16 year erratum

 There is an erratum published in Physical Review C this week which corrects paper from 16 years ago.  In their opening sentence they own up to adjusting the data in a way they shouldn't have:

In our original publication of level scheme of $104\mathrm{Nb}$, we determined level energies based on certain transitions and subsequently adjusted the raw data for other transitions to fit these energies. This is not correct scientific procedure as it alters original data, and it risks introducing incorrect transition and level energies into the literature. The main purpose of this erratum is to provide the original data.

and go on at the end to 

thank the Physical Review C editors and the data scientists at the National Nuclear Data Center at BNL for calling our attention to these corrections

To me, this shows that processes are working and that the data from their paper, which made its way into databases, was found to be erroneous, the error investigated, and corrected.  Too bad it happened in the first place, but the data remained available for scrutiny and the scrutiny worked.