Re: [OSM-talk] The French Corine Import has started
2009/9/29 Joseph Reeves > Hi Emilie, > > Apologies for not replying in French - although I have an interest in > what you're doing, I don't have the language skills to speak to you > about it natively. > > I'm very pleased to see that the Corine data import is going forward > and am looking forward to seeing the results when it's complete. At my > place of employment we make some use of OSM (see [1] for example), > including in France, and we are considering using it to a much greater > degree in the future. We are having ongoing discussions, however, > about OSM completeness surveys - in short there are people who want to > use OSM but want to be able to be explicit about what areas are > covered and which are not. Which seems fair enough... > > I was wondering if the Corine data import could give us such > statistics for France. Is it possible to say, for examples, how many > polygons exist that are tagged landuse:residential that don't overlap > a way tagged highway:residential ? Is this the sort of thing other > people see value in? > > Regards, > > Joseph > Hello, don't worry about answering in French. I have been living in English speaking countries for many many years. Corine can provide you with some of the statistics that you are asking for. There are some imprecision about some polygons as you cannot see some details sometimes, but generally speaking it would be not that difficult to get some basic statistics. I am doing most of my statistics with postgis since it is giving me lots of power over what I can do. The most difficult part is always to give a clear description of what you want. If you are not careful, you will get meaningless statistics. There are of course caveats to that due to the way roads are being tagged, their length, the precision of the polygons, etc... Corine land cover is only mapping areas more than 25 ha so some towns in my home area are not appearing as landuse=residential, even if they are clearly in a village. In addition, data is at least from 2006, maybe older. I will have to tweak landuse residential polygons around my area to make sure that they match reality. Emilie Laffray ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] The French Corine Import has started
Hi Emilie, Apologies for not replying in French - although I have an interest in what you're doing, I don't have the language skills to speak to you about it natively. I'm very pleased to see that the Corine data import is going forward and am looking forward to seeing the results when it's complete. At my place of employment we make some use of OSM (see [1] for example), including in France, and we are considering using it to a much greater degree in the future. We are having ongoing discussions, however, about OSM completeness surveys - in short there are people who want to use OSM but want to be able to be explicit about what areas are covered and which are not. Which seems fair enough... I was wondering if the Corine data import could give us such statistics for France. Is it possible to say, for examples, how many polygons exist that are tagged landuse:residential that don't overlap a way tagged highway:residential ? Is this the sort of thing other people see value in? Regards, Joseph [1] http://mapdata.thehumanjourney.net/office.html 2009/9/27 Emilie Laffray : > Hello, > > I am sending a quick message to mention that the French import of Corine > has started. We created an user for the occasion: > > http://www.openstreetmap.org/user/CLCF06 > > The polygons are already starting to appear all over France, since we > are uploading polygons by polygons. The import will take about a week to > complete. The final green light was given by the French community after > reviewing some mapnik overlay and review of the logic used for the > calculation of overlaps. > The maximum size for a way is 2000 nodes. > I hope you will enjoy looking at France and see polygons appearing as > much as we do currently. > This is the first phase of the import. We are importing only polygons > with little overlaps with existing OSM geometries. The scripts will be > published this week. The goal of this import was to minimize any impacts > on the user who have been working hard in completing the map of France. > The second phase will be the use of a web interface to allow individual > users to import polygons that were not imported manually. The interface > will be made public once the main import will be finished. > > Emilie Laffray > > > ___ > talk mailing list > talk@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk > > ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] The French Corine Import has started
On 28 Sep 2009, at 10:06 , Ruben Wisniewski wrote: > I think the user speaks for itself, you can look at the history if you > want to know where the data came from. But I think there should be a > tag > like on the tiger data which indicates if the data was reviewed > locally, > to prove it is valid. > such a tag is entirely useless. tiger data is the best example. - everyone can delete, copy, change tags - any tag can be changed - some mappers will follow the advice others will not, found a tiger way yesterday which was transformed into a portion of the coastline and added the coastline tag but kept most tiger tags. at the end this creates just more confusion. no application can rely on a reviewed=no tag If you ask mappers to actively add a source=* or reviewed=* tag then there is at least a higher chance it is there for a reason. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] The French Corine Import has started
I think the user speaks for itself, you can look at the history if you want to know where the data came from. But I think there should be a tag like on the tiger data which indicates if the data was reviewed locally, to prove it is valid. Regards, Ruben Marc Schütz wrote: >> I am sending a quick message to mention that the French import of Corine >> has started. We created an user for the occasion: >> >> http://www.openstreetmap.org/user/CLCF06 >> >> ... >> > > Shouldn't the attribution (source) go into the changeset? I don't know what > the current consensus about this is. AFAIK a bot is removing these things > from the TIGER data in the US right now. > > Regards, Marc > ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] The French Corine Import has started
On Mon, Sep 28, 2009 at 12:39 PM, Frederik Ramm wrote: > It will disappear from there anyway as soon as somebody modifies the way. > > Bye Hi > Frederik Yes, I know. And I would say it is normal if the original landuse is improved by local survey. It is not if the landuse polygon is not modified at all. It is not easy. But I still make a difference between some attributions missing and no attribution at all in exports and dumps for the 170.000 polygons. Pieren ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] The French Corine Import has started
Hi, Pieren wrote: > I would like too. But if we move the attribution (which is a legal > statement we must insert) into the changesets, it disappears in the > exports and planet dumps. It will disappear from there anyway as soon as somebody modifies the way. Bye Frederik ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] The French Corine Import has started
On Mon, Sep 28, 2009 at 11:03 AM, "Marc Schütz" wrote: > Shouldn't the attribution (source) go into the changeset? I don't know what > the current consensus about this is. AFAIK a bot is removing these things > from the TIGER data in the US right now. > > Regards, Marc I would like too. But if we move the attribution (which is a legal statement we must insert) into the changesets, it disappears in the exports and planet dumps. If you have a solution for that, we will take it. Pieren ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] The French Corine Import has started
Hi, Marc Schütz wrote: >> I am sending a quick message to mention that the French import of >> Corine has started. We created an user for the occasion: >> >> http://www.openstreetmap.org/user/CLCF06 > > Shouldn't the attribution (source) go into the changeset? I don't > know what the current consensus about this is. AFAIK a bot is > removing these things from the TIGER data in the US right now. I would suggest to put attribution in the changeset; you can then be very specific without filling up the database with vast amounts of duplicated data. Plus, if a way is later changed manually by someone, they often forget to remove or amend the "source" tag if it is placed on the way. The current TIGER changes are a different matter though; there, import and attribution information is removed from the nodes because it is already on the ways. Bye Frederik ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] The French Corine Import has started
> I am sending a quick message to mention that the French import of Corine > has started. We created an user for the occasion: > > http://www.openstreetmap.org/user/CLCF06 > > ... Shouldn't the attribution (source) go into the changeset? I don't know what the current consensus about this is. AFAIK a bot is removing these things from the TIGER data in the US right now. Regards, Marc -- Jetzt kostenlos herunterladen: Internet Explorer 8 und Mozilla Firefox 3.5 - sicherer, schneller und einfacher! http://portal.gmx.net/de/go/chbrowser ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
[OSM-talk] The French Corine Import has started
Hello, I am sending a quick message to mention that the French import of Corine has started. We created an user for the occasion: http://www.openstreetmap.org/user/CLCF06 The polygons are already starting to appear all over France, since we are uploading polygons by polygons. The import will take about a week to complete. The final green light was given by the French community after reviewing some mapnik overlay and review of the logic used for the calculation of overlaps. The maximum size for a way is 2000 nodes. I hope you will enjoy looking at France and see polygons appearing as much as we do currently. This is the first phase of the import. We are importing only polygons with little overlaps with existing OSM geometries. The scripts will be published this week. The goal of this import was to minimize any impacts on the user who have been working hard in completing the map of France. The second phase will be the use of a web interface to allow individual users to import polygons that were not imported manually. The interface will be made public once the main import will be finished. Emilie Laffray signature.asc Description: OpenPGP digital signature ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
