[mkgmap-dev] splitter r325: improved split algo and new option

Wed May 7 11:50:05 BST 2014

Am Dienstag, 6. Mai 2014, 13:56:42 schrieb Gerd Petermann:
> It is now less likely that splitter creates tiles with a low number of 
> nodes, it is more likely that all tiles have nearly  the same number of
> nodes, and typically you will see fewer tiles.
> Maybe this also means that you can increase the max-nodes value.

some infos about splitter 321 vs. 325

Intel i5 QUAD @ 2.7 Ghz, 16 GB RAM
ramsize = -Xmx6G
maxnodes = 1600000

o5m as input and output

rel		tiles	time		use_areas

sri lanka
o5m size 17.1 M
321		1		2s		yes
325		1		3s		no	
325		1		3s		yes

oal (Ost-Allgäu)
o5m size 78.7 M
321		4		7s		yes
325		3		8s		no
325		3		7s		yes

china
o5m size 261.1 M
321		16		18s		yes
325		15		25s		no
325		15		18s		yes

bonn +100km
o5m size 739.9 M
321		35		53s		yes
325		28		62s		no
325		28		54s		yes

germany
o5m size 3.2 G
321		152		250s		yes
325		124		256s		yes

DACH+
o5m size 8.3 G
321		357		703s		yes
325		392		866s		no
325		392		733s		yes

europe (just for fun)
o5m size 22.7 G
321		1038		4645s	no
325		1046		5267s	no
325		1046		4681s	yes

the whole planet
o5m size 45.1 G
not tested yet, maybe tonight

-- 
amarok2 now playing: