[Talk-us] exit_to vs destination

[Paul Norman]

On 7/15/2014 3:00 PM, Martijn van Exel wrote:
> I will follow up with some numbers, unless Paul changes his mind about 
> this not having much interest and crunches the numbers before me. 

Of the approximately 28k exit_to nodes, 24k are children of two or three 
ways with exactly one of those ways being a highway=motorway_link and 
the exit_to node being the first node of that way, which is a reasonable 
first-pass approximation for what can be mechanically edited.

If US usage follows the overall pattern, 14k of 17k exit_to nodes will 
meet the same criteria.