Re: gEDA-user: centroid data generated by pcb-20050315

[DJ Delorie <dj@xxxxxxxxxxx>]

I asked a friend in the business about this when we were first talking
about it, and there really are no standards for pick-n-place as all
the machines are different.  So, no matter what algorithm we choose,
it's going to be wrong for someone.  We could probably provide a few
"obvious" algorithms and let the user choose them, but we can't
possibly support everyone.

Having said that, some thoughts:

1. Yes, documenting what we *actually* do would be a good idea.  Note
   the recent threads about documentation ;-)

2. Median, average, and bounds algorithms all seem useful.

3. We also have the element's mark, which we could define as
   "pick-n-place location", but that would mean that users would have
   to be able to move those on placed elements.

4. Rotation is another can of worms too.