Originally Posted by
#19
I love this. In programming I am always looking for alternate representations of the same data. In graphing terms, we label the edges of a graph while they label the nodes.
Or to visualize this, we would describe a triangle as "this side, that side, and that other side, connected to form a triangle" where they would say "this point, that point, and that other point, connected to form a triangle".
What is not immediately apparent, however, is that while the systems can be considered equivalent, they are not equal. They can be leveraged differently. Our system makes it immediately apparent that 221 Baker Street and 1221 Baker Street are connected, albeit at quite some distance, but hides the fact that 1805 Chapel Street is twenty feet away from 221 Baker. The result? In a lot of US cities we see big streets hatched with tiny "feeder" streets.
The block-naming strategy emphasises distance over direction. You immediately see that two addresses share a common block, or ward, or city--and this also has an effect on the street layout.
When you get your organization scheme right, it really works. Salt Lake City, Utah, is laid out on a grid, so if you are at "1300 West 10600 South" and you want to get to "700 East 3300 South", you know exactly what direction to go, what streets to take, and about how far you have to go to get there....
But when you get your organization scheme wrong, it really stinks. Take a city that was built on the locality premise, like Boston, and then superimpose the street-names scheme on top of it, you end up with an unnavigable mess in which people still use landmarks to get around.
Sorry for such a long comment, but the takeaway for me here is not only to see if your business models are reversible, but to ask in what ways would reversing the model make things easier--and in what ways would it make things harder?