Write SOP for updating a dataset to a new data version

[Clare72]

Some datasets are not 'final', but 'snapshots' that need to be updated every so often. We should have an SOP for this (to add to the curation manual wiki?) to make sure all steps get carried out each time. This might be different depending on whether VFB IDs are generated by us (e.g. hemibrain) or supplied by the source (e.g. FlyWire).

Things to consider:
updating node annotations - label, comment etc. (anat record)
adding new (and removing old?) site/api xrefs
updating typing info (newmeta record? - but may want to remove old info)
deprecation of any nodes not in the new version