Eli Rykoff What have you ended up doing for the em_min / em_max attributes? Something like the ad-hoc table that Andy Salnikov uses for the ObsCore extraction in dax_obscore/config/dp02.yaml?

Eli Rykoff Would you code be able to provide multiple instances of prov_progenitor in the file (that's standards-consistent - see the "*" on page 18 of the standard) or is the output based on a dict and so only allowing one instance of each keyword?

For em_min/em_max I just put in dummy variables so far. If this is vital I can add it, with a hacky ad-hoc config dict.

For prov_progenitor I just have a freeform text field that you can put in the config. If it needs to be per-band I can change that configuration variable. But if it should be some cookie-cutter template that the band is substituted in (this is my plan for the creator_did for example), then that's even better.

At last - properties recommendations attached: hips-properties-recommendation.txt

Thanks! To be clear there should be multiple entries for prov_progenitor, and not one long entry?

1. em_min / em_max: it's not "vital", to be sure, but it seems reasonable to be consistent with the em_min / em_max we're reporting in the ObsCore table for the input images.
2. prov_progenitor: yes, multiple instances is clearly what the standard envisions:

Some keywords may be repeated to specify multiple values like provenance (suggested by (*) in the following table).

However, if this is architecturally difficult, because your whole code is based on a dict, say, I'm OK if you concatenate them all, perhaps with semicolons, for now.

A template with {band} substitution should do the trick.

Not hard at all, there's no dict. I'll just change the config to a ListField. Just wanted to make sure I understood completely.

Okay, here are my proposed property files that go with your suggestions. I'm attaching one for i-band, one for the gri color. The other bands look a lot like i-band with suitable substitutions/wavelengths.

Note that the hipsgen linter complains that temp:// is not a valid identifier, but Aladin.jar doesn't complain.

dp02_properties_i.txt dp02_properties_gri.txt

After feedback, here are updates to the properties files. MOC fraction is just for ci_imsim, that will be filled in properly on the real run. dp02_properties_i_v2.txt dp02_properties_gri_v2.txt

These look good to me; I think we can close this ticket now?

We have converged!