John Aravosis wants to know how a database of everybody’s telephone records can possibly beat every other database in the world in terms of size. The simple answer is that it can’t. A modern mainframe could handle a database like that with room to spare. [Update – I should point out that this is an off-the-cuff observation. Better-informed individuals could easily prove me wrong.] John thinks that it would get a lot bigger if you started throwing audio files (e.g., the actual phone calls) into the mix, but I think that he’s barking up the wrong tree. The phone companies might hand over names and numbers but there’s no way that they would record and turn over every single conversation. Besides being obviously, screamingly illegal it demands an expensive infrastructure upgrade that somebody would have to pay for.
When you think about what the government might use to flesh out a phone-records database something else comes to mind. Imagine sifting through a massive database looking for questionable activity. What sort of cross-referencing would you find particularly useful? FBI files, for one. Since 9/11 most of us have one and those of us who already had one saw it get a lot fatter. Both databases contain a phone number so matching one to the other would be a cinch. As long as I’m compiling I would probably throw in the TALON database as well. Why not tax records? It’s illegal as hell of course, but on the other hand people are out there plotting to kill us all. It shows how far we have slid as a country that the idea doesn’t even sound that crazy.
Honestly I have no idea what you need to throw in to crack the top ten databases in the world, but John’s right that it has to be more than phone numbers, times and dates. Doubtless we’ll know more soon.