Adam has some ruminations on what should happen when a data loss has been reported, and it turns out the data wasn’t really lost (the tape was found, the laptop was in the closet, etc.). While I can understand the temptation to strike out that entry in wherever it was logged, I think it’s important to keep both the original report and a new report of the data being found. Why don’t we see statistics on data that wasn’t really lost, anyway? Is it because lost data is almost never found? Or just nobody thought to compile such statistics?

-jsq