need project review

[Sam Hartman]

>>>>> "Henry" == Henry B Hotz <hotz at jpl.nasa.gov> writes:

    Henry> On Apr 5, 2008, at 9:15 AM, krbdev-request at mit.edu wrote:
    Henry> That's OK, but as was said before it's hard to get approval
    Henry> for a change if there isn't a tested rollback procedure.  A
    Henry> bit if documented thought about how to do that would be a
    Henry> good thing, even if it's "restore from a backup".  This
    Henry> change seems low risk and likely to fail quickly if it does
    Henry> fail.

I agree that there should be a rollback procedure.  I think restore from copy or backup seems like a fine answer though.

--Sam