DM-5249 and (to a lesser extent) DM-4887 have raised some issues with the plugin system for measurements. We’re finding that we need to support more plugin signatures than we originally anticipated, and the current system makes adding a new signature much more heavyweight than we’d like, especially when we want to also run plugins that only require a subset of the arguments in the plugin signature.
A few of the involved parties (@reiss, @yusra, @rowen, @ksk, @natelust, and myself) will gather tomorrow, Thursday, April 21 at 11am Pacific to discuss, location TBD. Others are of course welcome to attend as well.
I’ll try to join. Another use-case that we need to support (without subclassing if at all possible – I assume that that what Jim means by “much more heavyweight than we’d like”) is https://hsc-jira.astro.princeton.edu/jira/browse/HSC-1395 (Add PSF-matched aperture measurements in child and parent images)