[mkgmap-dev] splitter: No solution found

Sun Jul 27 06:39:37 BST 2014

Hi Andrzej,

I can't reproduce the problem on my machine because I only have 8GB
mem and the grid for south-america with resolution=16 requires more,
so debugging is a pain as the machine is constantly swapping.

I tried todays south-america-latest.osm.pbf with resolution=15 and r412
finds a 
rather good solution with 104 tiles. 

Do you have a good reason to use resolution=16 ?
I think this value is useful for small input files, but I see no 
need to use it with such a large one. It just seems to blow up
the amount of data.

I have an idea how to improve the algo if 
the "No solution found for partition..." message
is written. I'll try to produce my own case with a smaller sample set
to reproduce the problem.

Gerd

popej wrote
> Hi,
> 
> splitter 412 failed, when processing extract for South America. I got 
> following message:
> 
> Warning: No solution found for partition 
> (-60.0018310546875,-137.362060546875) to (29.783935546875,-
> 24.9993896484375) with 56 760 324 nodes
> 
> South America data was downloaded from geofabric.de:
> http://download.geofabrik.de/south-america-latest.osm.pbf
> 
> I have made some tests with following command line:
> splitter.jar --stop-after=split --resolution=16 --max-nodes=600000 
> south-america-latest.osm.pbf
> 
> Splitting was done correctly with splitter r395 but failed with r411 and 
> r412. I'm attaching logs.
> 
> -- 
> Best regards,
> Andrzej
> 
> _______________________________________________
> mkgmap-dev mailing list

> mkgmap-dev at .org

> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
> 
> log.zip (10K)
> &lt;http://gis.19327.n5.nabble.com/attachment/5810384/0/log.zip&gt;

--
View this message in context: http://gis.19327.n5.nabble.com/splitter-No-solution-found-tp5810384p5812803.html
Sent from the Mkgmap Development mailing list archive at Nabble.com.