Support feature versioning

See https://stackoverflow.com/questions/75074182/feature-and-featureview-versioning

This can make it easy for data scientists to do experiments and compare which version of feature can produce better results.

Start with the linked Stack Overflow discussion, then inspect FeatHub's feature and feature view APIs to understand their current lifecycle. Define what versioning means for each and how experiments should select and compare versions; the issue is done only when those semantics and acceptance checks are documented and implemented.

Written by the indexing model from the issue text.