DP0.2 Known Issues and Major Changes vs DP0.1

Current requests for backporting to 23.0.x for DP0.2 

Approved requests for backporting to 23.0.x for DP0.2 

Completed backports to 23.0.x for DP0.2 

Notes from team meeting Sept 29 2021:

Known Issues:  If new ones come up use label:  backport-v23,  DP0-patch

• Forced phot on good seeing coadds
• Tons of deprecation warnings log bombing for computeShape() deprecation. 
• ComputeShape  errors on small fraction of imageDifference quanta
• We write Parquet without compression leaving 10% savings. Each column is compressed individually. 
• Configs:
  • Increase required template fraction. 
• maybe issues?
  • Are the forced photometry on the diaObjects sensitive to crappy single frame WCSs. e.g. DM-30300. Need to trim on sky coords.

Tech Debt: 

• Diffim Pipeline and forced phot (on object and diaobject) pipeline don't use jointcal/WCS or FGCM photocalibs. 
• Post RFC-807 APDB psfFlux, and primary key to index.
• Are the WCSs in forced photometry robust to extrapolation?  Use a distorted camera geom into lsstcam and then using the affine transform.  How to make HSC camera geom of less problematic. DC2 has a model, distortions are not applied. there is no focal plane distortion simulated in DC2 says Dominique. Josh disagrees.
• Double check that proper Null values are used for missing patches in the object table. 

Notes from team meeting Sept 22 2021. 

• Moving CcdVisit and Visit to global step7 (done)
• Jeff's ticket to add the faro steps DM-31848 (tonight)
• Let's not push for the wcs=None ticket 
• Major changes coming? That need to be merged AFTER the release branch is made.
  • RFC-775. 
  • PIFF: 2022

Tickets for adding elements to DRP in the gen3 pipeline: including template generation, difference imaging, and association.  To track tickets we want to get in before w_2021_40 pipeline freeze for DP0.2. 

See Difference Imaging Pipeline for DP0.2 (Planning Page) for Diffim Pipeline planning specifically