This is an almost-entirely-mechanical change that probably isn't worth reviewing in detail across the board. To be specific, most packages have one commit with the message "Switch to YAML for PropertySet/PropertyList storage" - that was entirely a mechanical find-and-replace for:
- BoostStorage -> YamlStorage
- .boost -> .yaml
...and since I already looked at what that did, I don't think you should feel like you need to as well. That is the only commit in most obs packages.
These packages have more than that:
- obs_base: updates to test code (in addition to a big mechanical commit)
- ip_diffim, meas_astrom, meas_deblender: old, bitrotted, example code that I've either removed or minimally changed to avoid reference to BoostStorage.
Looking at those would make this a pretty tiny review.
Meredith Rawls, I'm sending this to you in part because I don't think I've sent a review to you before, and I'm trying to branch out a bit, and partly because I'm not sure if you're using the Task metadata in ap_pipe in a way that might exercise this change in ways that are different from what I've tested (lsst_distrib and ci_hsc). If that's the case I was hoping to learn what else I should do to test that this does no harm.
This is an almost-entirely-mechanical change that probably isn't worth reviewing in detail across the board. To be specific, most packages have one commit with the message "Switch to YAML for PropertySet/PropertyList storage" - that was entirely a mechanical find-and-replace for:
...and since I already looked at what that did, I don't think you should feel like you need to as well. That is the only commit in most obs packages.
These packages have more than that:
Looking at those would make this a pretty tiny review.
Meredith Rawls, I'm sending this to you in part because I don't think I've sent a review to you before, and I'm trying to branch out a bit, and partly because I'm not sure if you're using the Task metadata in ap_pipe in a way that might exercise this change in ways that are different from what I've tested (lsst_distrib and ci_hsc). If that's the case I was hoping to learn what else I should do to test that this does no harm.