[mkgmap-dev] addr:place and housenumbers

Sat Dec 13 08:12:16 GMT 2014

Hi all,

seems nobody else is interested ?
If I got it right, the combination
addr:place=* & addr:housenumber=* & addr:street!=* 
is not often used, mappers seem to prefer to 
give all streets in a hamlet the name of the hamlet,
so that they can use the normal scheme with
addr:housenumber=* & addr:street=* 

I used osmfilter on europe
osmfilter.exe europe.o5m --keep="addr:place= and addr:housenumber=" --drop="addr:street=" -o=addr_place.o5m
and found that most elements are in and around 
the Czech Republic (I see more housenumbers with addr:place than with addr:street)

So, I wonder if those users who create maps for this area found
a trick to make address search (with housenumbers) work?

@WanMil:
I must confess that I did not fully understand your initial post in 2013.

Gerd

> Date: Thu, 11 Dec 2014 21:24:36 +0100
> From: wmgcnfg at web.de
> To: mkgmap-dev at lists.mkgmap.org.uk
> Subject: Re: [mkgmap-dev] addr:place and housenumbers
> 
> Hi Gerd,
> 
> some time ago I asked if anybody is interested in adding place 
> boundaries to the precompiled boundaries 
> (http://www.mkgmap.org.uk/pipermail/mkgmap-dev/2013q2/018003.html).
> 
> I never got a feedback (apart from you) about that so I didn't realize it.
> 
> Your idea expands my idea in creating virtual bounds on the fly based on 
> node data. My personal opinion is that this supports insufficient 
> tagging - polygons for hamlets should be created in the OSM DB if are 
> required - but there seem to be regions where evaluating this tagging 
> might improve address data a lot.
> 
> My ideas for your thoughts:
> For step 2 you can use an algorithm that creates the smallest polygon 
> around the node set. I don't remember a name of such an algorithm but 
> you should find one on wikipedia. After that you can increase the 
> polygons to march upon the neighbour hamlet polygon with an upper 
> expansion limit (500m?).
> 
> This step might also be perfomed while preprocessing the bounds so it 
> will not decrease the map creation time very much.
> 
> WanMil
> 
> > Hi all,
> >
> > I have an idea how to support the evaluation of the addr:place tag.
> > I first thought this can all be done with a few rules in the style, but
> > it seems that
> >   the boundaries don't contain the hamlet boundaries, so it is difficult
> > to find the
> > matching names for streets and nodes.
> >
> > Idea:
> > Step 1:
> > - collect all nodes with addr:housenumber=* and addr:place=* and
> > addr:street!=*
> > - collect all nodes with place=hamlet and name=*
> > - collect all unnamed roads
> > Step 2:
> > For each hamlet, calculate the area (bbox or maybe a circle) containing
> > the corresponding
> > addr:place nodes (the place=hamlet nodes should help in case that there
> > are two or more
> > hamlets with the same name in one tile)
> > Step 3: Find out the unnamed streets in these areas. Give them the name
> > of the hamlet
> > and interpret the addr:place tag in the nodes like mkgmap:street (using
> > the normal housenumber
> > routines)
> >
> > Is anybody interested in that?
> >
> > Gerd
> >
> >
> >
> >
> > _______________________________________________
> > mkgmap-dev mailing list
> > mkgmap-dev at lists.mkgmap.org.uk
> > http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
> >
> 
> _______________________________________________
> mkgmap-dev mailing list
> mkgmap-dev at lists.mkgmap.org.uk
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.mkgmap.org.uk/pipermail/mkgmap-dev/attachments/20141213/d62d12f6/attachment.html>