On 3/3/2011 11:16 PM, Alan Mintz wrote:
Is this data really worth importing?
With the required amount of review, I wouldn't classify them as a data import so much as a requested work list. These addresses are important to the owners in some way: historically GPS's have gone to the wrong location, the road was misnamed, etc. And many people set these locations as navigation / map search targets. From the point of view that it is good to cater to those who consume OSM data, it is good to have these address points on the map.
Why should this data be any different? Why is dragging a point to the nearest building and more correct that dragging it to the next block, or the next town, if you don't know the feature personally?
In at least one case I found, it would be wrong to drag it to the nearest building - and since it began in the middle of the road, you can't even tell which side it belongs. This was probably a case of "GPS's always send my customers to the wrong place". Most of these should be surveyed before import to meet the expected OSM quality level. Only a few are clear from aerials alone - such as a church or school in a residential / rural area where it is the address in question.
That brings up another point: workflow. When / if someone decides to add the few addresses that they can survey, it would be good to mark them as having been "imported" and remove them from the master list. Otherwise it will be hard to find new reports when MapQuest publishes an update. Throwing them up on MapDust / OSM Bugs would almost be appropriate, but I'd say they are a bit more informed than the average OSM Bugs / Mapdust report.
_______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
