[mkgmap-dev] Coast branch generates bogus-looking multipolygon errors

Mon Dec 27 14:56:09 GMT 2010

Hi WanMil,

>>The mkgmap.log.0 is 322973 bytes with r1751 and 3855601 bytes with 
>>r1755, which are comparable to the numbers that I posted for r1728, 
>>r1753 and the map data from 3 days ago (December 23). In other words, 
>>the change in the mkgmap.log.0 size seems to be due to the merge of 
>>the coast branch.
>great work! Your detailed analysis helped a lot to find the problem. It 
>is fixed in r1760.

Thank you for fixing the bug. I confirm that r1760 is OK. The 
mkgmap.log.0 for today's dump is 323685 bytes with both r1751 and r1760, 
and after stripping the timestamps and sorting the lines, the output is 
identical, except for some generated polygon IDs. Also the gmapsupp.img 
is the same size with both r1751 and r1760.

	Marko