[Talk-us] Karlruhe Scheme addressing ways from 2009 TIGER data - Discussion

[Dave Hansen]

On Wed, 2009-11-18 at 08:57 -0600, Travis Rayhons wrote:
> 1) How do you see this information being imported into OSM where
> existing TIGER data already exists?  For instance, I know that I've
> added roads to OSM that didn't exist in the first TIGER import.  I
> noticed that they exist in the OSM2 file that you provided me.  How
> would those roads be matched up to assign the new address information
> to them?

The plan is to make those files available without any of the existing
roads being in them.  Then, if someone wants addressing, they take the
addressing ways and manually lay them over the existing data, making
sure to line them up if they don't line up.

That's what I think SteveC and I agreed to.  Although, it was his idea,
so I could be remembering wrong. :)

> A) With either version is the intention to eventually remove the
> generic address information for specific addresses of locations?  I
> can see as the mapping increases that more buildings and locations
> will be mapped with specific addresses.  What is the intention here
> where two addressing schemes (a general and specific addressing) will
> be used?  This should be spelled out in the wiki for all mappers to
> refer to.

Yes, I do believe that is the intention.  But, practically, we'll always
have both schemes around.  I think this is covered on the Karlruhe
Scheme wiki page, though.

> B) With either version, does this allow me to search for addresses in
> my GPS?  If so, could someone provide some assistance to me in how to
> do that?  I've manually added a way of each version to OSM to test
> this out on my GPS.  However, I was unsuccessful with using mkgmap to
> allow my GPS to search for any addresses within the ranges I manually
> added.  It is possible that this isn't the intended result.  If it
> isn't the intended result, what is the intended result other than
> having this data in OSM?

I don't know if mkgmap has support for addressing off the top of my
head.  I'd have to go look at the code.

The intent is to get the data into the map.  Having all of the other
devices and geocoders start to use it will happen later.

> If it were up to me, my preference would be to use OSM2.  However, I
> see that this could be a big challenge with an automated import.  To
> further suggest a solution, which could be used for either version,
> provide the files to local mappers who can either confirm the auto
> import or choose to manually add the address data for their region.  I
> would stick with one address scheme so that the whole country uses one
> scheme throughout.

There's not really a debate over the format, at least in my mind.  I
don't *like* Karlruhe, but I don't have a better suggestion.  It's kinda
like perl.  I don't like perl, but I seem to use it for everything. :)

There are a number of those OSM files (1...6 in some cases) and I
honestly haven't tried to figure out what's in all of them.

The plan now is to use the Karlruhe Scheme and not tag the actual street
ways.  The on-way tagging in those files is simply a convenience while
we're looking at them to make sure that the Karlruhe ways got created
and we can check them quickly.

-- Dave