[mkgmap-dev] Commit: r2013: Create and use "lies in" information in preprocessed bounds

Wed Aug 24 19:57:03 BST 2011

On Tue, 23 Aug 2011 22:46:19 +0200, WanMil wrote:
> I have just uploaded the new europe bounds:
> http://www.navmaps.eu/wanmil/europe_bounds_20110822.zip
>
Hi WanMil

just testing the new 2013 version with the new boundary files i'm 
getting the severe error message 'All levels are finished. Remaining 
boundaries 34' (this number is different per tile)
Do I have to worry about this error message?

Cheers, Johan

> The world file will follow but that takes some time...

>
> WanMil
>
>
>> Just for your information:
>>
>> To benefit from the changes you need to use bounds that are 
>> preprocessed
>> by r2013 or later. I will do that within the next days and upload 
>> them
>> to the navmaps page.
>>
>> r2013 still can use the old preprocessed bounds because the format 
>> is
>> the same.
>>
>> WanMil
>>
>>
>>>
>>> Version 2013 was commited by wanmil on 2011-08-22 21:33:02 +0100 
>>> (Mon, 22 Aug 2011)
>>>
>>> Create and use "lies in" information in preprocessed bounds
>>>
>>> The "lies in" information of a preprocessed boundary defines in 
>>> which other boundaries a given boundary is completely contained.
>>> Example:
>>> Boundary A10 (admin_level=10) lies in A8 (admin_level=8), A4 and 
>>> A2.
>>> All elements in A10 are automatically in A8, A4 and A2. So the 
>>> admin_levels 8, 4 and 2 are defined just be assigning the 
>>> admin_level=10 boundary.
>>>
>>> This has two advantages:
>>> 1. Perfomance: Time for index creation with bounds option can be 
>>> reduced by 25% max.
>>> 2. Bugfix: Boundary information was inconsistent for elements 
>>> located in two boundaries of the same level, e.g. a street crossing a 
>>> city border and a region border could have created one city assigned 
>>> to both regions. Now a city assigned by the bounds locator should 
>>> always have the same region and country.
>>> _______________________________________________
>>> 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
>
> _______________________________________________
> mkgmap-dev mailing list
> mkgmap-dev at lists.mkgmap.org.uk
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev