[Imports] NYC building + address import - to merge or not to merge?

Danijel Schorlemmer osm at schorlemmer.net
Wed Oct 23 09:04:44 UTC 2013

On Wednesday, October 23, 2013 10:44:04 AM Pieren wrote:
> On Wed, Oct 23, 2013 at 12:39 AM, Alex Barth <alex at mapbox.com> wrote:
> > case for sticking to that convention.
> Taginfo tells us that we have 15.3 millions addr:housenumber's on
> nodes vs 9.7 millions on ways. 

Have you actually checked how many house-number nodes are part of a building 
polygon/way and how many house-number nodes are lacking any building 
footprint? The latter ones can hardly be counted against the notion of having 
house numbers attached to the building footprint as long as the building has 
only one house number. Having the entrance as part of the building polygon 
being tagged with the house number and tagging the building footprint with the 
house number if the entrance is unknown seems the best solution to me.



> Of course, it's hard to say which part
> is coming from individual decisions or mass imports. But the fact is
> that, at the end, the "convention" is addr:housenumber's on nodes. I
> cannot influence your decision but just tell you that you will regret
> this one in the future. And the "convention" argument is simply wrong.
> Pieren
> _______________________________________________
> Imports mailing list
> Imports at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/imports

More information about the Imports mailing list