ANetAstrometryTask and all other astrometry.net-related code has moved from meas_astrom into a new package meas_extensions_astrometryNet, along with all lower-level DM code that depends on astrometry.net… The APIs have not changed…
For now ANetAstrometryTask is still the default astrometry solver, so pipe_tasks depends on meas_extensions_astrometryNet. Once we change that default (on DM-10242) we can remove that dependency. I don’t know when we will consider ourselves ready to do that.
This split was made on DM-2186 and merged to master on 2014-04-28.
We have run lots of data through the new matcher, and we expect it will be even better once @cmorrison’s improvements have been made. I’m all for switching the default to the matchOptimisticB now, and then to @cmorrison’s version once that’s been validated extensively (perhaps as part of @hsinfang’s processing runs?).
I think @hsinfang’s first run really ought to be with basically the algorithms we’re running on in the current HSC internal release (i.e. w_2017_17), but I’d like to commission the new matcher (and the new WCS fitter, which has also not yet been enabled as the default) either in parallel or shortly after it’s done.