EDIT: I read too quickly and didn’t notice it was a full component, not just two pin headers. Flipping necessarily causes a mismatch of pins/texts and re-routing. I earned 1½, did I?
Yes, it seems to be, as can be expected. T/P = around x is the default, which makes me wonder if @straubm has found the setting after all, changed it and then forgot.
I didn’t change the setting and less so did I forget. I just mixed up flipping and mirroring and rotating and everything. Never post before you had your first coffee.
To whom should I bow down and cover my head in shame? @eelik ? @jmk ? The world at large?
Well, it was flipped L/R to solve Fearless’ problem, and left to right is the Y axis, but if an 8 pin socket is placed on the now, underside, it will need re-routing, so both still right, so still ½ each.
No shame, no grovelling, both half right, or is that left… I suppose it depends on the flipping.
Reading the original post the third time, I noticed I still didn’t understand the purpose of the board. Now I see the component itself is added to the board and its pads are routed to the pin headers. The component is not a “module” with pin headers, this board will be the module. The two pin headers are separate, and it’s possible to mount them in whatever direction because they don’t affect routing. It would have helped to see the component in the 3D image.
I have had pin headers in the “wrong side”, so it’s possible if you don’t care about the silkscreen lines which show the borders of the plastic parts. What is important is to make sure the silkscreen text and the component pins routed to the header pins match. If you flip the pin header footprints you have to make sure it didn’t mirror the pin order. You can flip the physical pin headers safely, but flipping the footprint may reverse the order of the pins and therefore the connections/routes.