R: [mkgmap-dev] [PATCH v4] - merge nodes to remove evil short arcs
From Mark Burton markb at ordern.com on Thu May 28 11:12:03 BST 2009
Hi Marco, > I tested the patch (V2) compiling Italy and I found that some problems are solved but others remain (maybe remain the ones coming form the exactly duplicated nodes, or some other "evil" configuration). As long as we end up with a net reduction in issues, we're winning! > But I hope remaining problems will disappear with the pach correctring the utilization of right HASH. Unfortunately I cannot test both patches contemporary since I'd need the .jar of 1044 + evil arc patch...). I will send one to you. > Anyway just a proposal for a V5: if it is feasible, instead of deleting arcs if lenght<3.5m, it whould be better to delete arcs > > if ((Utils.toMapUnits(latitude1) == Utils.toMapUnits(latitude2)) && > (Utils.toMapUnits(longitude1) == Utils.toMapUnits(longitude2))) > > You would avoid to delete the arcs which ends will not really collapse in a single Garmin IMG point (you avoid the un-needed deletion of about 75% of arcs with respect to the simpler arc lenght test). Yes, that makes sense. I will look at doing that. Thanks again for your help. Cheers, Mark
- Previous message: R: [mkgmap-dev] [PATCH v4] - merge nodes to remove evil short arcs
- Next message: [mkgmap-dev] [PATCH v5] - merge nodes to remove evil short arcs
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the mkgmap-dev mailing list