A few years ago, while cycling, I had the idea of creating a wikipedia like cyclemap, where everyone who likes, can help. For the time being I discarded this idea, because I could not find enough time to program all the needed software and stuff. Now I found out about this project and almost everything is laid down in front of my feets. Hurray :-)
On one of the Freiburger Geocaching Events I heard about OSM. I was eager to do mapping myself, but unfortunately I had no cable to connect my GPS-device (a Geko 201) to my computer so I used waypoints for the first view streets I mapped. After about two weeks I finally got the cable and started mapping Freiburg. Mostly I concentrated on roads and their names omiting almost everything else.
A Freiburg Mapping Party was held. During that party my feeling, that the used datastructure is not the best, increased and I finally decided to leave the project.
Autumn 2007 and Winter 2008
In September 2007 I bought my CSx 60 and suddenly needed maps. Well I perfectly well knew where to get some and soon I was back in the project. Still I did not like the datastructure. Soon there was a change with the segments being removed, from where on I liked the datastructure a little bit more (this was probably also due to the invention of relations). I decided to map footways in a nearby forest, leaving a lot of the problems of the datastructure aside. This continued for a long time and only once in a while I also added POIs or other stuff, mainly in Freiburg.
Having a new girl friend in Haltern am See I suddenly came to an area, that almost has not been mapped at all. Together with the book of Jochen Topf and Frederik Ramm, that gave me some new energy, I started to map Haltern, and finally I finished on the main city somewhen in April. The only thing left there is part of the pedestrien area in the center, where I was swamped by the number of pois and stuff that needed to be considered. During that time I thought a lot about the cycle problem, which - in my opinion - is still not solved.
While being in holydays in Leksand, Sweden, I mapped a larger part of the city. Again I thought about the datamodel. Finally I came to the conclusion, that we need two "layers" of data: The first one should contain only polygons, the stuff that is often refered to as areamapping. The second one should mainly contain ways and pois, the stuff that is needed for routing. This would solve a lot of recent problems: Stacking of ways (should areas be created by reusing streets or not), the cycleway-problem (should cycleways be mapped as separate ways or as cycleway-tags), and probably some more.