There are many examples of important paper records (property, birth/death, etc) being lost in fires or floods, so it's not the case that "everything worked perfectly fine" in those days either.
There are many examples of important paper records (property, birth/death, etc) being lost in fires or floods, so it's not the case that "everything worked perfectly fine" in those days either.
All those examples are not different from not having backups of digital data too. Making copies when you microfiche, having duplicate stores, it's all exceptionally easy and a solved problem.
And of course everything didn't work perfectly, it did however work "perfectly fine", which means "very well" or "good enough". Meanwhile, adding in network connectivity to anything vital these days is just insanely dumb.
No software is secure, and will never ever be secure. Ever. Anyone who thinks that software can be made secure, is 100% wrong, period. My point is that the advantages aren't worth the disadvantages.
Your overall assessment of advantages and disadvantages seems like you're comparing paper with backups to software without backups, though. No competent system can have all the records destroyed remotely. And we know how to backup digital systems even better than we know how to backup paper ones.
And as a third option we can have efficient digital systems that aren't plugged into the Internet. (Presumably the Internet could have a copy that's regularly updated.)
That's not precisely true, depending upon the definition of 'competent'.
I've seen 'competent' backup solutions which had backup rotations that expired older content. And I've read post-incident responses where hackers slowly corrupted older records. And (via reading corporate wikis and docs), determined how long they had to wait, for backups to become tainted.
But I agree. True competent backup methods, including periodic backup restoration tests and other methods, can ensure a degree of protection. Still, one can end up with months of "bad backups" due to a hacker interceding in the process or corrupting records slowly, thus invalidating more recent backup sets post-hack. While this didn't happen with Equifax, the hackers were in there for 8 months slowly exfiling information.
So you can have a load of backups with questionable integrity.
Really, what I'm comparing here is indelible backups vs not. It could be holographic, write only storage for all I care. It's just that "paper" backups, which move to microfiche, have a century long history of how to deal with it. How to ensure they're good. How to keep duplicates, resolve security, and all that.
Meanwhile, most people dealing with the software side simple think "Oh, we can make this situation secure. We can make software secure."
This thought process is entirely wrong. Software is never secure. If you use software + a database or other store for data, it's not secure. And so, placing it online is just plain stupid.
Of course, you speak to other options. No external network connection, for example. And this is all well and good! And it significantly reduces the attack vector.
But of course, and lots of high security environments do this, you then have to ban staff from bringing in all phones, computers, and other devices. I read a post-action report where people's phones were hacked, and basically acted as a 2G modem (at the time) into wifi on an isolated network of a nuclear power plant. And due to the thought process of "We're 100% air gapped, who cares!", the hack was apparently an easy one.
But really, the difference is... how many people can attack your citadel.
The internet means billions. No network connection means hundreds.
It's just that simple. And I think you agree, mostly.
I was ready to argue until I got to the end "My point is that the advantages aren't worth the disadvantages."
Suddenly your entire argument shifted in my head and I fully agree.
Might be something there I can learn about myself then :D