in the geometry routine to find the 2nd point. I fixed the (2nd point) geometry to account for mirroring an ellipse… broke something else. 
2 Likes
Ok… here’s what I got on this one…
First an issue seems to exist in piece mode way back:
Produced this:
I was able to fix the endpoint angle:
But… what I can’t fix is the seam allowance angle. Here’s why. The ellipse uses 2 radii, if no rotation is added the 1st radius is on the Y axis (270 degs) and the 2nd on the X axis (0 deg). Using A15_A16 as a radius does not place point A16 on the ellipse… it’s close, but not on…
therefore the seam allowance angle type at A16 will not work.
Which brings me to the question… why use an ellipse for the waist length when the other 3 lengths are a curve? That being said, using the ellipse, the better option is to not bother with A16… create the ellipse using 270 and 0. Then using the intersect arc and axis create a point on the arc at the axis angle of A15_A3… then the point in on the arc, and the seam allowance angle type works correctly.
2 Likes
I’m beginning to wonder that myself. I think it’s because being such a big percentage of such a small arc I am not confident of being able to get it to drape evenly. (Yes, I know that I ought to just get it pretty close, & trim it to dimension after it’s hung for a few days, but I’m in denial about it. The main reasons I didn’t use the ellipse for the others is because of all the trouble I had getting that ellipse [not quite close enough to] right, & being smaller sections of larger arcs they fudge nicer. I think for the sake of pattern stability I’ll just trace the ellipse with a curve this time.
Thank you very much for the advice, it is much appreciated.
1 Like
Very well spotted! @Douglas
But the ellipse is still going past the angle point.
If I roughly add a node to the end of the ellipse, it’s a few degrees out:
If it was ending at the correct degree, then it wouldn’t have made the difference to the pattern piece. It would’ve just jumped that small gap with very little or no distortion.
Yes… I haven’t uploaded the changes yet. 
1 Like
Something else that dawned on me… since the 2 radii are different ie NOT a circle… if you were to sweep through various angle values going down from the approx 350 deg… eventually A16 will be on the Arc… then it will go outside the arc as you head towards waist. So yes, there is some angle at which A16 is on the arc, just not at the angle of A15_A16.
Basically the radii are NOT necessarily the start and end point angles… only if you select 0-90, 90-180, 180-270, or 270-360… OR those angles ± the rotation.
Also in regards to the seam allowance and the angle type… within the seam allowance routines there is some fuzzy math going on… if A16 were closer to the arc, the angle type may then work even if it not on the arc. I don’t think it’s worth fudging the accuracy factor to make this example work.
1 Like