OpenStreetMap
I'm in the US now visiting Mom. I'd downloaded the Garmin routable maps from Lambertus for California and uploaded to my Garmin eTrek Vista HCx. Before I left Thailand I played with these maps using Garmin Road Trip (Mac). The routing seemed to work just fine.
However, when trying to drive from the San Fernando Valley (LA, California) to Vasquez Rocks, northeast of LA I ran into massive trouble. The routing started out fine, heading us up Interstate 5 and over to California route 14; the same route I'd pick looking at a paper map.
But then, about half way there, the GPSr said to get off the highway. At first I figured this was the "shortest route" that I'd specified, but I soon realized that things were seriously amiss. The GPSr was advising us start heading back south to Interstate 5, the exact opposite direction from our goal. And, when I looked at the GPSr map, I realized that areas south of us were being displayed as being north of us. In short, it seemed that map tiles were either swapped (north for south) or overlapped. I guess this caused the routing peculiarities.
I turned navigation "off" and continued on with the paper map.
Once we got out of the overlap area things seemed back to normal.
I'm away from home and don't have any Garmin software available, so I can't explore this more fully. Will look again when I'm back in Thailand.
Discussion
Comment from Zartbitter on 12 November 2009 at 21:06
Lambertus wrote that routing can has problems in overlapping areas. He's waiting on a bugfix release of the splitter program which should fix this issue. That's what I've read.
Comment from wer-ist-roger on 12 November 2009 at 23:15
Considering your title it seamed that there is something wrong with the actual OSM-data. But to me it looks more like there was either a problem with the garmin map or with your device.
Comment from Buadhai on 14 November 2009 at 06:42
If this project was really all about data and not about maps then it should have been titled "Open Geodata". But, it's called Open Street Map and, aside from the nerds, most people think of geodata as, quite simply, maps. If the maps created from the data are unusable, then we have a problem.
This project will never appeal to the masses unless we manage to present the data in a way that ordinary people can use. The original diary entry, above, simply points out that the data is not yet in a form usable by jerks like me.