Folks, you don't need a local/wide area disaster to cause these problems. How many houses burn down every day? Keeping your 'stuff' off the floor helps, but the only thing that REALLY works is some form of off site backup. Be that "swap the USB drive with the one kept in the bank once a week" (then you only lose a weeks worth of stuff) to automated backups. The DISADVANTAGE of 'swap to the bank' is that MOST folks have their bank close to home, and whatever takes out "home" could prevent you from getting to your backup

That said, backup plans can get a tad 'extreme'. Back when I was a kid, and just getting interested in computers (aka I think this was circa 1972) I got a tour of a data center where the folks were truly paranoid (but they were also using it for load sharing)

The had 2 redundant mainframes at each data location. The location I toured was in Downtown Manhattan. There was another center in Midtown, then there was one outside Boston, One in Chicago, One in California, London, Munich, Paris, and Tokyo. The LAST data center was in Alice Springs. Yes - their disaster plan DID figure in Nuclear War

Well, we go one step further - to WHOM is the cleanup valuable? Why shouldn't the people who value the cleanup PAY for the cleanup. The owner can say "I'm happy the way it is, YOU are the one who wants it cleaned, YOU pay to clean it"

I guess being way older than most /. readers, I still remember when it was fairly common

True, but we looked at it as "This is what we do" vs "I'm a police officer stuck doing calibrations"

It's a wierd thing, we did our jobs because well, it was our job, so we did them as best as we could. I know when I was tuning a box to go out, my goal was the box went out with every number nominal, all it took was a little pride and skill

Wow! Back in the 80s, when I worked for a contractor, the rule (which was a RELAXED version) was
You calibrate at the Mfg recomendation. If found Out of Cal (OOC), you HALVED the calibration period for that unit. If found IN cal for two or more consecutive cal cycles with NO adjustment, you got to increase the calibration period 50% (making a unit that had been OOC 75% of the Mfgs recomedation). Eventually, the units tended to settle into being just in cal, but way off nominal and getting calibrated at the end of cycle. Units that were lightly used tended to move to 200% Mfg recomended (which if I remember right was the max you were allowed to go), and heavily used units were somewhat lower, depending on how conservative the Mfg was in setting their unit
There was also a max cal time - if I remember right, it was 5 years, and that's where standards were (you know, the stuff the cal lab used for reference)
A real fun story. Rule of thumb was you had to have a calibrator that was one significant figure BETTER than the unit under calibration to calibrate it. Easy enough, until you start calibrating what is the most accurate unit of it's kind ever made (which we did - even NIST used one of our units for calibration). We had to go back to first principals to calibrate that unit. It was fun, and took circa 2-3 weeks to do that calibration (I know, I helped on that one). Joke? We NEVER found one out of calibration. The unit was all ratio transformers, and the ratios just don't change on potted transformers. You can make them FAIL (too much power), but you CAN'T make them drift. That said, we never gundecked (read pencil whipped/lick and sticked) those units. We were getting paid to do it, so we did it

OK, yes, I know I'm a LOT older than the average /. reader (Hint, I'm starting to get mailings from AARP), but when I was a kid, when you needed to replace a set of pistons on an engine, your FIRST job was to measure the cylinders, and see if you needed 0 +5, +10 +20 etc pistons. Yep, the Mfg tolerances on the engines was such that car A's parts were NOT necessarly interchangable with car B's.

Modern machining, and in particular modern carbide tooling (one of the first real nano technologies - the particles use to make them are nano sized), and probably more interesting to the /. crowd cutter ware compensation build into CNC machine tools have allowed Mfgs to hold tolerances that are WAY WAY tighter than they were

BTW by tighter tolerances, I don't mean FITS, I mean tolerances - if I put a 0.50000 pin into a 1.00000 hole, I've spec'd a TIGHT tollerance, but a VERT loose fit

I know a guy who was a Crew Chief of the 135Qs (mostly out of Beale), and he said even when he was in (about a decade before they took them OOS) they were beat. Has some interesting stories, including what happens when you have you boom box (dating the era there) sitting up on the top of the Vertical stab, and you knock it off, and how much trouble you end up in

One of my regrets is never seeing an SR-71 in the air

THAT explains the Citytime fiasco, eh - maybe he's looking to get in on the Nth version
http://www.nytimes.com/2011/11/01/nyregion/bloomberg-administration-admits-mishandling-citytime-and-nycaps-programs.html

I used to encrypt my emails to the few people I knew who supported PGP, but guess what? Most folks don't have PGP and frankly, they don't care, and most emails I write you can read - if you really want to know where my daughter's cell phone was, or what I was planning on making for dinner....

Yeah - I had one (no details) where I was calling a 8, when they realized what I HAD, they were saying most people call it a 9-10, and I said I could imagine worse. 12 days of high doses of morphine. What fun. I'd rather NOT go through that again. Something about being in too much pain to even scream, just moan, BP and heart rate spiked, curled up, semi concious. They told me that if I had come to the hospital 12-24 hours later, I would not have come home