Virtual1 (en-GB)

You lost me!

You shouldn’t trust everything the Internet tells you.

The Internet is in essence a huge international network and requires quite a lot of admin to keep it going and there are still a lot of bits that you might think are tightly defined but the truth is generally the opposite. As an example, let’s take a look at geo-location data.

When you ask your car satnav to plot a route to get somewhere, it can use a post code, one of a set of rough location points around the country run by a trusted database owner we call the Royal Mail. They have matched each post code to a real world set of co-ordinates that the satnav system can use to determine the destination on a map. The accuracy only needs to be within a few metres and generally it works pretty well.

Now consider the Internet. You type something in your web browser and the server you need to reach is looked-up on what for simplicity we will call “the Internet routing system”. The key technical element to identify the server you need is its IP address. So you could think of the IP address as a sort of post code for the purposes of our example.

But what if you want to merge these two things and know physically where in the world something on the Internet physically exists? Say for example, you are an international airline and you want to present your home page in the language that best matches where the customer is connecting from. The real problem is that there isn’t a universally adopted system that maps IP addresses to physical locations.

Being resourceful, various parties simply built their own databases and borrowed address information from related objects in public databases, taking the head office of the service provider or the address of the administrator. This allows a reasonable guess. But it isn’t “official”. It is however, the best we have at the moment. Geo-location data fields have been added to the relevant official databases in recent years but adoption is low and so we are swimming in a sea of unofficial databases, very far removed from the situation with the Royal Mail’s post codes.

In the meantime, we get some interesting situations. I know of cases where customers report their address is listed as being in Russia by some sites, in Germany in the next. In the official databases, they are clearly listed as being in the UK, but somebody running their own database in the USA didn’t think that was right. As these databases are privately owned, the owners are not obliged to listen and errors can persist. All you can do is let them know and hope they care about data integrity or that fear of losing their paying customers through inaccurate results forces their hand.

Recently, the BBC reported on a couple from Kansas who are suing such a company for $75,000 after five years of problems. As they seem to geographically be the location for a load of dodgy websites, they have been visited by police for everything from missing children to stolen trucks about which they know nothing. Of course the servers for those sites are nowhere near their little rural farm. Taking legal action against the database owner was the only course of action they had left.

So bear all this in mind next time you visit an online shop and they think you are based in Kazakhstan. Their Internet satnav just took them the long way round.