I'm curious to know how astrometry.net
handles proper motion, stellar parallax and aberration. Whereas precession, nutation and the like affect all stars uniformly through a rotation, and hence do not change the skymarks, the aforementioned effects affect different stars differently and would hence alter the skymarks. As we move further from the year 2000, I imagine that the indexes will need to be regenerated for the present epoch to account for proper motion. But parallax and aberration having a period of a year, I imagine this is not a useful strategy in this case. Assuming a "perfect" imaging system producing a perfect plate of the sky, I presume that parallax and aberration would be the primary sources of error in the solution?