We are screwing ourselves. This really isn't anything new at all. But what's the other option, clay tablets? Paper doesn't last forever, technology is always progressing. There's no reason why we can't think of fading technologies as fading languages; we've been decoding forgotten languages for years. Imagine it's 2045, you've just found a CD left by your great grandfather. Pick up a phone book, skip over the historians, skip over the linguists, and call up a computer scientist.
Of course, the biggest problem with this is not on our end, but on the business and government end. I have two external hard drives, 200GB in total, which are bound to be outdated soon enough, but I'm sure I'll buy a terabyte hard drive in a few years and transfer everything onto that one drive. For me, that's easy and not too expensive. I'm in a position to do this, since it's an easy process and I'll (assumedly) always want access to all of my data. Companies, on the other hand, archive stuff away and never touch it again. So, in 20 years, when they (for whatever reason) need to access it, the technology's been replaced by five better products and no one has any idea how to get to that old information.
So, sucks for them, I guess? The first question to ask is how important is all that old data? Who cares that some company can't bring up its employee records from 20 years ago? Similarly, is the most important data the public data or the private data? If it's the former, then it'll be on the Internet (ideally) or at least housed with some sort of redundancy. If it's the latter, then we're fucked.
I can see technology going in an entirely different direction than people might expect though, archiving wise at least. Laserdiscs, tapes, Zip drives, floppy disks, CDs, DVDs, etc. etc. etc.... new portable technologies come out all the time, but hard drives seem pretty constant. IDE hard drives have been around forever, and SCSI is a pretty time-tested technology as well (if not too standardized). Hitachi just came out with a new parallel bit architecture for their hard drives, increasing hard drive space and lowering cost and size. Right now you can get a 60GB hard drive the size of a deck of cards (iPod anyone?); how quickly until 300GB hard drives that size come out and a 60GB drive costs as much as a DVD-RW?
What I'm saying is, if people are willing to spend the money, this isn't a big problem. Back everything up to a few terabyte RAID servers and you'll be set for awhile. Back everything up to a case-load of tapes and you'll be fucked.
Anyway, the all-consuming question is always, "Are we progressing or just moving forward?"
We have already screwed ourselves. This is what happens when you outlaw pirates.
This really isn't anything new at all. But what's the other option, clay tablets? Paper doesn't last forever, technology is always progressing. There's no reason why we can't think of fading technologies as fading languages; we've been decoding forgotten languages for years. Imagine it's 2045, you've just found a CD left by your great grandfather. Pick up a phone book, skip over the historians, skip over the linguists, and call up a computer scientist.
Of course, the biggest problem with this is not on our end, but on the business and government end. I have two external hard drives, 200GB in total, which are bound to be outdated soon enough, but I'm sure I'll buy a terabyte hard drive in a few years and transfer everything onto that one drive. For me, that's easy and not too expensive. I'm in a position to do this, since it's an easy process and I'll (assumedly) always want access to all of my data. Companies, on the other hand, archive stuff away and never touch it again. So, in 20 years, when they (for whatever reason) need to access it, the technology's been replaced by five better products and no one has any idea how to get to that old information.
So, sucks for them, I guess? The first question to ask is how important is all that old data? Who cares that some company can't bring up its employee records from 20 years ago? Similarly, is the most important data the public data or the private data? If it's the former, then it'll be on the Internet (ideally) or at least housed with some sort of redundancy. If it's the latter, then we're fucked.
I can see technology going in an entirely different direction than people might expect though, archiving wise at least. Laserdiscs, tapes, Zip drives, floppy disks, CDs, DVDs, etc. etc. etc.... new portable technologies come out all the time, but hard drives seem pretty constant. IDE hard drives have been around forever, and SCSI is a pretty time-tested technology as well (if not too standardized). Hitachi just came out with a new parallel bit architecture for their hard drives, increasing hard drive space and lowering cost and size. Right now you can get a 60GB hard drive the size of a deck of cards (iPod anyone?); how quickly until 300GB hard drives that size come out and a 60GB drive costs as much as a DVD-RW?
What I'm saying is, if people are willing to spend the money, this isn't a big problem. Back everything up to a few terabyte RAID servers and you'll be set for awhile. Back everything up to a case-load of tapes and you'll be fucked.
Anyway, the all-consuming question is always, "Are we progressing or just moving forward?"
We have already screwed ourselves.
This is what happens when you outlaw pirates.
Reply
Leave a comment