Horses comma dead comma beating, re: drives hard Noisy old profile Reviving

Rules of thumb are just that. They aren't meant to be Science but Craft; analysis beyond some point(*) is generally worthless. Hard drives last longer than food, but not much. Reviving for data recovery is great, but all of the commodity winchester types are of limited lifespan. If you get 25K hours, consider yourself skating on thin ice. PS: If a drive is dead due to stiction, then rotating it sharply parallel to the platter (I can't believe rotational direction matters nor that you can know which way they turn w/o opening), or rapping during power up, or taking the cover off all work "a lot". On the latter, the once or twice I've done it I just wiped everything off with a clean rag first, hit reset, rotate the platter (finger on enter hub, I forget) then xcopy (it was some DOS box) the data off. On those old ST225 type drives, sometimes the little brush (tab) finger thing that grounds the rotating spindle, on the bottom outside of the drive (under? the PCB?) squeals. Bend it. (*) I suppose you could use real methodology to determine exactly what the failure modes really are, and possibly disk manu's do that. But there's not much value in it outside disk-drive {manufacture,repair} worlds. Drive construction techniques come and go, by the time the likes of us (non-disk-drive-manu-engineers) figure one thing out, it's changed again.