I'm tempted to say the same thing myself. I don't care what license my data is under and I have accepted the new license but I do not want to see data deleted because I think that this will kill the project. OSM data is useful because it is relatively accurate and complete in many parts of the world and deleting data will create holes in the map which will make the data useless (e.g. suppose 1/4 of road segments were created by users who didn't accept the new license, but they are scattered semi-randomly - deleting this data makes the map data totally useless for end users.) Then users will have to waste a bunch of time repairing this mess but the incompleteness of the data will scare off users. As long as there are a significant number of users who have not accepted the license terms (and there may be a significant number of contributions by users who contributed for a while then left the project and forgot about OSM and didn't bother to respond to emails), this will be a problem. I want to spend my time mapping not wasting it cleaning up a mess made by legalese that 99.9% of non-lawyer OSM contributors couldn't care less about. If this crazy idea of deleting data goes ahead I guarantee that everyone will migrate to a fork with the old license and intact data versus the ODBL version with a quarter of the data missing.

There's lots of inactive users in my area (Toronto) as well. Maybe we should send them emails to encourage them to contribute?

You need substantially more than 1 tile to make a map of decent size. I think that this statistic is meaningless, sorry.

If the imagery is properly licensed, by all means trace from it. Of course, surveying is very much encouraged since the imagery may be out of date/unclear and some things cannot be added from imagery (road names, small businesses, house numbers...)

This seems to be a perennial problem with Google Maps and other proprietary data. For instance, Google recently (a few days ago) rolled out new data for Canada, which is mostly derived from GeoBase government data (which OpenStreetMap is also using). It is full of errors and omissions, for instance the lack of the recently opened northern extension of Highway 410 in Brampton which is correct on OSM. My only advice is to keep mapping and NEVER use Google as a data source.

spam

spam

Just realign the roads to match the satellite imagery. There are a couple of GPS traces in the area you are looking at, it is obvious that the satellite imagery is aligned properly in the Memphis area. (Don't forget that GPS can also be inaccurate as well.)

Get up early on a Sunday morning and do it - no traffic and hardly anyone outside, but it's still light out (in the summer).

TIGER 2009 data is better than the old TIGER data which is currently in OSM. See: .

Spam

It might be useful to copy from Tiger 2009: http://www.census.gov/geo/www/tiger/tgrshp2009/tgrshp2009.html. This is public domain government data and significantly better quality than what is currently imported into OSM.

Spam

If you click on a section of the generated route a dialog box will appear allowing you to rate that section's perceived safety on a scale of 1-7, this will affect future generated routes. Also the system tends to prefer residential streets and cycleways over major roads, and prefers roads with bike lanes or signed bicycle routes.

There's a minor technicality that means that the Toronto data is not compatible with OSM (for now):

http://lists.openstreetmap.org/pipermail/legal-talk/2009-November/003017.html

I hope that this gets fixed sooner or later.

I will normally unglue parks and landuse (because the edge of the park is NOT the centerline of the road, because the width of the road is positive), but I usually glue boundaries to the middle of roads/rivers if they follow them precisely.

Generally this is a result of misaligned Yahoo imagery. To align it to GPS press space+arrow keys (in Potlatch) or use the little move WMS button (in JOSM). Usually once the Yahoo imagery is aligned properly, you can rely on it for tracing. In some cases Yahoo is spot on, in other cases it needs a bit of manual adjusting.

I've been mapping sidewalks separately as ways in Toronto too (still many, many more sidewalks to be done though). Haven't got around to mapping them as areas yet though (would need more accurate aerial imagery to do this, such as the Nearmap aerial imagery in Australia). I'd be careful using the crosswalk tag for pedestrian crossings as I normally reserve this for pedestrian crossings with flashing lights activated by the pedestrian (which are very common here).

spam

In many areas Yahoo imagery is more accurate than GPS traces, if the area hasn't changed since the photo was taken and the photo is aligned properly (or is aligned manually to the GPS trace in the editor). Remember that consumer GPS isn't all that accurate. In general, I recommend comparing both sources with what is actually on the ground and using whichever is more accurate.