[mkgmap-dev] --add-pois-to-areas behaviour

Thu Jan 28 14:03:23 GMT 2010

Felix Hartmann schrieb:

> That's because it's only a dirty hack alltogether. It would be bit
> cleaner if there was a seperate file inside the style-file where one can
> specify the poi to areas keys. Currently mkgmap can only know what kind
> of POI to create, if it also has a matching Polygon. That's as far as I
> understood the principle.

I think I don't understand the point. Why do I need a separate file?
Why this is not just the following algorithm?

if --add-pois-to-areas is set
1. look if there is a matching polygon rule, make a polygon of specified type
2. independent if there was a polygon rule, look if there is a point rule for this tag and create a point of specified type if so

> Even with a seperate file it would still be a
> rather dirty hack, as it would be much better to have the POI inside
> openstretmap at first hand. I don't really think it is good style to map
> POIs for areas.

No, I disagree. If you map a POI as polygon in Openstreetmap there is enough information for a later renderer to get a point out of this if needed.
Why should I map an additional point in OSM? That would only increase redundancy and brings not more information.

> Maybe a better solution would be to not create a poi,
> but just match in the search indexes (but I don't know if that is
> possible, I deem this to be impossible).

I haven't seen yet, that garmin supports searching of areas, but I don't know enough about that.

Christoph

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: OpenPGP digital signature
Url : http://lists.mkgmap.org.uk/pipermail/mkgmap-dev/attachments/20100128/995cdb33/attachment.bin 