logo separator

[mkgmap-dev] [PATCH v1] Multipolygon: do not warn about joining problems for ways outside the bbox

From Marko Mäkelä marko.makela at iki.fi on Mon Nov 8 20:19:45 GMT 2010

On Mon, Nov 08, 2010 at 10:20:55AM -0800, aighes wrote:
>I tried r1728 with the lake Päijänne, as Marko said. No improvments.  
>Would it help, if I upload an tile as osm.gz?

I guess that it would help if you supplied all your scripts and files, 
similar to what I have done at http://www.polkupyoraily.net/osm/.

At the very least, you should get my logging.properties and 
logging.ignore from there, and do

java -Dlog.config=logging.properties ... -jar mkgmap.jar ...

instead of your usual invocation. Does

grep -vf logging.ignore logging.properties

show anything useful?

By the way, someone noted on 
that the Osmarender rendition of the northern Päijänne is broken.  
Without looking at it, I guess that also it is a problem with tile 
boundaries crossing the multipolygon.

I haven't seen any warnings for Päijänne ever since I moved my tile 
borders so that Päijänne fits within a single tile. The bogus warnings 
that WanMil's fix removed were for the lake Saimaa (which has been split 
to several multipolygons).

>If you code coastline-processing with an additional inputfile, please 
>keep the actual way alive. I think not everyone would like this way, 
>because of problems with an older coastline and new objects, which 
>where situated now in water. So every user should deside which errors 
>are "better".

I don't completely understand what you are trying to say, but I 
understood that the plan is as follows. If a separate coastline file is 
given, the natural=coastline tags from the main input tile file are 
dropped, but the ways and their other attributes than natural=coastline 
are preserved.


More information about the mkgmap-dev mailing list