[mkgmap-dev] changes in dem-tdb branch
From Gerd Petermann GPetermann_muenchen at hotmail.com on Mon Feb 12 08:40:16 GMT 2018
Hi all, slowly I seem to get closer to a solution for the crash problem. I've re-activated the dem-tdb branch to test some changes. With r4111 I've implemented a special option to force the problem case, so that the crash doesn't depend on the content of the hgt file or the dem-poly, see http://www.mkgmap.org.uk/websvn/revision.php?repname=mkgmap&rev=4111 When the extra flag is set, the crash is related to the position of the tile boundaries (TRE), the height and width - and therefore the overlaps - of the DEM area(s), the dem-dists values, the length of the route, the place where the route starts and ends, maybe more. I assume that MapSource makes certain assumptions about these values when the extra flag is set AND "something else" is true or false. This something might be a value in TRE or tdb, for example I've calculated a map for some tiles with the same (?) boundaries as in the Garmin map TopoGuide Hungary 2.12c. I've calculated some routes in the Garmin map and stored them in a *.gdb file. (no crashes with "Show profile..." button, data looks reasonable) I've used the same routes for the map calculated by mkgmap, I see crashes for some of them. I replaced the *.DEM data in the Garmin map by that from mkgmap and I see no crashes(!) (I made sure that I see the data calculated by mkgmap) It is difficult to create tiles with exactly the same boundary values in TRE and DEM and *.tdb, so maybe I still have to improve the code regarding rounding errors or calculation of the position. That's what I've done so far and I am still making progress... Another explanation is my conspiracy theory: It might be that MapSource is much more restrictive when the tdb file says that the map is not from Garmin, e.g. it is not lockable or whatever MapSource might use to detect that. Let's hope that this is not the case. BTW: Reg. "Show Profile.." /Graph I've learned that both MapSource and Basecamp use different DEM levels depending on the overall length of the route. With dem-dist=5520,16560,y the DEM level 0 (5520) is used for routes < ~ 29 km, 16560 is used for 29..86 km, y >= 87 With dem-dist=3312,x the DEM level 0 (3312) is used for routes < ~ 17,4 km The ratios show a constant value : 17,4 / 3312 ~ 29/5520 ~ 87 / 16560 ~ 190 The odd numbers are probably caused by different units (feet / metres). Gerd
- Previous message: [mkgmap-dev] order of subfiles in tdb and hasDem flag
- Next message: [mkgmap-dev] changes in dem-tdb branch
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the mkgmap-dev mailing list