<div dir="ltr"><div><div><div><div><div><div><div><div>Hi,<br><br></div><div>I don't think that Pieren was complaining about someone or because the first example were in France, it's more about the way to do it : on the french list we get complains about polygon deletion, people already call it vandalism as there was no related message on the french list ...<br>
</div><div><br></div>among the "few" change that where already made (<a href="http://zverik.osm.rambler.ru/whodidit/index.html?lat=45.66096&lon=4.57577&zoom=11&layers=BTT&user=pnorman">http://zverik.osm.rambler.ru/whodidit/index.html?lat=45.66096&lon=4.57577&zoom=11&layers=BTT&user=pnorman</a>), some where closed to my city :<br>
- <a href="http://www.openstreetmap.org/changeset/21258184">http://www.openstreetmap.org/changeset/21258184</a> : deleted relation was v11, ways where from v8 to v23<br>- <a href="http://www.openstreetmap.org/changeset/21257976">http://www.openstreetmap.org/changeset/21257976</a> : deleted relations in v2 or v3, ways from v2 to v19<br>
</div>=> doesn't looks like an original not modified import ...<br><br><br></div>I think we all agree that CLC import is not of good quality now, but at that time we didn't had bing, and yahoo imagery was not high resolution in this kind of places ...<br>
</div>Big multipolygon are also a pain to edit, but it's perhaps better to split them in smal parts than just delete them all ?<br><br></div>Have you some more detailled statistics about all this "bad" meadow polygons ?<br>
</div>- for France, how many relation are in V1 AND from user CLCF06 ? after the first import a lot of people already worked on the data, and created new (splitted, modified, ..) V1 with the same tags but not the same geometry<br>
</div>- among the polygons, what was the proportion between meadow and other elements ? if the non meadow is just a few %, then the meadow polygon still give a usefull information (especially for big meadow lands like in Poitou)<br>
</div>- what is the total area of all the polygon you want to delete ? <br><div><div><div><div><div><br><div><div><div>I find it bad to have to mass delete some 4 years data. Is there no other way to improve the data, like to use a quality insurance tool or task manager ?<br>
</div><div>- ex places that need more mapping: <a href="http://tile.openstreetmap.fr/?zoom=11&lat=45.64801&lon=4.54062&layers=B0000000FFFFTF">http://tile.openstreetmap.fr/?zoom=11&lat=45.64801&lon=4.54062&layers=B0000000FFFFTF</a><br>
</div><div>- ex multipolygon with errors: <a href="http://osmose.openstreetmap.fr/fr/map/#zoom=10&lat=45.613&lon=3.926&layer=Mapnik&overlays=FFFFFFFFFFFFFFFFFFT&item=1170&level=1%2C2%2C3&tags=&fixable=">http://osmose.openstreetmap.fr/fr/map/#zoom=10&lat=45.613&lon=3.926&layer=Mapnik&overlays=FFFFFFFFFFFFFFFFFFT&item=1170&level=1%2C2%2C3&tags=&fixable=</a><br>
</div><div><br></div><div>In France we already made some (big) long term work with such tools, like to map all the +36 000 french communes. I'm sure we can deal with some meadow ;)<br><br><br></div><div>cheers,<br>Sylvain<br>
</div><div><br></div></div></div></div></div></div></div></div></div>