I don't want to make this into an LDM-level test specification. Let's reserve those specifications for tests which are comparing delivered functionality against documented requirements. This, on the other hand, is a fairly ad-hoc specification of a test relative to meas_mosaic which is not tied back to fundamental requirements documentation and which we won't use to fill in a verification matrix: let's not overcomplicate the procedure.

Obviously, there will need to be (a) test specification(s) to verify the ultimate compliance of Jointcal with the requirements described in in LDM-562 (DM-10732). Naïvely, I would suggest that those tests are described as part of LDM-534 (the L2 test plan) rather than in a Jointcal-specific testing document.

Eric Bellm, John Parejko: I've made some small updates (highlighted in red) to the confluence page (https://confluence.lsstcorp.org/display/DM/JointCal+Acceptance+Tests) in response to Robert Lupton's comments. Please take a look at shout if anything there bothers you; I plan to call this issue complete at the end of this week at the end of this week unless there is an active discussion going.

The only change I see is the "mappings should be manually inspected" sentence at the bottom. Is that it? I'm not particularly sure how you plan to do that, but it can't hurt, I guess.

Yup, that's it. For photometry, you could visualize them as images of the spatially-varying PhotoCalib. Astrometry is trickier, but I imagine some kind of whisker plot showing distortions could work?

No objections to closing from me.