Am 29.11.2009 um 18:20 schrieb Johann Gail: > Hi Thilo, > > what is the state of this patch? Is it runable, or have you stopped work > on it because of unsolvable problems? Maybe the garmin strucutres need > to be made more dynamic, getting its final values in a later step? > > I remembered that there was some problems, but have not investigated > further. In general the concept seems ok to me.
Hi Johann, I couldn't get the patch working properly with routing. The main problem is that to merge the ways I need the processed ways, so that I can compare the attributes that are actually put into the IMG files. It is of no use to compare the OSM tags, because then even minor differences would prevent merging of the ways, even if most tags won't influence the resulting Garmin ways at all. The IMG files are build "along the way" when the OSM data is processed, so I never have all processed ways available. This is made worse by the structure of the code, where the classes get nested deeper and deeper during processing... I had to change quite a lot of code to make all *processed* OSM data available at some point. But somehow this broke the routing. Probably this has to do with the way the ways that are split along the way are tracked. But it ended in a massive rewrite anyway, so it would have been a nightmare to maintain as a patch. And the resulting code wasn't nice and clean either, because it went against the work flow of the existing code. Therefore, I stopped working on it. Regards Thilo _______________________________________________ mkgmap-dev mailing list [email protected] http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
