Alan Millar wrote:
>> Can/should this be done with relations instead of separate paths?  The
>> idea is to have it stick to the road.  With this it creates a lot more
>> nodes, and there is no easy way to manually get the spacing right
>> between the road and the addresses if something is moved.  Also the
>> spacing needed is render dependent.
> 
> Can it be done with relations instead of [or in addition to] separate
> paths?  Yes.  Should it be done instead?  No.
> 
> Sure, the majority of addresses are for houses that are easily,
> unambiguously identified with the street that their address belongs to. 
> But reality I've seen is that there are great numbers of exceptions, and
> so the Karlsruhe schema is necessary to indicate where an addressed
> building actually is.
> 
> Take a look at the Clackamas Corner library:
> 
> http://www.openstreetmap.org/browse/node/365787985
> 
> Now find its street, SE 82nd Ave.  It is more than half a mile away.  And
> the library is in a strip of stores, so an addressing way would/could
> apply here.

Good catch on the Clackamas Town Center situation; I'm surprised I
didn't come up with it first, as I used to drive deliveries around that
area as a private courier and would often lose valuable time and fuel
trying to find addresses along the Cascades Highway (of which 82nd
Avenue is a member) only to find my destination is five or six blocks
off the highway in a maze of unnamed ways that don't appear on TeleAtlas
or Navteq.

This situation is by no means unique, I can think of many locations
throughout eastern Multnomah County where this would apply, such as all
addresses on NE Frontage Road in Troutdale, which is split over either
side of a freeway, with a few addresses behind other addresses for good
measure:  I imagine you'd need to either terrace it out (easy in that
area because of the large, easily identified lots; give each lot an
outline, and tag the outline with Karlsruhe and landuse= tags) or have
two interpolation ways (one each tier of lots).

Out of the two, I think terracing works out better and wouldn't be a bad
idea in the long term, since this would give a general idea where plot
outlines are.  I see interpolation as more useful as an interim measure
prior to terracing, and handy for locations where terracing just won't
work (such as suite numbers, "subhouse" numbers, etc) in trailer parks,
and sprawling apartment blocks and business parks.




Attachment: signature.asc
Description: OpenPGP digital signature

_______________________________________________
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us

Reply via email to