round-trip error with ###NaN

[plexus]

https://circleci.com/gh/lambdaisland/deep-diff/65?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link

Something the generative tests caught, not sure this is a big issue but something to look into.

[alysbrooks]

I think there are two issues, and I think they're really two manifestations with the same issue. (not= ##NaN ##NaN) which breaks diff, but it also breaks our tests.

matchers-combinators and clojure.data/diff fail (if that's the right word) in the same way. Maybe this just needs to be documented?

[alysbrooks]

Glad I put my research into writing, but I don't think this should be a priority for the moment.

[plexus]

Seems that CircleCI link no longer works for me... should've taken a screenshot.

Adding a section to the README with "known limitations" would perhaps be the best short-term action here.

[alysbrooks]

I think this makes sense to leave something open as a documentation fix. Fixing the underlying problem probably won't be a priority, so this can be closed as "PR welcome" once the documentation is done.