Workflow question: Reconciling database local <-> remote

[eluser]Crimp[/eluser]
Two instances of the same data is not always a bad idea ;-)

It's the latter part that poses the challenge. There are situations where updates to schema/content needs to be pushed from test to live, for example. Like you propose, I currently do this with some variation of "manual" labor, outlined in the first post.

So the answer to my question about how others best handle this may be: they prefer not to.