One thing experienced engineers know is that all things are transient, everything fails eventually, and if you're good at what you do you'll make sure that when it fails it will fail gracefully. S.M.A.R.T. for hard drives is a good example. It dutifully reminded me that my primary drive was going to fail for several days before it did, yesterday night, at about 8pm.
Thus, I was able to backup everything I could recall as being particularly important. This included my BYOND work, which I manually backed up anyway every time I coded in a major feature: that's just veteran computer user practice. Also veteran computer user practice is coping with data loss: probably the worst thing I lost was some save game files - I'll be starting over from scratch if I ever decide to play Fallout 3 New Vegas again.
Seagate, while generally a reliable brand, sort of dropped the ball on their Barracudas, especially the 7200.11s, this is the second time I had one fail on me. You might recall that I did a software RAID 0 awhile back. Clearly, I was mad with power, because I was forewarned, I knew firsthand that these drives were unreliable, and yet I chose to do a RAID 0 instead of a RAID 1. Thus, aside from what I backed up, I lost everything. Next time, I will be doing that RAID 1 and, with any luck, the software RAID implementation will do accelerated reading.