There seems to have been a change in the Positions file in 5.1.6 with respect to 5.1.5 (or whatever the current version was back in February).
I had this board that I ordered and was manufactured … (happy ending … I got the boards).
I need to re-order, and I made one very minor change: I rotated one component on the top side.
I generate the Positions file now, keeping the previous positions file.
When I run:
[···]$ diff positions-file-original-board.csv positions-file-rev-2.csv
I see the difference in the one component I changed, but I also see all of the components on the bottom layer with the X coordinated negated (i.e., negative values — all of them have the same magnitude, just a minus sign in front). Example (excerpt):
< "C7","0.1uF","C_0603_1608Metric",72.850000,55.310000,0.000000,bottom < "C8","0.1uF","C_0603_1608Metric",65.220000,54.040000,180.000000,bottom --- > "C7","0.1uF","C_0603_1608Metric",-72.850000,55.310000,0.000000,bottom > "C8","0.1uF","C_0603_1608Metric",-65.220000,54.040000,180.000000,bottom 84c84 < "D1","TVS","D_SOD-323",66.810000,46.650000,270.000000,top --- > "D1","TVS","D_SOD-323",67.200000,47.450000,0.000000,top
(D1 is the one that I changed)
I omitted the complete output, but it would seem like all of the components on the bottom layer got the x-coordinate multipled times ‒1
Is this a bug introduced in 5.1.6?
Or maybe I inadvertently changed some settings on the board or when generating the output? (since I did change some traces, I did go to the Board Settings page and made changes; However, I’m essentially 100%… say, 99.99% certain that I didn’t change any settings or any other changes to the board; just added a trace width and changed the differential pairs values)