Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
My French is no good, so I just tried to open part of discussion here also http://wiki.openstreetmap.org/wiki/Talk:WikiProject_Corine_Land_Cover/Taggin g_scheme. Anyway, I see from French page that the simplification ideas are actually quite similar (e.g. forest is just one forest). We have 55% of land as forest in Estonia, so it is very important for us. My second major proposal would be to skip import of (most) agricultural lands. So please feel free to extend your discussions to the Talk-page I mentioned. Of course, by end of the day tagging in different countries could become slightly different, even Corine classification use itself is not fully unified as far as I know. /Jaak -Original Message- From: talk-boun...@openstreetmap.org [mailto:talk- boun...@openstreetmap.org] On Behalf Of Pieren Sent: 28. mai 2009. a. 13:48 To: talk@openstreetmap.org Subject: Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source... On Thu, May 28, 2009 at 11:17 AM, Jaak Laineste (Nutiteq) j...@nutiteq.com wrote: Hello, are there separate general discussion lists or wiki regarding this Corine land data import ? We just got hands on and permissions for our local (Estonian) data, and some practical questions and ideas have arisen. Yes, we created a Wiki page which is here now: http://wiki.openstreetmap.org/wiki/WikiProject_Corine_Land_Cover We have a group discussing the tagging scheme. A first proposal is available here: http://wiki.openstreetmap.org/wiki/WikiProject_Corine_Land_Cover/Taggin g_scheme but we are going through the list and we already decided some changes. You can see the current progress in the french group (page is in french but tags are in english): http://wiki.openstreetmap.org/wiki/WikiProject_France/Corine_Land_Cover /Nomenclature The discussion is on the french mailing list but we would appreciate if the discussion is internationalized. We also have a page looking for a solution about the import and the conflict resolution with existing landuse in OSM: http://wiki.openstreetmap.org/wiki/WikiProject_Corine_Land_Cover/Corine _Data_Import Pieren ___ 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] Corine Land Cover becomes a potential OSM data source...
Hello, are there separate general discussion lists or wiki regarding this Corine land data import ? We just got hands on and permissions for our local (Estonian) data, and some practical questions and ideas have arisen. First of all strategy in general level: one option is to import and tag everything (huge data amounts) like it is, another would be to combine data categories (e.g. different types of forests) and do not farmlands at all (I assume these are not rendered anyway on maps). I estimate that the second approach could reduce data amounts up to 10 times. /Jaak 2009/5/14 Andy Robinson (blackadder-lists) ajrli...@googlemail.com: Pieren, A new OSMF Working Group is being formed to support groups and individuals with the import of new public and private data. I've copied SteveC who will be leading the group so that your email reaches the new groups radar. Cheers Andy -Original Message- From: talk-boun...@openstreetmap.org [mailto:talk- boun...@openstreetmap.org] On Behalf Of Pieren Sent: 13 May 2009 10:13 PM To: OSM Subject: [OSM-talk] Corine Land Cover becomes a potential OSM data source... at least in France. The Corine Land Cover (CLC) is refering to a european programme establishing a computerised inventory on land cover of the 27 EC member states and other European countries, at an original scale of 1: 100 000, using 44 classes of the 3-level Corine nomenclature. It is produced by the European Environment Agency (EEA) and its member countries and is based on the results of IMAGE2000, a satellite imaging programme undertaken jointly by the Joint Research Centre of the European Commision and the EEA . Until now, the terms of use did not allow commercial use unless the Agency has expressly granted the right to do so. This was stopping any possibility to use the land use data for OSM. But, beginning of 2009, french environment agency (IFEN) released the new version of the dataset of year 2006, called CLC2006 with a newer version of the terms of use which explicitely allow commercial use. After some discussions with the french authorities responsible for the programme in France, it has been clearly stated that the CLC2006 data for France can be imported into OSM. During this discussion, it appeared that the same way of opening the data access has been mentionned at the EEA committee. The ad hoc committee in Q1/2009 suggested the following proposal for the new terms of use: Use rights : EEA is promoting the widest possible use of all data produced during the project. All core land cover data (national and European CLC 2000-2006 changes, national and European CLC2006 and related metadata, high resolution built-up areas, including degree of soil sealing, 2006 and high resolution forest areas, 2006) will be made available free of charge via the web, for non-commercial as well as commercial uses. But until it is released at EEA level, only specific national programmes who officially adopted new terms of use compatible with the OSM licence could take this data as a potential source. For France, we are now looking how we will be able to import some of 44 classes. We also created a wiki page trying to translate the CLC nomenclature to OSM tags: http://wiki.openstreetmap.org/wiki/Corine_Land_Cover I would like to see your comments about this translation table but also more in general, about this data source. It is possible that some other states already used CLC data for OSM, in which case, we would appreciate if they could share their experience. regards, Pieren EEA web site for CLC2006: http://etc-lusi.eionet.europa.eu/CLC2006/ ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk -- Jaak Laineste www.nutiteq.com Skype: jaakl3000 Mobile: +372 509 2586 ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
On Thu, May 28, 2009 at 11:17 AM, Jaak Laineste (Nutiteq) j...@nutiteq.com wrote: Hello, are there separate general discussion lists or wiki regarding this Corine land data import ? We just got hands on and permissions for our local (Estonian) data, and some practical questions and ideas have arisen. Yes, we created a Wiki page which is here now: http://wiki.openstreetmap.org/wiki/WikiProject_Corine_Land_Cover We have a group discussing the tagging scheme. A first proposal is available here: http://wiki.openstreetmap.org/wiki/WikiProject_Corine_Land_Cover/Tagging_scheme but we are going through the list and we already decided some changes. You can see the current progress in the french group (page is in french but tags are in english): http://wiki.openstreetmap.org/wiki/WikiProject_France/Corine_Land_Cover/Nomenclature The discussion is on the french mailing list but we would appreciate if the discussion is internationalized. We also have a page looking for a solution about the import and the conflict resolution with existing landuse in OSM: http://wiki.openstreetmap.org/wiki/WikiProject_Corine_Land_Cover/Corine_Data_Import Pieren ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
Andy Robinson (blackadder-lists) wrote: Pieren, A new OSMF Working Group is being formed to support groups and individuals with the import of new public and private data. I've copied SteveC who will be leading the group so that your email reaches the new groups radar. I am happy to announce that as of today Estonian Environment Information Centre has transferred to us full CORINE 2006 dataset for Estonia (30+ MB zipped ESRI geodatabase) for importing to OSM. So CORINE has already become a data source for OSM. We are definitely open to all ideas and suggestions regarding importing this data. As Estonia has quite limited OSM land cover information I expect import to be relatively painless. Regards, - M ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
The French community has been working hard on getting tags converted to OSM format. A wiki has been created. We have also been analyzing a bit the data to see if it was overlapping with existing polygons to facilitate import of data. It is nice to see that the other countries are following the release of the data. Emilie Laffray On Tue, May 26, 2009 at 3:23 PM, Margus Värton mar...@dakar.ee wrote: Andy Robinson (blackadder-lists) wrote: Pieren, A new OSMF Working Group is being formed to support groups and individuals with the import of new public and private data. I've copied SteveC who will be leading the group so that your email reaches the new groups radar. I am happy to announce that as of today Estonian Environment Information Centre has transferred to us full CORINE 2006 dataset for Estonia (30+ MB zipped ESRI geodatabase) for importing to OSM. So CORINE has already become a data source for OSM. We are definitely open to all ideas and suggestions regarding importing this data. As Estonia has quite limited OSM land cover information I expect import to be relatively painless. Regards, - M ___ 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] Corine Land Cover becomes a potential OSM data source...
Steve Singer ssinger_pg at sympatico.ca writes: On Thu, 14 May 2009, Pieren wrote: For the Canadian GeoBase road import I've been using a plugin to JUMP called RoadMatcher[1] to that detects common roads between two datasets. I've then been excluding ones that aren't likely to not cause conflicts. A few more details at: http://wiki.openstreetmap.org/wiki/Geobase_NRN_-_OSM_Map_Feature RoadMatcher plugin has recently adopted to work with OpenJUMP which is improved and actively developed version of the original JUMP. OpenJUMP is here: http://openjump.org and the corresponding RoadMatcher here: https://sourceforge.net/project/showfiles.php?group_id=118054package_id=320958 ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
On Fri, May 15, 2009 at 3:05 PM, Steve Singer ssinger...@sympatico.ca wrote: On Thu, 14 May 2009, Pieren wrote: For the Canadian GeoBase road import I've been using a plugin to JUMP called RoadMatcher[1] to that detects common roads between two datasets. I've then been excluding ones that aren't likely to not cause conflicts. A few more details at: http://wiki.openstreetmap.org/wiki/Geobase_NRN_-_OSM_Map_Feature Wow, this looks really promising. During the time of the original import a tool like this was discussed but we didn't know if it existed and didn't want to write it ourselves. However, for future reference it would be *really* nice to be able to automatically generate diffs between OSM and external datasets and be able to merge them somehow. I'm thinking of the interface where the two dataset appear next to eachother with differences higlights and you can select which one is correct. Have a nice dat, -- Martijn van Oosterhout klep...@gmail.com http://svana.org/kleptog/ ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
On Thu, 14 May 2009, Pieren wrote: For the Canadian GeoBase road import I've been using a plugin to JUMP called RoadMatcher[1] to that detects common roads between two datasets. I've then been excluding ones that aren't likely to not cause conflicts. A few more details at: http://wiki.openstreetmap.org/wiki/Geobase_NRN_-_OSM_Map_Feature For land-use you could probably write some SQL queries that run against a osm2pgsql instance that find which of your land-use polygons don't intersect with existing osm land-use tagged areas. I don't mind joining a mailing list for the data import working group, but I have enough on my plate with the various Canadian datasets that I won't be able to provide assistance scripting or importing to others for sometime. I'm not sure if I'll be able to make weekly conference calls either. [1] - http://www.jump-project.org/portal.php?PID=PL Steve Thanks for your suggestions. In our case, it is more about 80% of the import dataset that will stay. And we speak about e.g. forests, tree plantations, agricultural areas,etc which is a huge surface in France. Tha't why our preference would go to some solution making the import as much as possible automatic. For instance, a tool detecting conflicts between the new landuse polygones and existing landuse polygones, then splitting the data in two sets, one without conflicts which could be directly uploaded and one with conflicts which would require some individual investigation/edits. Pieren ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk-nl] [OSM-talk] Corine Land Cover becomes a potential OSM data source...
- Original Message From: OpenStreetMap NL discussion list talk-nl@openstreetmap.org To: OpenStreetMap NL discussion list talk-nl@openstreetmap.org Subject: Re: [OSM-talk-nl] [OSM-talk] Corine Land Cover becomes a potential OSM data source... Date: 14/05/09 18:31 Martijn van Exel wrote: gt; I disagree with you on relevance. The NWB data is likely to be more gt; current and much more detailed than the AND data, so much of it will gt; be relevant and usable - although not directly. Onzin :) Van iemand enorm dicht bij de bron: gt; Verwacht trouwens niet te veel van het NWB zelf. Openstreetmap heeft gt; nu al een betere kwaliteit dan het NWB en dat verschil zal alleen gt; maar toenemen. Maar het NWB is wel de sleutel tot een heleboel gt; overheidsdata en daarom is een koppeling NWB-openstreetmap meer dan gt; wenselijk. Met als bijeffect dat je dan dus fouten in beide bestanden gt; op kunt sporen. gt; While the level of detail would make for a massive import, setting up gt; a NWB WMS layer to be used in JOSM would probably be a better idea gt; than to import the entire NWB into the OSM db. OSM zal eerst eens wat read only lagen moeten introduceren. Om zo scheiding van data te bewerkstelligen en daarvanaf branches te kunnen maken. Dan winnen er meer partijen... overheid en OSM. Helemaal mee eens. Dingen als grenzen en plaatsnamen enzo. Stefan ___ Talk-nl mailing list Talk-nl@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-nl Message sent using UebiMiau 2.7.10 ___ Talk-nl mailing list Talk-nl@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-nl
Re: [OSM-talk-nl] [OSM-talk] Corine Land Cover becomes a potential OSM data source...
2009/5/14 Stefan de Konink ste...@konink.de: Martijn van Exel wrote: I disagree with you on relevance. The NWB data is likely to be more current and much more detailed than the AND data, so much of it will be relevant and usable - although not directly. Onzin :) Van iemand enorm dicht bij de bron: Verwacht trouwens niet te veel van het NWB zelf. Openstreetmap heeft nu al een betere kwaliteit dan het NWB en dat verschil zal alleen maar toenemen. Maar het NWB is wel de sleutel tot een heleboel overheidsdata en daarom is een koppeling NWB-openstreetmap meer dan wenselijk. Met als bijeffect dat je dan dus fouten in beide bestanden op kunt sporen. Als dat zo is - wie is diegene dicht bij de bron en vanwaar de geheimzinnigheid? - dan moeten we onze strategie daarop aanpassen. Ik neem niet zomaar aan dat A of B beter is en denk dat we tzt de mogelijkheid moeten creëren om de kwaliteit onderling te vergelijken. Daarvoor is het noodzakelijk om referentiedata te hebben - deels kunnen daar onze collectieve GPS-tracks voor dienen, maar een recente, kwalitatief goede sedt luchtfoto's lijkt me nog het beste. While the level of detail would make for a massive import, setting up a NWB WMS layer to be used in JOSM would probably be a better idea than to import the entire NWB into the OSM db. OSM zal eerst eens wat read only lagen moeten introduceren. Om zo scheiding van data te bewerkstelligen en daarvanaf branches te kunnen maken. Dan winnen er meer partijen... overheid en OSM. Het ligt eraan wat je uitgangspunt is voor OSM. Sommigen zien OSM als zuiver 'de beste vrije wereldkaart', terwijl anderen - waartoe ik mezelf reken, en jij ook toch Stefan? - OSM uiteindelijk zien als de ultieme repository voor vrije geodata. Eerstgenoemden zullen elke import / toegoeging zien in het licht van het kaartproduct, zij het op het scherm, zij het wellicht in andere vormen. Laatstgenoemden zullen neigen naar het openstellen van de OSM-database voor (al dan niet read-only) lagen / tabellen met 'externe' bronnen. Zo krijg je de synergie tussen aanleverende partijen waar jij het ook over hebt. Op dat punt zijn we het dus volgens mij dan weer wel eens. Tenslotte, plug: Zaterdag mappen in Hoofddorp! Kom ook! http://wiki.openstreetmap.org/wiki/Netherlands_Mapping_Parties_2009#Mapping_Party_-_Hoofddorp_-_zaterdag_16_mei martijn van exel -+- mve...@gmail.com -+- http://www.schaaltreinen.nl/ ___ Talk-nl mailing list Talk-nl@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-nl
Re: [OSM-talk-nl] [OSM-talk] Corine Land Cover becomes a potential OSM data source...
Martijn van Exel wrote: 2009/5/14 Stefan de Konink ste...@konink.de: Martijn van Exel wrote: I disagree with you on relevance. The NWB data is likely to be more current and much more detailed than the AND data, so much of it will be relevant and usable - although not directly. Onzin :) Van iemand enorm dicht bij de bron: Verwacht trouwens niet te veel van het NWB zelf. Openstreetmap heeft nu al een betere kwaliteit dan het NWB en dat verschil zal alleen maar toenemen. Maar het NWB is wel de sleutel tot een heleboel overheidsdata en daarom is een koppeling NWB-openstreetmap meer dan wenselijk. Met als bijeffect dat je dan dus fouten in beide bestanden op kunt sporen. Als dat zo is - wie is diegene dicht bij de bron en vanwaar de geheimzinnigheid? - dan moeten we onze strategie daarop aanpassen. Ik neem niet zomaar aan dat A of B beter is en denk dat we tzt de mogelijkheid moeten creëren om de kwaliteit onderling te vergelijken. De kwaliteit kun je al vergelijken, ga naar maximumsnelheden.nl en zoom in op de mapnik layer. Dan zul je zien dat het NWB is vervuild met 'nieuwe kaart' materiaal. En er nog een aantal dingen niet in orde zijn. Daarvoor is het noodzakelijk om referentiedata te hebben - deels kunnen daar onze collectieve GPS-tracks voor dienen, maar een recente, kwalitatief goede sedt luchtfoto's lijkt me nog het beste. 15m is al beschikbaar in CC-BY-SA. 1m is al beschikbaar in CC-BY-NC. Gezien we zelf bezig zijn met de 1px=7cm set lijkt me de optie 'vooral doorgaan' en het aanbod om een getfeaturebyXY te gaan doen op het NWB om zo al onze wegen te taggen met een maxspeed een heel erg goede suggestie. While the level of detail would make for a massive import, setting up a NWB WMS layer to be used in JOSM would probably be a better idea than to import the entire NWB into the OSM db. OSM zal eerst eens wat read only lagen moeten introduceren. Om zo scheiding van data te bewerkstelligen en daarvanaf branches te kunnen maken. Dan winnen er meer partijen... overheid en OSM. Het ligt eraan wat je uitgangspunt is voor OSM. Sommigen zien OSM als zuiver 'de beste vrije wereldkaart', terwijl anderen - waartoe ik mezelf reken, en jij ook toch Stefan? - OSM uiteindelijk zien als de ultieme repository voor vrije geodata. ...dat zien we inderdaad beide zo, maar zoals al aangegeven niet in deze vorm. Te veel mensen kunnen te eenvoudig feitelijke data slopen. Eerstgenoemden zullen elke import / toegoeging zien in het licht van het kaartproduct, zij het op het scherm, zij het wellicht in andere vormen. Laatstgenoemden zullen neigen naar het openstellen van de OSM-database voor (al dan niet read-only) lagen / tabellen met 'externe' bronnen. Zo krijg je de synergie tussen aanleverende partijen waar jij het ook over hebt. Op dat punt zijn we het dus volgens mij dan weer wel eens. Eigenlijk zijn we het altijd met elkaar eens ;) Tenslotte, plug: Zaterdag mappen in Hoofddorp! Kom ook! http://wiki.openstreetmap.org/wiki/Netherlands_Mapping_Parties_2009#Mapping_Party_-_Hoofddorp_-_zaterdag_16_mei Het staat op de planning, Floris weet m'n 'constraint'. Stefan ___ Talk-nl mailing list Talk-nl@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-nl
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
Pieren, A new OSMF Working Group is being formed to support groups and individuals with the import of new public and private data. I've copied SteveC who will be leading the group so that your email reaches the new groups radar. Cheers Andy -Original Message- From: talk-boun...@openstreetmap.org [mailto:talk- boun...@openstreetmap.org] On Behalf Of Pieren Sent: 13 May 2009 10:13 PM To: OSM Subject: [OSM-talk] Corine Land Cover becomes a potential OSM data source... at least in France. The Corine Land Cover (CLC) is refering to a european programme establishing a computerised inventory on land cover of the 27 EC member states and other European countries, at an original scale of 1: 100 000, using 44 classes of the 3-level Corine nomenclature. It is produced by the European Environment Agency (EEA) and its member countries and is based on the results of IMAGE2000, a satellite imaging programme undertaken jointly by the Joint Research Centre of the European Commision and the EEA . Until now, the terms of use did not allow commercial use unless the Agency has expressly granted the right to do so. This was stopping any possibility to use the land use data for OSM. But, beginning of 2009, french environment agency (IFEN) released the new version of the dataset of year 2006, called CLC2006 with a newer version of the terms of use which explicitely allow commercial use. After some discussions with the french authorities responsible for the programme in France, it has been clearly stated that the CLC2006 data for France can be imported into OSM. During this discussion, it appeared that the same way of opening the data access has been mentionned at the EEA committee. The ad hoc committee in Q1/2009 suggested the following proposal for the new terms of use: Use rights : EEA is promoting the widest possible use of all data produced during the project. All core land cover data (national and European CLC 2000-2006 changes, national and European CLC2006 and related metadata, high resolution built-up areas, including degree of soil sealing, 2006 and high resolution forest areas, 2006) will be made available free of charge via the web, for non-commercial as well as commercial uses. But until it is released at EEA level, only specific national programmes who officially adopted new terms of use compatible with the OSM licence could take this data as a potential source. For France, we are now looking how we will be able to import some of 44 classes. We also created a wiki page trying to translate the CLC nomenclature to OSM tags: http://wiki.openstreetmap.org/wiki/Corine_Land_Cover I would like to see your comments about this translation table but also more in general, about this data source. It is possible that some other states already used CLC data for OSM, in which case, we would appreciate if they could share their experience. regards, Pieren EEA web site for CLC2006: http://etc-lusi.eionet.europa.eu/CLC2006/ ___ 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] Corine Land Cover becomes a potential OSM data source...
On Thu, May 14, 2009 at 11:05 AM, Andy Robinson (blackadder-lists) ajrli...@googlemail.com wrote: Pieren, A new OSMF Working Group is being formed to support groups and individuals with the import of new public and private data. I've copied SteveC who will be leading the group so that your email reaches the new groups radar. Cheers Andy Thanks for your reply. I think we are exactly in this case. Once we agree which layers will be imported and the tag mapping, we can easily convert the data to osm format. But we are facing the problem of how to detect and resolve the conflicts with existing landuse data in OSM which are usually (but not always) more accurate (Yahoo imagery or french cadastre). I know we are not the first country having this problem and we would like to know how it was solved with other mass imports. Pieren ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
Hi Pieren wrote: But we are facing the problem of how to detect and resolve the conflicts with existing landuse data in OSM which are usually (but not always) more accurate (Yahoo imagery or french cadastre). I know we are not the first country having this problem and we would like to know how it was solved with other mass imports. We are currently importing public transport information for the UK (NaPTAN) and are having a similar problem with existing data in OSM. Our approach is to tag the imported data specially so that it can easily be found in the database but does not show up on the map (i.e. all imported bus stops are tagged with source=naptan_import but do not have highway=bus_stop tag). Mappers can then check the imported data and activate bus stops by adding a highway=bus_stop tag to it or copy the imported tags over to an existing OSM bus stop. To aid this task a web-based tool is being developed at the moment. The tool displays all bus stops from OSM and all newly imported stops on a map and can find pairs of similar bus stops in the existing and the new dataset (based on proximity and similar names). The user can then easily select pairs and merge them. Christoph ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
Interesting. We will be having the same kind of import versus existing data issue once more (if and) when the Nationaal Wegenbestand (National road database) is made available by the transport ministry. I did not give it much thought yet - may be some of my Dutch co-OSM-ers will have a more elaborate train of thought - but I was thinking along similar lines: import the data but tag it so that it won't show up in the map. Have a dedicated map layer on a local server that can be used to check existing user data, may be even a Potlatch layer if at all possible. I guess JOSM support would be easy enough through the WMS plugin. Other noteworthy fact is that we have had this issue before, in 2007 when importing the AND data. There was much discussion then on how to approach, and as far as I can remember we had an opt-out system then: if you don't want your current user data overwritten by AND data, you could indicate such and this way some particularly well mapped areas were untouched by the AND data. But back then, the Netherlands had only isolated pockets of good OSM coverage. martijn van exel -+- mve...@gmail.com -+- http://www.schaaltreinen.nl/ On Thu, May 14, 2009 at 12:23, Christoph Boehme christ...@b3e.net wrote: Hi Pieren wrote: But we are facing the problem of how to detect and resolve the conflicts with existing landuse data in OSM which are usually (but not always) more accurate (Yahoo imagery or french cadastre). I know we are not the first country having this problem and we would like to know how it was solved with other mass imports. We are currently importing public transport information for the UK (NaPTAN) and are having a similar problem with existing data in OSM. Our approach is to tag the imported data specially so that it can easily be found in the database but does not show up on the map (i.e. all imported bus stops are tagged with source=naptan_import but do not have highway=bus_stop tag). Mappers can then check the imported data and activate bus stops by adding a highway=bus_stop tag to it or copy the imported tags over to an existing OSM bus stop. To aid this task a web-based tool is being developed at the moment. The tool displays all bus stops from OSM and all newly imported stops on a map and can find pairs of similar bus stops in the existing and the new dataset (based on proximity and similar names). The user can then easily select pairs and merge them. Christoph ___ 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] Corine Land Cover becomes a potential OSM data source...
Hi, Martijn van Exel wrote: Interesting. We will be having the same kind of import versus existing data issue once more (if and) when the Nationaal Wegenbestand (National road database) is made available by the transport ministry. I did not give it much thought yet - may be some of my Dutch co-OSM-ers will have a more elaborate train of thought - but I was thinking along similar lines: import the data but tag it so that it won't show up in the map. Have a dedicated map layer on a local server that can be used to check existing user data, may be even a Potlatch layer if at all possible. I guess JOSM support would be easy enough through the WMS plugin. This way (import everything with funny tagging and let users then work with it) is, in my eyes, suitable if you have reason to believe that a lot of what you import is actually going to be used. If, on the other hand, it is likely that 80% of what you import will later be deleted because it was there already, then it may be better to run the import like we did the Nordrhein-Westfalen import: Set up a completely separate WMS layer that can be shown in the backgroundd and allow people to import individual objects where they think it's worthwile. Bye Frederik ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
Thanks for your suggestions. In our case, it is more about 80% of the import dataset that will stay. And we speak about e.g. forests, tree plantations, agricultural areas,etc which is a huge surface in France. Tha't why our preference would go to some solution making the import as much as possible automatic. For instance, a tool detecting conflicts between the new landuse polygones and existing landuse polygones, then splitting the data in two sets, one without conflicts which could be directly uploaded and one with conflicts which would require some individual investigation/edits. Pieren ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
On 14/05/09 11:23, Christoph Boehme wrote: We are currently importing public transport information for the UK (NaPTAN) and are having a similar problem with existing data in OSM. Our approach is to tag the imported data specially so that it can easily be found in the database but does not show up on the map (i.e. all imported bus stops are tagged with source=naptan_import but do not have highway=bus_stop tag). Mappers can then check the imported data and activate bus stops by adding a highway=bus_stop tag to it or copy the imported tags over to an existing OSM bus stop. Has consideration been given in any future API/database schema updates of having support for layers, like e.g. the GIMP does it for graphics? So you could import data into a new layer with all the appropriate tags and then merge down as appropriate. The slippy map would only show the base layer. This sort of support might also be useful for people reusing OSM software and schemas who want to overlay their own data onto the OSM data. Gerv ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
Martijn van Exel wrote: Interesting. We will be having the same kind of import versus existing data issue once more (if and) when the Nationaal Wegenbestand (National road database) is made available by the transport ministry. I did not give it much thought yet - may be some of my Dutch co-OSM-ers will have a more elaborate train of thought - but I was thinking along similar lines: import the data but tag it so that it won't show up in the map. Have a dedicated map layer on a local server that can be used to check existing user data, may be even a Potlatch layer if at all possible. I guess JOSM support would be easy enough through the WMS plugin. I have to agree with Frederik on this. With the NWB you mention, it's likely the vast majority of it will not be relevant to NL coverage, since we already have had a mass import of AND data. Why should we 'pollute' the OSM db with data that's not going to be used on a large scale. I'd too rather see such a source of data on a separate layer/db, through WMS for instance, so that we can compare, and (like Gervase puts it) merge down what we want. -- Lennard ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Corine Land Cover becomes a potential OSM data source...
On Thu, May 14, 2009 at 16:52, Lennard l...@xs4all.nl wrote: Martijn van Exel wrote: Interesting. We will be having the same kind of import versus existing data issue once more (if and) when the Nationaal Wegenbestand (National road database) is made available by the transport ministry. I did not give it much thought yet - may be some of my Dutch co-OSM-ers will have a more elaborate train of thought - but I was thinking along similar lines: import the data but tag it so that it won't show up in the map. Have a dedicated map layer on a local server that can be used to check existing user data, may be even a Potlatch layer if at all possible. I guess JOSM support would be easy enough through the WMS plugin. I have to agree with Frederik on this. With the NWB you mention, it's likely the vast majority of it will not be relevant to NL coverage, since we already have had a mass import of AND data. Why should we 'pollute' the OSM db with data that's not going to be used on a large scale. I disagree with you on relevance. The NWB data is likely to be more current and much more detailed than the AND data, so much of it will be relevant and usable - although not directly. While the level of detail would make for a massive import, setting up a NWB WMS layer to be used in JOSM would probably be a better idea than to import the entire NWB into the OSM db. I don't know what kind of layer Potlatch would be able to handle, but we could probably try and provide a yahoo-compatible tile layer on top of the WMS. T o be discussed on talk-nl - we're getting OT here ;) I'd too rather see such a source of data on a separate layer/db, through WMS for instance, so that we can compare, and (like Gervase puts it) merge down what we want. Yes - I'm with you there, although we should take a close look at NWB to see if there's some direct import possibilities. I'm thinking for example cycle paths along roads, but I don't want to get ahead of things here. -- Martijn ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk-nl] [OSM-talk] Corine Land Cover becomes a potential OSM data source...
On Thu, May 14, 2009 at 16:52, Lennard l...@xs4all.nl wrote: Martijn van Exel wrote: Interesting. We will be having the same kind of import versus existing data issue once more (if and) when the Nationaal Wegenbestand (National road database) is made available by the transport ministry. I did not give it much thought yet - may be some of my Dutch co-OSM-ers will have a more elaborate train of thought - but I was thinking along similar lines: import the data but tag it so that it won't show up in the map. Have a dedicated map layer on a local server that can be used to check existing user data, may be even a Potlatch layer if at all possible. I guess JOSM support would be easy enough through the WMS plugin. I have to agree with Frederik on this. With the NWB you mention, it's likely the vast majority of it will not be relevant to NL coverage, since we already have had a mass import of AND data. Why should we 'pollute' the OSM db with data that's not going to be used on a large scale. I disagree with you on relevance. The NWB data is likely to be more current and much more detailed than the AND data, so much of it will be relevant and usable - although not directly. While the level of detail would make for a massive import, setting up a NWB WMS layer to be used in JOSM would probably be a better idea than to import the entire NWB into the OSM db. I don't know what kind of layer Potlatch would be able to handle, but we could probably try and provide a yahoo-compatible tile layer on top of the WMS. T o be discussed on talk-nl - we're getting OT here ;) I'd too rather see such a source of data on a separate layer/db, through WMS for instance, so that we can compare, and (like Gervase puts it) merge down what we want. Yes - I'm with you there, although we should take a close look at NWB to see if there's some direct import possibilities. I'm thinking for example cycle paths along roads, but I don't want to get ahead of things here. -- Martijn ___ Talk-nl mailing list Talk-nl@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-nl
Re: [OSM-talk-nl] [OSM-talk] Corine Land Cover becomes a potential OSM data source...
Martijn van Exel wrote: I disagree with you on relevance. The NWB data is likely to be more current and much more detailed than the AND data, so much of it will be relevant and usable - although not directly. Onzin :) Van iemand enorm dicht bij de bron: Verwacht trouwens niet te veel van het NWB zelf. Openstreetmap heeft nu al een betere kwaliteit dan het NWB en dat verschil zal alleen maar toenemen. Maar het NWB is wel de sleutel tot een heleboel overheidsdata en daarom is een koppeling NWB-openstreetmap meer dan wenselijk. Met als bijeffect dat je dan dus fouten in beide bestanden op kunt sporen. While the level of detail would make for a massive import, setting up a NWB WMS layer to be used in JOSM would probably be a better idea than to import the entire NWB into the OSM db. OSM zal eerst eens wat read only lagen moeten introduceren. Om zo scheiding van data te bewerkstelligen en daarvanaf branches te kunnen maken. Dan winnen er meer partijen... overheid en OSM. Stefan ___ Talk-nl mailing list Talk-nl@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-nl
[OSM-talk] Corine Land Cover becomes a potential OSM data source...
at least in France. The Corine Land Cover (CLC) is refering to a european programme establishing a computerised inventory on land cover of the 27 EC member states and other European countries, at an original scale of 1: 100 000, using 44 classes of the 3-level Corine nomenclature. It is produced by the European Environment Agency (EEA) and its member countries and is based on the results of IMAGE2000, a satellite imaging programme undertaken jointly by the Joint Research Centre of the European Commision and the EEA . Until now, the terms of use did not allow commercial use unless the Agency has expressly granted the right to do so. This was stopping any possibility to use the land use data for OSM. But, beginning of 2009, french environment agency (IFEN) released the new version of the dataset of year 2006, called CLC2006 with a newer version of the terms of use which explicitely allow commercial use. After some discussions with the french authorities responsible for the programme in France, it has been clearly stated that the CLC2006 data for France can be imported into OSM. During this discussion, it appeared that the same way of opening the data access has been mentionned at the EEA committee. The ad hoc committee in Q1/2009 suggested the following proposal for the new terms of use: Use rights : EEA is promoting the widest possible use of all data produced during the project. All core land cover data (national and European CLC 2000-2006 changes, national and European CLC2006 and related metadata, high resolution built-up areas, including degree of soil sealing, 2006 and high resolution forest areas, 2006) will be made available free of charge via the web, for non-commercial as well as commercial uses. But until it is released at EEA level, only specific national programmes who officially adopted new terms of use compatible with the OSM licence could take this data as a potential source. For France, we are now looking how we will be able to import some of 44 classes. We also created a wiki page trying to translate the CLC nomenclature to OSM tags: http://wiki.openstreetmap.org/wiki/Corine_Land_Cover I would like to see your comments about this translation table but also more in general, about this data source. It is possible that some other states already used CLC data for OSM, in which case, we would appreciate if they could share their experience. regards, Pieren EEA web site for CLC2006: http://etc-lusi.eionet.europa.eu/CLC2006/ ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
