Code review comment for lp:~thisfred/u1db/sync_integration_tests

Ah, no: The code is triggered (just put a print in there) but all it does is shortcut the put_doc_if_newer, because what the target has is already strictly newer (this is the case where _validate_source returns 'superseded' rather than ok.) In the case where something is really wrong we raise an InvalidTransactionId, and that code is also exercised by the tests.

I'll try and change the relevant tests so that they detect that the put_doc_if_newer is broken off early, although I'm not sure that it's important, since it should not have any effect when it isn't (and indeed doesn't seem to, as your commenting out the check shows.)