I was running the other day when my brain stumbled in an interesting direction. I was thinking about how the government houses vast amounts of information about us. Each person has to have several thousands of lines of information in databases somewhere. I was thinking how one could efficiently query this information in a manageable amount of time. You could have a huge set of storage with a monster front end system, or perhaps there would be an alternative. I then thought of how DNS works.
If you are looking for a .com, you hit the root servers, and work your way down to the relevant server that holds the information. If you are looking for a .org, you head in a different direction. Why not use this same system for massive databases?
You could store the person’s name in the “root” database server. If you need tax information, you are directed to the “.tax” server. If you need their arrest record, you head to the “.cops” server. I know you could just write nested queries or the like, but what if the back-end DB system was designed and built to do all of this for you? You wouldn’t have thousands of programmers cranking away for hours every time they wanted to develop a system like this, it would already work like this, and you could just query the system. I suppose you could just create views that did all of the joins and such for you…eh.
I figure there is something already like this, or perhaps my logic is flawed…I didn’t put too much thought into it, it just seemed like an interesting concept.
What does the peanut gallery have to say about it?
Leave a Reply