Github user ahgittin commented on the issue:
https://github.com/apache/brooklyn-server/pull/969
I'm uncomfortable with the changes in third commit
(https://github.com/apache/brooklyn-server/pull/969/commits/e6dee4e591cb2d14acb24a9270a01d2c4bbf5e4b).
Legacy plans that have been stored with the text-based comparison might fail.
Probably not, as Brooklyn will have been restarted and these tags regenerated
so it's probably okay, but the minor efficiency gain of only computing one hash
seems not worth it. I think the code only gets invoked when plans have
identical IDs so it will rarely make a difference -- thus err on side of
correctness not speed.
Suggest a comment to the effect of your idea, which I think on balance is
probably right, but just I'm not certain and caution seems the better part of
valor here.
---
