On Feb 13, 2009, at 13:51, David wrote:
As I am new here, I introduce myself.
Welcome!
I am the author of a Garmin map made with the well known cgpsmapper. I discovered OpenStreetMap few months earlier and I started to work with mkgmap. My goal is to produce my map with mkgmap. I am not yet ready because of these "problems" :
- mkgmap cannot compile one of my big polygons even if I slice it with mapedit.
Does mkgmap crash? Could you send the crash output and/or a small sample .mp-file?
- mkgmap produces bigger files than cgpsmapper.
I think mkgmap is not very much optimised for small file size, but how large is the difference? Are you comparing maps generated from .mp for both tools?
- derspite data come from OSM, routing does not work with my maps in Mapsource (it is not a real problem, I cannot also do that with free cgpsmapper).
Routing should now work. You need to use the "--route" option (but you appear to know that), and if you start with a .osm-file, you need to check out the nod branch from svn for the moment. What's going wrong for you?
- mkgmap produces a different draw of the map when compiled with net and route options : roads name appear and cities name are not displayed in Mapsource.
Not sure what's wrong here. Anyone else see this?
For mkgmap routing revision 856 and later :
I haven't looked at these modifications in detail, since I don't know those parts of the code too well. Regarding this:
In the case I want to contribute to the code (to correct bugs or activate functions for example), how can I do this ?
It would be best if you post your changes in patch form: Edit a local checkout of svn, then do "svn diff".
As I was looking for why my contour lines and summits with altitude in meters in my file (polish format .mp) are displayed in meters divided by 3.2808399 (a feet to meters conversion) - is there a setting to solve this ?
No idea... What does an entry for a contour line in a .mp-file look like? I would imagine mkgmap just writes a label, so it's surprising this should be wrong.
- I found the function distance() where the earth is supposed to be a sphere which is a pretty good approximation.
It should be good enough for our purposes. Cheers Robert