Comment by jodrellblank

Why shouldn't I complain? Transactions and snapshots and checkdisk were invented lifetimes ago, and building resilient systems out of fragile parts is so ancient that Joe Armstrong is dead and buried and "cattle not pets" is the mantra of the era for cloud scale computing.

Holier-than-thou victim blaming may be a good excuse in FOSS land but we pay Microsoft money for Windows and Apple money for macOS X, and both companies are capable of building stuff which can recover from interruption. Even JETDB is surprisingly resilient when you force-reboot an Exchange server these days, so is SQL Server, so is Active Directory, so is NTFS.