Dave.

There’s a lot to be said for “http://yourISP.com/~username” being available 24/7 at no particular effort to you.

90% of users when they are presented with the UAC popup when they do something:

“Yes yes whateverrr” <click>

I’ve got photos in Flickr dating from 1999 onwards. Ten thousand or so of them, and a couple of the early ones are now corrupted.

But they are my “other backup” for Google photos so I don’t mind too much. I also have a USB Blu-ray drive at home that I use to periodically burn M-Discs that I hand out to a few relatives.

That’s about as good as I can conveniently do for backup, and it’s probably better than the single-point-of-failure box of negatives that my parents have in their cupboard.

when they’re powered down.

There’s no periodic cell refresh in flash memory like there is in DRAM. When USB sticks are plugged in, all you are doing is powering up the flash chip and interface ICs.

You’d have to read a block then write it back to actually refresh the stored charges in the cells.

I work in OT. The number of “best practice” IT mantras that companies mindlessly pick up and then slavishly follow to the detriment of their mainly-OT business is alarming.

Make your own damn best practice that suits your business best, don’t copy and paste something from a megacorp. Sure, include elements from megacorp’s best practice if they are applicable, but don’t be a slave to the entirety of it.

Ok I’ll just inject a little bit of context into those numbers because it smells faintly of a hit piece by Reuters, simply because of the timeframe used and the number of employees spaceX has.

My experience is 30 years in the mining industry, which in that time has become pretty good at managing safety, and reporting on it.

So I’ll dig in a little.

Since 2014, so nine years.

SpaceX employee count : 13000 approximately.

Take about a quarter of that to weed out the paper pushers and company growth since 2014, gives us 3500 or so employees in the line of fire (that is, manufacturing and such).

600 reportable injuries, so about 66 injuries a year. About 5.5 a month on average, over 9 years.

Now those 3500 employees work 60 hour weeks (because: spaceX). So 5.5 injuries and 840,000 man-hours a month. I’m going to round those hours up to 1 million for convenience and to counter the fact that I ditched quite a few people in my initial assessment of SpaceX employees in the line of fire before.

And with a bit of half-assery , I say, “ta-da!” and get 5.5 reportable injuries per million man-hours at SpaceX over the last 9 years.

So, what kind of number is that? Well for tracking this kind of thing normally you would work on a value called that “lost time injury frequency rate” - LTIFR - which is the number of injuries per million hours worked. Oh look, my previous rounding to a million has become very convenient.

Looking at the data that Reuters has given, and my half-assed guesses about employees, spaceX has a long term LTIFR of 5.5. Note that number drops significantly if you use SpaceX’s entire employee base, which as a single entity, they would be quite entitled to use and report.

How does that number stand up against industry norms? 5.5 is middle of the road for manufacturing and construction, generally, but that includes all sorts of manufacturing, from building houses, to steel foundries , to making cars.

The fact that Reuters had to take 9 years of data to make the raw numbers sound alarming enough is a bad smell. They could have calculated LTIFR numbers for each year and figured out a trend and if that was alarming enough, they could have reported on it, like “SpaceX increasingly dangerous to work at!”. The fact that they didn’t makes me suspect it’s a hit piece, although I am willing to accept they didn’t want to get into LTIFR numbers and are dumbing it down for the general public.

Absolutely the number of serious injuries is a concern. Serious injuries are also at the top of a “injury pyramid”, with every layer underneath broader, all the way down to “Ow, I stubbed my toe”. If you have real figures for one layer (like a layer where an employee can’t hide an injury), you can get a good idea of what the other layers should look like.

Judging from Reuters’ numbers, the bottom “minor” layers aren’t getting reported enough, which suggests a lack of safety culture at SpaceX. Although that could simply be from Reuters’ using only public records, which, you know, only keep track of injuries worth keeping track of, so the bottom of that pyramid might only be seen by SpaceX internally.

In conclusion, the reporting by Reuters of raw numbers over long timeframes is suspect. That’s not how things are done in the safety industry, which works with weighted metrics to get results they can compare between companies. Dig in a bit further yourself.