[Talk-us] United States Roadway Classification Guidelines

[Nathan Edgars II]

On Wed, Jul 28, 2010 at 9:11 AM, McGuire, Matthew
<Matt.McGuire at metc.state.mn.us> wrote:
> Tag for truth!
>
> Not to speak for Ian, but what I took that to mean was that tags should be based on observation, not based on how a feature should be represented on a map.

There's no observation that will tell you whether a road is primary or
secondary.
>
> Tagging incorrectly is not good. And tagging correctly is good. Tagging for the renderer is not correct.

Tagging railway=rail for a pair of rails rather than mapping each rail
separately and using rail=steel is tagging for the renderer. It's also
tagging correctly, as would be the rail=steel method.

> All other maps that I have ever seen - including all I have made - have this crude quality. Sensor derived geographic material such as satellite and aerial imagery does not have this quality when viewed at appropriate scales. OSM can transcend the crudeness of vector maps and the scale limitations of sensor derived data if mappers act as agents on one feature at a time.

What you mean is that it can transcend usefulness and become a sea of
unclassified roads. Gotcha.