Please forgive me for a long and image laden post.
I don't know that this is a rotation/scale issue. I am thinking it is something to do with how the fabric deals with breaking up curves when they are part of a remainder. Here is another case: Part of lot 1 of the North Silver Lake Subdivision is being amended by Bellevue Phase One. Both subdivisions use the same tie point, the lines shared by both subdivisions are identically described. When I join Bellevue to the fabric, the residuals are pretty tight - the highest was around 0.005. The rotation on Bellevue before the remainder is 359.99954°. The rotation for North Silver Lake Sub is 359.99981°. Scales are 1.000006 and 1.000005 respectively. Here is a screen shot of North Silver Lake Sub before the remainder, note the radial bearing for curve 4-6:
Here is what it looks like after the remainder, notice the radial bearing for curve 29-30 is not the same as the parent curve:
My theory is that the chord bearing from the parent curve is somehow corrupting the new curve. I am willing to bet that point 30 lies along the chord bearing of the parent curve and throwing all of the other lines off.
I just did a simple test that proves my theory. I first created this subdivision -
Then this one inside of the parent and joined it.
After performing the remainder, I get the following:
And -
Everything looks fine after the remainder, but no acreage is calculated. You only see the problem after you open the two resulting parcels from the remainder task. Both of those curves are retaining the original chord bearing of the parent parcel instead of being recalculated. Because of the mountainous terrain of our county, splits happen on curves quite frequently (at least 3 times in the past month).
I am thinking this is a bug.
Jeff Ward
Summit County, Utah