I'd have that high up on the project risk register for many obvious reasons, not least the time it might take but most critical the scope for error.
It does sound unbelievable for this day and age, and I'd be questioning it. I wrote software to transfer data between formats on many occasions when I worked for a manufacturer. It wasn't really very clever software as it was usually something we were doing without a budget and proper time allowance. But it was always do-able.
I can remember in some migration projects there would be a bulk load of historic data up to a particular (recent) cut-off point, and anything after that had to be treated differently, but I can't ever remember that that involving re-keying it.
|