|
Hi Using http://www.openstreetmap.org/?lat=-33.76155&lon=151.13008&zoom=16&layers=M as an example, there are many streets without names on them. If you zoom out a little and scan around, you will see that all around that area there are usually more blank street names than streets with names. The strange thing is that I had used OSM before and all those street names were there! As best I can tell, I have an extract from 26/5/2012 that includes all street names. I have another extract from 26/9/2012 that has many street names missing. I thought I must have picked a bad day in September, but the street names are still mostly missing now at the end of November. Can anyone help fix this? |
|
OpenStreetMap has changed its license to ODbL (Open Database License) roughly two months ago, a few days before you created your more recent extract. This change required mappers' permission to continue the use of their contributions under the new license. The vast majority of mappers have done so, but a few did not want to or were unable to agree to the new license, so their contributions were removed. Even though this has affected less than 1% of our data globally, some areas have been disproportionately damaged by the data loss. Sydney is among those areas. To fix the gaps, the street names unfortunately need to be added again. (We would be happy about your help with that!) Simply using the old data as a source is not permissible, however, because that would go against the will - and maybe copyright claims - of those mappers who chose not to accept the license. So these names need to be derived from local knowledge, survey, or sources explicitly permitted for use in OSM. 2
Thanks for your helpful explanation. I'm familiar with various licenses and associated issues, so I can understand why things may have regressed with a change of licenses. I like the look of the new license, and will be happy to contribute. Time to go searching for best way to do this! Cheers
(01 Dec '12, 06:50)
dbIsMe
2
http://qa.poole.ch/?zoom=12&lat=-33.88925&lon=151.02801&layers=TB0 can help with locating the most affected areas. I suppose the good news is that because of the high density of missing names in some areas re-surveying them should be very efficient.
(01 Dec '12, 08:07)
SimonPoole ♦
|
