Re: [OSM-talk] Organizational mapping policy
Hi Paul, I'm curious how HOT projects which are mentioned relate to this. What I mean is we frequently train other non-governmental organizations (NGOs), governments and universities in OSM. The proposed guidelines read to me that people would have to declare that they were being paid to map. A frequent scenario is we find a group that could benefit from OSM mapping, for example a civil society group (CSO) and show them how OSM works. They may then decide to incorporate it into one of their own projects. Some of the CSOs have a mix of volunteers and staff, so would both types of participants need to declare what they were doing? Frequently people are more being paid to provide training than to map directly in OSM. Is this another scenario? Should HOT contractors/staff then declare that they are being paid to train people in OSM? I don't think using separate accounts is a great idea for me personally I would have no idea when I should use one account versus an another. I would be perfectly happy to declare that I work for HOT on my user page, which it already does(1). Thanks, -Kate (1) http://www.openstreetmap.org/user/wonderchook On Wed, May 14, 2014 at 6:44 AM, Paul Norman penor...@mac.com wrote: We have more and more organizations and businesses mapping in OSM. Multiple organizations have been conducting paid editing in Europe and the US. This generally comes to light *after* complaints are made - with the company usually not identifying who they are, what their goals are, and what they want, beforehand. There have also been difficulties determining what has been mapped on behalf of an organization. We will likely see more of this type of editing in the future, and while not necessarily bad, there are differences between it and normal editing. Recent events in a project similar to OpenStreetMap - Wikipedia - have demonstrated that the participation of organizations in data editing can occasionally lead to misunderstandings or disharmony in the project, particularly where a lack of transparency is involved. For this reason the DWG is considering if it is necessary to issue guidelines for organizational editing. Some previous discussion is at http://lists.osm.org/pipermail/osmf-talk/2013-November/002344.html There are some activities we do not want to cover in the guidelines - Unorganized editing by employees, e.g. a shop owner adding their shop or nearby details to the map - Editors mapping in response to a contest or similar where the contest organizer does not have the power to require them to edit - Individuals who, on their own accord, decide to participate in an organised effort or challenge, like local mapping parties, Mapathons, HOT projects, etc Some possible guideline requirements could involve - Disclosing those who are directing them (e.g. employers or who they are contracting for) on the users page - Creating a wiki page with links to user pages of users mapping under an organization's direction - Requiring those working on broader projects to communicate and get feedback from the community before starting - Requiring disclosure of proprietary third-party sources used. Organizations may have data from third parties that they can legally use when contributing to OSM, but aren't able to directly show others the data - Maintaining separate accounts if doing both personal and organizational editing The extent of editing activities covered is something else that needs to be discussed. Some types of activities that *could* be covered are - Teachers requiring their students to edit OSM as part of a course - Consultants editing for multiple clients - Being required to edit as part of an employment relationship SEO spammers would be covered by this policy, but are not the target. They would ignore it, so we'll just end up using the existing tools of reverting and blocking. Paul Norman For the Data Working Group ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
Paul Are there actually examples of larger scale paid mapping that haven't been disclosed and that have been detected by the DWG? I know we have now and then had companies adding their outlets/operations but that would typically be covered by the import guidelines in any case. Simon signature.asc Description: OpenPGP digital signature ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, On 05/14/14 08:55, Simon Poole wrote: Are there actually examples of larger scale paid mapping that haven't been disclosed and that have been detected by the DWG? Not by the DWG - but by the community, yes, this happens quite frequently. In Germany alone in the past year there have been four instances that I know of where a company or agency started a project that involved modifying OSM data and they were then found out by the community (one case involved the micro mapping of railway stations, one was cycle routes in Lower Saxony, and two were projects that aimed to improve disabled/OAP mobility by mapping sidewalks as separate footpaths). All of these cases had in common that there was to bad intent on the part of the organisation doing the mapping, but sometimes dangerously little knowledge about OSM and its community, and in each time it took the community a while to identify the people behind it and establish a working channel of communication. I don't know much about cases in other parts of the world; both Mapbox and Telenav are quite open about employing mappers (Telenav have had a talk at last SOTM-US about this). Their openness about this is welcome but at the moment totally voluntary; there are likely a bunch of other companies who do the same kinds of things but do not talk about it. Bye Frederik - -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09 E008°23'33 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQEcBAEBAgAGBQJTcxqpAAoJEOx/uhGAJu9HGLQH/j/Ig59gdK3kYNXdUomYAVGL tweZCVCP9NaQXq9brISVHy9wuHyqv6Q4/ZU5Erj2h1D5rACx08nq6+DpfRLTRwpk NTFpcpBRqzPmTWXyKoupZ1WWCWB8IUqHGmr3FsL4+AKJh3hIBjXXPXtNjiu3O215 At63EAguJZfCZyivCQWsSSAURzmWYbHjJudGd0PsS1lNRJ4h0a0a9QFzpgkGD5ud UCxMafPna2TJhRCewIR5TcjQVWTgCbAJuNR0FT+Qn0a9vbALFgMmyTP21K1zztLa W1ZswxfyReh/8x5afa/1R/k0YV/NvsJddkFUdgrNaHVonxlnstdpUxIZG/AyHu0= =0VcQ -END PGP SIGNATURE- ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
Kate, On 05/14/14 08:45, Kate Chapman wrote: I'm curious how HOT projects which are mentioned relate to this. What I mean is we frequently train other non-governmental organizations (NGOs), governments and universities in OSM. The proposed guidelines read to me that people would have to declare that they were being paid to map. This is something that we'll have to discuss and find a good definition. Personally, I think that a policy like that should cover any kind of (for lack of better word) directed mapping, where a mapper doesn't act on their own accord (because they want to) but on someone else's (because they're told to). The boundary is of course blurry - if you report for duty at your local CSO, on your own accord, because you want to make the work a better place, and then are told that this week's project is fixing TIGER roads in rural Pennsylvania - are you directed? You could even be paid without being directed, even if that might be a rare exception - in most cases your employer or client would at least give you a general direction like improve routing in the US or so. And that's the interesting bit - not whether or not someone is paid, but accounts A,B,C are employees of the X project at Y company and they're working on Z. In my view, any kind of training would also be directed, whether or not people receive compensation for it. It would be good for a group of trainees to be somehow identifiable by the community - a simple disclosure like I'm part of the XY class instructed by Z and we're mapping from aerial imagery in Rwanda would already go a long way. (Only last week I had to employ the help of OSM admins to puzzle together a list of likely pupils from a school class in Kazakhstan, most of whom were making fictional edits - it would have helped a lot to have some up-front indication about who was their instructor.) Should HOT contractors/staff then declare that they are being paid to train people in OSM? I don't think using separate accounts is a great idea for me personally I would have no idea when I should use one account versus an another. I would be perfectly happy to declare that I work for HOT on my user page, which it already does(1). You are probably an exceptional case because even when you map for HOT it will likely be in the role of a director rather than directee ;) I guess I'd expect someone who partly does directed work and partly his own - say, they armchair-map Brazil for their employer during the day but survey their home town in private on weekends - to use different accounts, but in cases where it is not clear-cut, people could pehaps still indicate affiliation on a per-changeset basis, e.g. if you were to do some mapping in the classroom as part of a training project you could simply throw in the project name in the changeset comment or so - something you'd likely do anyway. Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09 E008°23'33 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, On 05/14/14 09:26, Frederik Ramm wrote: All of these cases had in common that there was to bad intent Should have been no bad intent. Bye Frederik - -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09 E008°23'33 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQEcBAEBAgAGBQJTcyKMAAoJEOx/uhGAJu9HNioH/iG1B5X1XvSRHQeewTppa3/Y 9j2N6sLrHQ0kvRKYYFY2eRhwGlwzwrpMYHtiX10LfhjEkmIBMnmetgChQY68fokZ s784PBmYaNQ77+nP8+4UN29KMOlY4YAJ5taIBMBv9x3Hqtb/iomnFOTYm2ZAjEXI jUUeXSwInIPDclmucG7KHxF3hCazxcydA1zG71s4OPsOtz9grXvuswQeToOwEeOE w8s3A2hQx6mrGRsI6OKuHreP5OcRVKS29w3QTiOMkBuQ+zKfM2rakJrjFwWJ5idj XF/eeb0/8zaSia50AZbMHeCGWgca99RI4g6GXBNjQ2q55ppKmSr0bMR7AQDnBe4= =qM1y -END PGP SIGNATURE- ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
On Wednesday 14 May 2014, Frederik Ramm wrote: Personally, I think that a policy like that should cover any kind of (for lack of better word) directed mapping, where a mapper doesn't act on their own accord (because they want to) but on someone else's (because they're told to). The boundary is of course blurry - if you report for duty at your local CSO, on your own accord, because you want to make the work a better place, and then are told that this week's project is fixing TIGER roads in rural Pennsylvania - are you directed? Exactly. It seems to me the distinction between those activities you specifically do not want to regulate and those you do want to cover is problematic. The core question seems to me what exactly the aim of such a policy is. If it is aimed at companies who have people edit in OSM the policy should define its scope in terms of these companies, not in terms of the editing activities they endorse. One possible point is that organizations developing certain rules for mapping on their own (like regarding tagging or use of geometries to represent certain things) and instruct others to use these rules they are required to discuss/document these with the community first. Such policy would be independent of how exactly people are instructed by the organization - if they are paid or just volunteers. If on the other hand the editing activities themselves are considered the primary issue the question is what aspect of them is considered to be the problem and this should make the core of the definition. Based on the issues in Wikipedia Paul refers to for example the possible conflicts of interest might be the main issue and if that is the case it might be best to require any mapper to disclose possible conflicts of interest on their user page. The use of proprietary third-party sources is for example an issue not limited to organizational mapping at all, it is a frequent occurence that people use proprietary data they have access to (for example as part of their work but without their employer being involved) as a mapping source - such sources should probably be required to be disclosed even if the mapping itself is a totally private activity. -- Christoph Hormann http://www.imagico.de/ ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
On Wed, May 14, 2014 at 1:44 AM, Paul Norman penor...@mac.com wrote: editing can occasionally lead to misunderstandings or disharmony in the project, harmony is not the first word coming in my mind about OSM... Some types of activities that *could* be covered are - Teachers requiring their students to edit OSM as part of a course I hope not. How can we have on one side a (foundation) group trying to reduce barriers for newcomers and on the other side a group increasing bureaucracy for newcomers and their teachers ? Pieren ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
[OSM-talk] Command-line program to remove untagged, unconnected nodes from .osm file
Dear list, I am trying to find a way to get rid of untagged, unconnected nodes in a .osm file. I know JOSM can do this but I need a command-line application. I've been trying to do it with osmosis and osmfilter without success (pattern *=* not allowed). Any ideas? Regards, Juan___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
Þann 14.05.2014 08:49, Pieren reit: On Wed, May 14, 2014 at 1:44 AM, Paul Norman penor...@mac.com wrote: Some types of activities that *could* be covered are - Teachers requiring their students to edit OSM as part of a course I hope not. How can we have on one side a (foundation) group trying to reduce barriers for newcomers and on the other side a group increasing bureaucracy for newcomers and their teachers ? Quite. I've been mulling over setting together a session that teachers could use in higher classes in primary school, aimed at introducing the children to the concept of not being only consumers but also participants and creators of material online. An introduction to not only OpenStreetMap but also Wikipedia, Project Gutenberg and other such open data initiatives. Requiring these teachers to get each of their students to not only create a OSM user but also to put a boiler-plate disclaimer on their user pages as they map the playgrounds or sport fields they attend mostly, or wastebins near their school, seems an overkill and it looks like the red tape that is choking the English Wikipedia (new contributor numbers and engagement is dropping steadily) is edging closer to OSM, to my personal dismay (I hardly touch the English Wikipedia these days, having to display signed sheets in triplicate from librarians confirming the knowledge is real - or that is what it feels like - when my edits haven't been reverted by disbelieving bots). The focus needs to be on the problem at hand, which I gather is companies hiring people to map things using their own methodology incompatible with current OSM tagging guidelines. Is that correct understanding? The focus needs to be on a problem at hand, not on increasing bureaucracy for everyone acting with a common instructor or goal in mind. --Jói ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Command-line program to remove untagged, unconnected nodes from .osm file
Hi, I am trying to find a way to get rid of untagged, unconnected nodes in a .osm file. I know JOSM can do this but I need a command-line application. I've been trying to do it with osmosis and osmfilter without success (pattern *=* not allowed). Any ideas? If your data file is well-formed, a simple grep will do: grep -v node .*/ Cheers, Nik signature.asc Description: This is a digitally signed message part. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
[OSM-talk] OSM France BANO project... openaddresses in France
OSM-FR has initiated a project to create an open database of addresses in France. Data sources will be OSM + available opendata + address data collected from the cadastre. The resulting dataset will be under ODbL (because of OSM data and also many opendata sets are also under ODbL). In the past weeks, we've been coding the collector scripts which are also matching non OSM data with OSM data, to get better street names (Avenue des Champs-Élysées instead of AV DES CHAMPS ELYSEES). We've starting collecting data a few days ago and the progress can be seen here: http://openstreetmap.fr/outils/bano/status A BANO overlay rendering is also available to view the coverage and density: http://tile.openstreetmap.fr/~cquest/leaflet/bano.html#6/46.988/0.912 green = OSM data orange = opendata blue = cadastre + OSM enhancements red = cadastre only The resulting dataset will not be imported as is in OSM as the french community considers it needs to be manually reviewed. As a first step, we prefer to concentrate on making sure we reduce the number of missing streets and street names and this is simplified by the automatic matching that is done between OSM and non-OSM data. When no match is found, we make it visible on the BANO rendered layer at the higher zoom levels, and showing the expected name and street code. Here is an example: http://tile.openstreetmap.fr/~cquest/leaflet/bano.html#17/48.82902/2.31073 We're about to publish the first datasets, one for each departement (roughtly 100.000 to 1.000.000 addresses depending on the area) which be of course listed on openaddresses.io We expect to have a coverage of 80%-90% (roughtly 20 million addresses). For more info, check the wiki page (in French only, translations welcome): The BANO team is also reachable on IRC #bano (irc://irc.oftc.net:6667/bano) and on twitter https://twitter.com/ProjetBANO -- Christian Quest - OpenStreetMap France ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Command-line program to remove untagged, unconnected nodes from .osm file
Unfortuately you're wrong here, as it does not keep untagged, but connected nodes. regards Peter Am 14.05.2014 11:16, schrieb Dominik George: Hi, I am trying to find a way to get rid of untagged, unconnected nodes in a .osm file. I know JOSM can do this but I need a command-line application. I've been trying to do it with osmosis and osmfilter without success (pattern *=* not allowed). Any ideas? If your data file is well-formed, a simple grep will do: grep -v node .*/ Cheers, Nik ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Command-line program to remove untagged, unconnected nodes from .osm file
Thanks. I'm starting to think that the task is too ambitious especially if the .osm file is a large one. Regards, Juan On Wednesday, May 14, 2014 11:30 AM, Peter Wendorff wendo...@uni-paderborn.de wrote: Unfortuately you're wrong here, as it does not keep untagged, but connected nodes. regards Peter Am 14.05.2014 11:16, schrieb Dominik George: Hi, I am trying to find a way to get rid of untagged, unconnected nodes in a .osm file. I know JOSM can do this but I need a command-line application. I've been trying to do it with osmosis and osmfilter without success (pattern *=* not allowed). Any ideas? If your data file is well-formed, a simple grep will do: grep -v node .*/ Cheers, Nik ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
[OSM-talk] Re: This has to stop: User Diaries Spam
Having just had over 20 such entries popping up not only on the RSS-feed but also on Twitter this indeed has become very irksome. I suggest first posting of a user with less than 10 edits to be held for moderation. Red tape which I generally am hesitant about but generally people with less than 10 edits don't post until they have some more experience. Google Translate might be of use to decipher postings if moderators don't understand them, just to get the gist of whether it is an ad for printer cartridges or a proper post. --Jói ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
2014-05-14 11:51 GMT+02:00 Pierre Béland pierz...@yahoo.fr: I dont think that we can justify to create a second account. This had the effect over the last few years to slow down some valuable Imports. I do not believe that requiring the creation of a dedicated account for paid mapping will shoo this kind of mappers away. It is done in 2 minutes. Maybe it slows down imports done by a large group of people each with small amounts of data, but if you are going to do a lot of mapping or importing it certainly won't be an obstacle. My guess is that paid mappers will most probably create a dedicated account anyway in order to be able to document the outcome of their work. cheers, Martin ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] This has to stop: User Diaries Spam
On Tue, May 13, 2014 at 11:31 PM, Andreas Goss andi...@t-online.de wrote: BUT something has to be done against the Spam on the User Diaries page. Soon or later, we will have more spam than real diary entries if it's not already the case. Since the devs have no will or resources to struggle against spam, I would suggest to remove the diary feature completely. Pieren ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] This has to stop: User Diaries Spam
Am 14.05.2014 14:46, schrieb Pieren: On Tue, May 13, 2014 at 11:31 PM, Andreas Goss andi...@t-online.de wrote: BUT something has to be done against the Spam on the User Diaries page. Soon or later, we will have more spam than real diary entries if it's not already the case. Since the devs have no will or resources to struggle against spam, I would suggest to remove the diary feature completely. I think it would be better to remove the twitter integration of all user diaries and only use certain official ones like the weekly summmary. Theodin Pieren ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
Kate, You bring up some really excellent questions. Like Frederik, the key that differentiates the organizational mapping is central planning/management. From your email it seems that HOT is involved in various project, some of which would qualify as this organizational mapping, while others would not, and some of it would be meta in the sense that you are organizing organizations. I'd like to keep this discussion focused on how we can address the current issues. We will surely be in a place where it will be necessary to update and clarify the policies in the future. Do you have suggestions that you want to share on addressing the concerns Paul brought up, and if you have concerns about how it may negatively impact your work, where that might be the case and how you would suggest changing the proposal to address it? - Serge ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] OSM France BANO project... openaddresses in France
On Wed, May 14, 2014 at 11:16 AM, Christian Quest cqu...@openstreetmap.fr wrote: OSM-FR has initiated a project to create an open database of addresses in France. Data sources will be OSM + available opendata + address data collected from the cadastre. The resulting dataset will be under ODbL (because of OSM data and also many opendata sets are also under ODbL). Christian, thanks for sharing this news, it looks promising and interesting. -Simone ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Command-line program to remove untagged, unconnected nodes from .osm file
2014-05-14 11:49 GMT+02:00 Juan Lucas Domínguez Rubio juan_lucas...@yahoo.com: Thanks. I'm starting to think that the task is too ambitious especially if the .osm file is a large one. I guess you are aware of other discussions about this topic, e.g. here: https://help.openstreetmap.org/questions/22384/highlight-untagged-and-unconnected-nodes-in-a-way and here (with further pointers to QM tools etc.): http://wiki.openstreetmap.org/wiki/Untagged_unconnected_node unfortunately it doesn't look as if there was a command line tool, and honestly for big files I can't imagine how this could work (you'll have to have some kind of temp. storage like a file or a database to keep track of all the nodes). Walking through all the ways and relations for every single node to check if it is used seems excessive ;-) cheers, Martin ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] This has to stop: User Diaries Spam
On 14/05/2014 5:53 AM, Jóhannes Birgir Jensson wrote: Having just had over 20 such entries popping up not only on the RSS-feed but also on Twitter this indeed has become very irksome. I suggest first posting of a user with less than 10 edits to be held for moderation. Red tape which I generally am hesitant about but generally people with less than 10 edits don't post until they have some more experience. Google Translate might be of use to decipher postings if moderators don't understand them, just to get the gist of whether it is an ad for printer cartridges or a proper post. --Jói Strictly speaking it's not according to IETF rules, but on the lists I help maintain I turn on moderation for originators of spam when I see it but leave it off for everyone else. I don`t know if it`s possible to do that for the diaries. Tom ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
Hi, On 14.05.2014 11:09, Jóhannes Birgir Jensson wrote: The focus needs to be on the problem at hand, The problem at hand is currently at a scale that can still be handled on a case-by-case basis; the reason DWG is thinking about a general guideline is not that we're trigger happy bureaucrats but that we'd like everyone to know the rules of play rather than making them up as we go along, and that we assume that the number of such cases might be on the rise. which I gather is companies hiring people to map things using their own methodology incompatible with current OSM tagging guidelines. No, that is one potential issue but by far not the only aspect. Consider a real-life situation like this: * User complains fictional data is added all over my city! * Investigation finds 10 accounts having added fictional data; * further investigation finds that 10 other accounts have signed up at the same time from the same network, but have added things that do not immediately look bad (things that might or might not be factual) * reaching out to those who edited the most brings zero reply (possibly because their native language is not English nor anything spoken by anyone in DWG) Even reconstructing the whole situation takes quite a bit of time; and then we have to decide which bits to revert and which to keep. Is this a course that was misunderstood, or just organized doodling, or what? How can we reach the teacher (if any)? In this specific instance we decided to revert everything contributed by the whole group - surely not the optimum outcome for an OSM training course! Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09 E008°23'33 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
Maybe someone could remind folks about the ...wasn't there a...did I hear about a test server? Something like a sandbox? A -- Alex On Thu, May 15, 2014 at 1:14 AM, Frederik Ramm frede...@remote.org wrote: Hi, On 14.05.2014 11:09, Jóhannes Birgir Jensson wrote: The focus needs to be on the problem at hand, The problem at hand is currently at a scale that can still be handled on a case-by-case basis; the reason DWG is thinking about a general guideline is not that we're trigger happy bureaucrats but that we'd like everyone to know the rules of play rather than making them up as we go along, and that we assume that the number of such cases might be on the rise. which I gather is companies hiring people to map things using their own methodology incompatible with current OSM tagging guidelines. No, that is one potential issue but by far not the only aspect. Consider a real-life situation like this: * User complains fictional data is added all over my city! * Investigation finds 10 accounts having added fictional data; * further investigation finds that 10 other accounts have signed up at the same time from the same network, but have added things that do not immediately look bad (things that might or might not be factual) * reaching out to those who edited the most brings zero reply (possibly because their native language is not English nor anything spoken by anyone in DWG) Even reconstructing the whole situation takes quite a bit of time; and then we have to decide which bits to revert and which to keep. Is this a course that was misunderstood, or just organized doodling, or what? How can we reach the teacher (if any)? In this specific instance we decided to revert everything contributed by the whole group - surely not the optimum outcome for an OSM training course! Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09 E008°23'33 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] This has to stop: User Diaries Spam
Am 5/14/14 15:59 , schrieb Theodin: I think it would be better to remove the twitter integration of all user diaries and only use certain official ones like the weekly summmary. If you only want weekly summaries you can already just only follow @openstreetmap or for the German one @OSMBlogde. Not sure if there are more. __ openstreetmap.org/user/AndiG88 wiki.openstreetmap.org/wiki/User:AndiG88 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] This has to stop: User Diaries Spam
Am 5/14/14 14:46 , schrieb Pieren: On Tue, May 13, 2014 at 11:31 PM, Andreas Goss andi...@t-online.de wrote: BUT something has to be done against the Spam on the User Diaries page. Soon or later, we will have more spam than real diary entries if it's not already the case. Since the devs have no will or resources to struggle against spam, I would suggest to remove the diary feature completely. Pieren I just think the diaries are too important for that, right now. They are pretty pretty much the only place where you can share something with the community outside of mailing lists. In addition those blog entries can actually be formatted and shared online compared to mailing lists. (e.g. http://www.reddit.com/r/duckduckgo/duplicates/24yqc2/nextduckduckgo_uses_openstreetmap_for_places/) OpenStreetMap does not really have a news page, which gets regular updates and submitting the larger announcements to the official blog to be published will be even more work. Also they are on the openstreetmap.org domain which means is really great from a SEO perspective (compared to private blogs), which unfortunately also is the main reason for all the Spam. __ openstreetmap.org/user/AndiG88 wiki.openstreetmap.org/wiki/User:AndiG88 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] This has to stop: User Diaries Spam
The best solution in my opinion -if possible- would be to limit the access to diaries to accounts that have existed for a certain time and have some edits (just like 1 week and 5-10 Edits) or maybe easier require moderator approval for every account that isn't 1 week or month old. A general approval for the diaries by moderator approval for the first posts only might also be a solution. Who is currently moderating and cleaning up this spam when discovered? Implementing part of the above would be pretty simple. Until some low threshold of participation is reached (by time or edits) diary entries can be written, but not syndicated in the RSS feed (which I guess is what is used to build blogs.openstreetmap.org, and @osmblogs). Just a line or two to change in this function https://github.com/openstreetmap/openstreetmap-website/blob/master/app/controllers/diary_entry_controller.rb#L120 Then make a pull request. Would be up to Tom Hughes to accept the request or not. For more active moderation, we'd need to know, how much time commitment does it really require? And what is the profile of the people doing so? Would need to be someone trusted by OWG. There is a block facility in the app, which prevents edits if applied (until the user reads the block message; there may also be different levels). I'm not sure if access to diary publishing respects the block. If the process starts to require even more active moderating, some kind of dashboard, to highlight suspicious behavior, could be useful. Could start to gather queries of users to investigate further ... like new users, who have made a lot of diary entries in a short span of time. Perhaps someone already has such a thing, and is using it to monitor things? Mikel * Mikel Maron * +14152835207 @mikel s:mikelmaron On Tuesday, May 13, 2014 5:34 PM, Andreas Goss andi...@t-online.de wrote: First of all let me say I'm a huge fan of the diaries, I think it is great that OpenStreetMap gives every user the chance to present his own blog to a larger audiance. BUT something has to be done against the Spam on the User Diaries page. This could be done by moderating faster, but then there is still the problem with Twitter and other feed based systems. Just today I got this: http://i.imgur.com/jKAxzLh.png You know what the first thing is I would do if I wasn't really committed to OpenStreetMap? I would UNFOLLOW that account. I actually had this set up for Facebook, but got rid of it, because it was so annoying. OpenStreetMap already isn't doing that much regarding social media and Marketin/PR, so can we at least do it right with the few channels that exist? The best solution in my opinion -if possible- would be to limit the access to diaries to accounts that have existed for a certain time and have some edits (just like 1 week and 5-10 Edits) or maybe easier require moderator approval for every account that isn't 1 week or month old. A general approval for the diaries by moderator approval for the first posts only might also be a solution. I also wrote 2 diary entries about this some time ago: http://www.openstreetmap.org/user/AndiG88/diary/21188 http://www.openstreetmap.org/user/AndiG88/diary/21156 Andi PS: Then there is also the issue with profile backlink spam and while there is a page on the wiki to report it, but usually very little happens if you do: http://wiki.openstreetmap.org/wiki/Spam __ openstreetmap.org/user/AndiG88 wiki.openstreetmap.org/wiki/User:AndiG88 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
I have to say, my initial reaction to this proposal was that it was heavy handed, unnecessarily punitive, over reaching, and not in the spirit of OSM. A cure worse than the disease. But I think there's something here, which if done right, can benefit all involved. In part it's framing this the right way, and in part the right tools, which offer something to all involved. And the right tool I reckon are OSM Groups (https://github.com/osmlab/openstreetmap-website/tree/groups-sketch). It's wise to contextualize all of this as part of best practice, rather than policy requirements. We want to encourage people how to do things well within OSM, rather than discourage them from getting involved at all. No reason to take a defensive stance, unless an epic problem is erupting. From the DWG and others, would be good to know more specifically about problem instances from the past. Exactly what scale of a problem are we talking about? To summarize what's new here, what the DWG is essential asking for is an association between organized groups and individual user, and details on those organized groups. And in fact, I bet most organized groups would be more than happy happy to represent themselves, and the people involved. Doing so on the wiki is clumsy. Doing so in openstreetmap-website could be very straightforward for all involved. Say I start a mapping project in a slum in Nairobi. I actually want to bring everyone involved together, to communicate among ourselves, to promote our project within the community, to help track and advise what folks are mapping. I want a Group, and I want to integrate it as part of the training process. As a bonus, anyone who has questions about what's going on, can find the association from individual users easily, ask questions of the whole group or group admin, if needed. So let's finish up the groups-sketch, get in deployed, and start using smart tools so its attractive and obvious how to do right by the map! Mikel * Mikel Maron * +14152835207 @mikel s:mikelmaron On Wednesday, May 14, 2014 2:26 PM, Alex Rollin alex.rol...@gmail.com wrote: Maybe someone could remind folks about the ...wasn't there a...did I hear about a test server? Something like a sandbox? A -- Alex On Thu, May 15, 2014 at 1:14 AM, Frederik Ramm frede...@remote.org wrote: Hi, On 14.05.2014 11:09, Jóhannes Birgir Jensson wrote: The focus needs to be on the problem at hand, The problem at hand is currently at a scale that can still be handled on a case-by-case basis; the reason DWG is thinking about a general guideline is not that we're trigger happy bureaucrats but that we'd like everyone to know the rules of play rather than making them up as we go along, and that we assume that the number of such cases might be on the rise. which I gather is companies hiring people to map things using their own methodology incompatible with current OSM tagging guidelines. No, that is one potential issue but by far not the only aspect. Consider a real-life situation like this: * User complains fictional data is added all over my city! * Investigation finds 10 accounts having added fictional data; * further investigation finds that 10 other accounts have signed up at the same time from the same network, but have added things that do not immediately look bad (things that might or might not be factual) * reaching out to those who edited the most brings zero reply (possibly because their native language is not English nor anything spoken by anyone in DWG) Even reconstructing the whole situation takes quite a bit of time; and then we have to decide which bits to revert and which to keep. Is this a course that was misunderstood, or just organized doodling, or what? How can we reach the teacher (if any)? In this specific instance we decided to revert everything contributed by the whole group - surely not the optimum outcome for an OSM training course! Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09 E008°23'33 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
[OSM-talk] This has to stop: User Diaries Spam
Am 14.05.2014 14:46, schrieb Pieren: I would suggest to remove the diary feature completely. In my opinion that would be a big step in the wrong direction. In fact we should be expanding the community side of OSM.org as the community should be at the heart of the project. Things like mailing lists and IRC are simply not known about by average Joe. I'm glad to see the Google Summer of Code project as a good first start: http://www.google-melange.com/gsoc/project/details/google/gsoc2014/ukasiu/5724160613416960 Rob ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] This has to stop: User Diaries Spam
Just to say, @osmblogs is not the diaries, it is just a conversion of the osm aggregate blogs RSS feed, using ifttt.com. If spam were not in the RSS feed it would not be on twitter either. David ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
Hi, On 14.05.2014 23:07, Mikel Maron wrote: I have to say, my initial reaction to this proposal was that it was heavy handed, unnecessarily punitive, over reaching, and not in the spirit of OSM. A cure worse than the disease. Somehow I have the impression that either the proposal was colossally mis-worded, or it as somehow acquired a spin or social dynamic (OMG they're targeting HOT!) that was never intended and is not, in my opinion, in any way present in Paul's request for consultation. Frankly I see nothing punitive or heavy handed here but that may really be a matter of perception. If you're in the HOT business then you might immediately see how this could apply to some of your projects and might make life harder. When I read the proposal, I think of the countless man-hours (and frustration and desperation and heated tempers) involved when mappers on the German forum once again find a strange edit pattern and over the course of days and hundreds of messages the truth slowly emerges. There's nothing punitive here; there's an attempt to make life easier for everyone. It is not about regulating anything - I don't think Paul said anything about anyone enforcing mapping rules or whatever - it is just about transparency and disclosure. If someone teaches OSM to a group of people and instructs them to set up an account - does it really make matters worse if you ask them to write one sentence on their profile page (I am Joe Smith and I am learning OSM in Mikel's OSM for Dummies course)? Would this not be good practice already, even if not expressly written anywhere? We want to encourage people how to do things well within OSM, rather than discourage them from getting involved at all. I should be surprised if the measures outlined by Paul were to affect more than a small franction of mappers. The overwhelming majority of OSM are ordinary mappers who do this as a hobby and who would hardly ever fall under these rules. They would likely not even read or know of them so how would they be discouraged? No reason to take a defensive stance, unless an epic problem is erupting. We're trying to look ahead just a tiny bit. We have seen the problems that Wikipedia had with this; we're already seeing SEO spam in OSM (which would not be helped by this policy but proves that we're not magically exempt from businesses abusing us) - can it hurt to be prepared? To summarize what's new here, what the DWG is essential asking for is an association between organized groups and individual user, and details on those organized groups. Yes, almost; personally I'd say we are concerned about those cases where, essentially, if the community wants someone to change their mapping they would have to speak to that person's supervisor. So it's not so much your average mapping party but really something where the organisation has some kind of command structure and the individual mapper isn't free to map what they want. In my eyes, this means a fundamental change in approach compared to the average mapper who does what the like in their spare time; one that is worth documenting. Say I start a mapping project in a slum in Nairobi. I actually want to bring everyone involved together, to communicate among ourselves, to promote our project within the community, to help track and advise what folks are mapping. I want a Group, and I want to integrate it as part of the training process. As a bonus, anyone who has questions about what's going on, can find the association from individual users easily, ask questions of the whole group or group admin, if needed. Yes, I think these Group ideas could go a long way to help. Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09 E008°23'33 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
Hi Frederik Somehow I have the impression that either the proposal was colossally mis-worded, or it as somehow acquired a spin or social dynamic (OMG they're targeting HOT!) that was never intended and is not, in my opinion, in any way present in Paul's request for consultation. No, didn't think this was intended. Yes perhaps mis-worded. I think as phrased, it's possible unintended consequences that could effect HOT, among other organizations, which haven't had any problems like the ones targeted by the proposed policy. There's nothing punitive here; there's an attempt to make life easier for everyone. It is not about regulating anything - I don't think Paul said anything about anyone enforcing mapping rules or whatever - it is just about transparency and disclosure. That's good. I agree, it's a lot about how it's phrased and communicating. When something is labelled as required, then there's a implication of enforcement. The alternative of a best community practice may be the same thing practically, but sounds pretty different. If someone teaches OSM to a group of people and instructs them to set up an account - does it really make matters worse if you ask them to write one sentence on their profile page (I am Joe Smith and I am learning OSM in Mikel's OSM for Dummies course)? Would this not be good practice already, even if not expressly written anywhere? Definitely a good practice. Is there a way to encourage this better during the sign up process? Or after sign up, to alert users that they have a little more to do to make their profile complete? Like some kind of alert next to the user drop down on the top - right? In my eyes, this means a fundamental change in approach compared to theaverage mapper who does what the like in their spare time; one that is worth documenting Ok, we're probably violently agreeing with each other at this point ;). Agreed. My point is that we should strike the right tone and perception. Yes, I think these Group ideas could go a long way to help. Right on! Let's get groups-sketch finished up and deployed. -Mikel * Mikel Maron * +14152835207 @mikel s:mikelmaron On Wednesday, May 14, 2014 7:03 PM, Frederik Ramm frede...@remote.org wrote: Hi, On 14.05.2014 23:07, Mikel Maron wrote: I have to say, my initial reaction to this proposal was that it was heavy handed, unnecessarily punitive, over reaching, and not in the spirit of OSM. A cure worse than the disease. Somehow I have the impression that either the proposal was colossally mis-worded, or it as somehow acquired a spin or social dynamic (OMG they're targeting HOT!) that was never intended and is not, in my opinion, in any way present in Paul's request for consultation. Frankly I see nothing punitive or heavy handed here but that may really be a matter of perception. If you're in the HOT business then you might immediately see how this could apply to some of your projects and might make life harder. When I read the proposal, I think of the countless man-hours (and frustration and desperation and heated tempers) involved when mappers on the German forum once again find a strange edit pattern and over the course of days and hundreds of messages the truth slowly emerges. There's nothing punitive here; there's an attempt to make life easier for everyone. It is not about regulating anything - I don't think Paul said anything about anyone enforcing mapping rules or whatever - it is just about transparency and disclosure. If someone teaches OSM to a group of people and instructs them to set up an account - does it really make matters worse if you ask them to write one sentence on their profile page (I am Joe Smith and I am learning OSM in Mikel's OSM for Dummies course)? Would this not be good practice already, even if not expressly written anywhere? We want to encourage people how to do things well within OSM, rather than discourage them from getting involved at all. I should be surprised if the measures outlined by Paul were to affect more than a small franction of mappers. The overwhelming majority of OSM are ordinary mappers who do this as a hobby and who would hardly ever fall under these rules. They would likely not even read or know of them so how would they be discouraged? No reason to take a defensive stance, unless an epic problem is erupting. We're trying to look ahead just a tiny bit. We have seen the problems that Wikipedia had with this; we're already seeing SEO spam in OSM (which would not be helped by this policy but proves that we're not magically exempt from businesses abusing us) - can it hurt to be prepared? To summarize what's new here, what the DWG is essential asking for is an association between organized groups and individual user, and details on those organized groups. Yes, almost; personally I'd say we are concerned about those cases where, essentially, if the community wants someone to change their mapping they would have to
Re: [OSM-talk] Organizational mapping policy
On 15 May 2014 01:03, Frederik Ramm frede...@remote.org wrote: If you're in the HOT business then you might immediately see how this could apply to some of your projects and might make life harder. When I read the proposal, I think of the countless man-hours (and frustration and desperation and heated tempers) involved when mappers on the German forum once again find a strange edit pattern and over the course of days and hundreds of messages the truth slowly emerges. There's nothing punitive here; there's an attempt to make life easier for everyone. It is not about regulating anything - I don't think Paul said anything about anyone enforcing mapping rules or whatever - it is just about transparency and disclosure. If someone teaches OSM to a group of people and instructs them to set up an account - does it really make matters worse if you ask them to write one sentence on their profile page (I am Joe Smith and I am learning OSM in Mikel's OSM for Dummies course)? Would this not be good practice already, even if not expressly written anywhere? The case of existing users mapping under the direction of others, or users switching between their own mapping and being directed by someone are probably relatively rare. So the main case being new users coming to OSM as part of a directed group, osm can maybe borrow one or two mechanisms used on other websites: * a field in the signup form where they can type in or select an option about how the arrived in OSM, whose value may only be retrievable by the DWG, * sign-up links with an embedded referer that a workshop teacher can create and mail to students, or if it is possible to implement sufficiently short links then even write them on the blackboard. Cheers ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Organizational mapping policy
From: Mikel Maron [mailto:mikel_ma...@yahoo.com] Sent: Wednesday, May 14, 2014 2:07 PM Subject: Re: [OSM-talk] Organizational mapping policy I have to say, my initial reaction to this proposal was that it was heavy handed, unnecessarily punitive, over reaching, and not in the spirit of OSM. A cure worse than the disease. To clarify (and I could have made this more explicit) there is *not* a proposed policy here. The DWG is considering if it is necessary to issue guidelines, it is not decided that something needs to be issued or the contents of anything we'd issue. The items listed are possible requirements and possible covered activities only. It is extremely unlikely that any policy resulting from this will include all the possible requirements and cover all the possible activities. I'm personally against some of the requirements listed as possibilities. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk-ie] Dublin street names from Dublin City Council
Hi Nick, This would have been a good idea, if you excluded the fact, that some local Irish names might not match between cities, and dialects alone. Also, some English names might have valid translations specific just for Dublin alone (see Smithfield in Dublin, which is named Margadh na Feirme, meaning Farm Market). I'd keep it specific to Dublin, and maybe Leinster province, if we can confirm the names actually get written in the same fashion as they do in Dublin. That said, this only refers to street names alone; I'm not concerned about shop names at the time. Best regards, On 13 May 2014 20:56, Nick Burrett n...@sqrt.co.uk wrote: Would it be better to run it across all of Ireland? The roads in one town should have the same translation as any other town. You could collect the existing translations in OSM to augment the city list, then apply that countrywide. I was thinking of a similar scheme for shops to rationalise discrepancies in say the shop tag given to Boots. On 13 May 2014 17:48, Mike Połtyn hole...@gmail.com wrote: Hi everybody. I am fairly new to OSM and this is my first post here, so I want to introduce myself. Hi, my name is Michał Połtyn; I am a web developer from Poland, currently living in Dublin. My nickname on OSM is Holek. Now, that's good enough, so let's cut to the chase, shall we? I have been adding Irish names to Dublin streets[1] by hand[2] recently[3], and I thought there's a better way. Turns out, Dublin City Council releases a PDF with all city names in English and Irish.[4] Now, I found out that this dataset is also available on dublinked.com[5] as XLS or CSV (the most important for us), where it's licensed as Public data, available to everyone. I would assume this means public domain; in the end it is just a list of properties. My idea was to add all missing Irish names to streets in Dublin in one big changeset, and report all other conflicting names on a Wiki page to review. I thought I could prepare a bot script to add all name:ga tags for all matching streets and run it during Dublin hackaton next Sunday under a supervision of OSM masters :) What are your thoughts, concerns? I'm open for suggestions, critique, and any good word in general. Take care, everybody! -- Michał Hołek Połtyn [1] http://www.openstreetmap.org/changeset/22218893 [2] http://www.openstreetmap.org/changeset/22234506 [3] http://www.openstreetmap.org/changeset/22218453 [4] http://www.dublincity.ie/SiteCollectionDocuments/Sr%C3%A1idainmneacha%20Bhaile%20Atha%20Cliath.pdf (PDF, 357 KB) [5] http://dublinked.com/datastore/datasets/dataset-213.php ___ Talk-ie mailing list Talk-ie@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-ie ___ Talk-ie mailing list Talk-ie@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-ie ___ Talk-ie mailing list Talk-ie@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-ie
[OSM-talk-ie] OpenStreetMap hack day in TOG, Dublin next Sunday 25th May
Hi all, Just a reminder that there's an OpenStreetMap hack day / meet up in TOG, the Dublin hackerspace, next Sunday 25th May, all afternoon (12-6). Everyone (TOG members and non-TOG-members, heavy OSM users, to people just curious about OSM) is welcome. There's no structure to the day, this will be a free form meet up, hack day, bar camp. I'd suggest bringing a laptop or the like if you'd like to do anything. There may be a spare laptop from TOG or two available on the day. We'll probably go for some beers afterwards, so if you're in Dublin city centre that evening, pop along. Some more info: http://www.tog.ie/2014/05/open-street-map-hack-day-2/ TOG has been mapping on OSM for a long time ( http://www.openstreetmap.org/node/475361416 ), and it's at Location: 53.341512, -6.268792. I'm the point-of-contact for this event, and should be there all day. Any questions ask me. TOG phone number: +353 76 6035359. See you all there, Rory ___ Talk-ie mailing list Talk-ie@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-ie
[OSM-talk-ie] Dublin street names from Dublin City Council
Hi, I agree that there are some particular problems with Dublin street names. This document: www.dublincity.ie/SiteCollectionDocuments/Sráidainmneacha Bhaile Atha Cliath.pdf lists Nassau Street = Sráid Nassau. The street sign lists Nassau Street = Sráid Thobair Phádraig. It is down to over-zealous linguistic nationalism: http://www.historyireland.com/volume-13/dublins-street-names/ Years ago, I spoke to one of the council officials who deals with street names and his opinion was the document could be considered 95% correct. My experience since then would lead me to think 90% would be the number. I'm not sure that the PDF matches the CSV. Colm --- Never doubt that a small group of thoughtful, committed citizens can change the world. Indeed, it is the only thing that ever has. Margaret Mead ___ Talk-ie mailing list Talk-ie@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-ie
Re: [OSM-talk-ie] Dublin street names from Dublin City Council
Hi Colm, I'd say even 90% of proper names is good enough to be used on OSM. At least during importing we have a document to reference (being that PDF or CSV, doesn't really matter), and if anything, we can reference that in the edit itself as well. On Wed, May 14, 2014 at 1:34 PM, Colm Moore colmmoor...@hotmail.com wrote: Hi, I agree that there are some particular problems with Dublin street names. This document: www.dublincity.ie/SiteCollectionDocuments/Sráidainmneacha Bhaile Atha Cliath.pdf lists Nassau Street = Sráid Nassau. The street sign lists Nassau Street = Sráid Thobair Phádraig. It is down to over-zealous linguistic nationalism: http://www.historyireland.com/volume-13/dublins-street-names/ Years ago, I spoke to one of the council officials who deals with street names and his opinion was the document could be considered 95% correct. My experience since then would lead me to think 90% would be the number. I'm not sure that the PDF matches the CSV. Colm --- Never doubt that a small group of thoughtful, committed citizens can change the world. Indeed, it is the only thing that ever has. Margaret Mead ___ Talk-ie mailing list Talk-ie@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-ie -- Michał Hołek Połtyn ___ Talk-ie mailing list Talk-ie@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-ie
Re: [Talk-br] Edição destrutiva no Rio de Janeiro
Aproveitando que o Paulo Carvalho é constante na lista ele poderia avaliar se considera o changeset destrutivo. Sim. É definitivamente destrutivo. Estou nos Estados Unidos, a trabalho, acessando o e-mail esporadicamente. Perdemos muita coisa que deu trabalho enorme para fazer!!! Não sei se tem como salvar o changeset num OSM à parte, mas acho péssimo ter meu trabalho perdido assim. Por mim reverte e depois a pessoa vai colocando o que de útil possa ter. Em 13 de maio de 2014 17:11, Arlindo Pereira openstreet...@arlindopereira.com escreveu: Ele me mandou agora esta mensagem. O que vocês acham? Preciso de ajuda para passar um pente fino nesse changeset, não terei tempo para verificar as modificações uma a uma. []s Arlindo Então Arlindo, Realmente, eu percebi que enquanto eu não havia efetuado o upload das minhas alterações o PauloCarvalhoRJ havia submetido um changeset exatamente na região em que eu estava trabalhando e obviamente que eu percebi o bom trabalho que ele estava realizando e não queria degradar o esforço dele (e pra falar a verdade fiquei até meio desesperado pois eu também não queria jogar tudo o que tinha feito fora), mas eu pesquisei e descobri que o JOSM tem uma ferramenta ótima para tratar dos conflitos que permite analisar caso a caso, nó por nó, via por via onde houve uma diferença de versão entre os elementos modificados no arquivo .osm e os mesmos elementos no banco de dados no momento do upload (ou a qualquer momento na verdade). E essa ferramenta permite que em cada caso de conflito (quando a versão do elemento é alterada no banco de dados no momento entre o meu download e o respectivo upload das modificações), por exemplo um nó ou uma via que sofreu uma modificação após eu ter baixado no JOSM e que na minha edição eu deletei, ou o contrário o elemento foi deletado mas eu mantive no meu e fiz alguma modificação mas alguém efetuou um edit e o apagou, o JOSM gera uma lista com todos os conflitos existentes naquele momento e permite que cada caso seja tratado permitindo que se opte tanto pela versão do servidor como pela versão local, isso a nível de atributo do tag e posição geográfica, por exemplo, tiveram muitos casos em que ele havia inserido atributos de nome da rua ou outros mas havia mantido a posição e eu optei por manter o atributo que ele gerou e a posição que eu havia ajustado, dessa forma atendendo a ambos, e a maioria dos casos foram exatamente desse tipo. Provavelmente o que eu apaguei da criação dele é porque eu também já havia criado um elemento com o mesmo objetivo, tanto que a região agora se encontra praticamente completa e só mesmo um morador seria capaz de identificar detalhes que possam ser melhorados em termos de mapeamento. Claro que ainda faltam alguns nomes de ruas e outros atributos ainda podem ser explorados haja visto a infinidade de possibilidades que o OSM nos permite, mas o básico para uma pessoa usar com o objetivo de navegação. Quanto às restrições, eu estou ciente e até sei onde ficavam, todas ali na região do retorno da Av. das Américas em frente ao Barra Shopping e 1 ou 2 cruzamentos depois desse. Essa foi a questão que eu disse que estou comprometido em corrigir iniciando exatamente naquela avenida. Na verdade eu ainda tenho o arquivo, mas deixar de remover os pontos os pontos que eu removi nunca seria capaz de refazer o trabalho feito pois como disse na outra mensagem grande parte das modificações se deram justamente na limpeza de desconformidades apontados pelo validador do JOSM que está de acordo com as regras do Wiki e otimização de excesso de pontos sem atributos principalmente em vias. O que você comentou sobre as vias separadas eu expliquei na mensagem anterior mesmo deixando claro que eu tomei esse cuidado de verificar em cada item se haviam relações associadas. Peço que antes de efetuar a reversão tente exportar uma imagem daquela região como está e como ficará. Reforçando, não há como efetuar otimização e correção de erros sem que haja exclusões. Aproveitando que o Paulo Carvalho é constante na lista ele poderia avaliar se considera o changeset destrutivo. Será bastante lamentável, mas ok, o critério é de vocês. 2014-05-13 15:23 GMT-03:00 Arlindo Pereira openstreet...@arlindopereira.com: Beleza, hoje eu não vou ter como reverter (trabalhando até tarde). Se ninguém mais o fizer, agradeço se vc botar a mão na massa. =) []s Arlindo 2014-05-13 15:12 GMT-03:00 Fernando Trebien fernando.treb...@gmail.com: Cara, eu tenho condições de reverter hoje lá pelas 18h (tô sem a infraestrutura necessária aqui), salvando uma cópia do estado atual (anterior à reversão), mas se alguém quiser se adiantar, dou todo o meu apoio. Eu começaria aplicando o Reverter do JOSM (vai levar muito tempo pra baixar todos os milhares de objetos modificados) e dando um Ctrl+Z para desfazer a reversão. Isso deve dar os objetos modificados no seu estado imediatamente após a edição.
Re: [Talk-br] Edição destrutiva no Rio de Janeiro
Pouco, Frenando? Detonaram o que eu fiz na região de Interlagos... Onde fica o respeito pelo trabalho dos outros que muito se fala aqui? Esses 20% de que você fala nada mais são do que HORAS do meu trabalho. E mais, meus changesets são pequenos, onde cada coisa é explicada em pormenores. Como tanta coisa foi apagada assim??? Em 14 de maio de 2014 14:33, Fernando Trebien fernando.treb...@gmail.comescreveu: Sim Paulo, está tudo salvo, o antes e o depois da edição. Eu gostaria de convencer o Thiago a consertar os problemas manualmente ele mesmo. Aparentemente são poucos os problemas (apesar de significativos). Seria um bom aprendizado para ele, para evitar os mesmos erros no futuro. (Thiago, você assume a bronca?) Obviamente, quando decidi que iríamos consertar e não reverter (depois de avaliar as principais seções do resultado do osmdiff), também assumi que, se ele mesmo não fizer a correção, eu mesmo farei. Eu decidi isso porque porque uns 80% do changeset não é destrutivo. São os 20% que sobraram que precisam ser consertados. Obviamente não podemos fazer essas análises longas o tempo todo. Se a coisa ficar impossível de gerenciar, a única coisa que poderemos fazer é adotar a potencialmente injusta prática de reverter tudo imediatamente e pedir que seja refeito, quebrando em changesets facilmente revisáveis e bem descritas. Ou seja, entendam que esse caso é a exceção, não a regra. Se eu não tivesse me prontificado a revisar (tenho interesse nisso porque estou construindo uma ferramenta justamente pra facilitar a revisão), e o Arlindo não pudesse, e ninguém mais pudesse, eu apoiaria a reversão imediata. O JOSM dá muito poder ao mapeador. E com grande poder, grandes responsabilidades. 2014-05-14 14:25 GMT-03:00 Paulo Carvalho paulo.r.m.carva...@gmail.com: Aproveitando que o Paulo Carvalho é constante na lista ele poderia avaliar se considera o changeset destrutivo. Sim. É definitivamente destrutivo. Estou nos Estados Unidos, a trabalho, acessando o e-mail esporadicamente. Perdemos muita coisa que deu trabalho enorme para fazer!!! Não sei se tem como salvar o changeset num OSM à parte, mas acho péssimo ter meu trabalho perdido assim. Por mim reverte e depois a pessoa vai colocando o que de útil possa ter. Em 13 de maio de 2014 17:11, Arlindo Pereira openstreet...@arlindopereira.com escreveu: Ele me mandou agora esta mensagem. O que vocês acham? Preciso de ajuda para passar um pente fino nesse changeset, não terei tempo para verificar as modificações uma a uma. []s Arlindo Então Arlindo, Realmente, eu percebi que enquanto eu não havia efetuado o upload das minhas alterações o PauloCarvalhoRJ havia submetido um changeset exatamente na região em que eu estava trabalhando e obviamente que eu percebi o bom trabalho que ele estava realizando e não queria degradar o esforço dele (e pra falar a verdade fiquei até meio desesperado pois eu também não queria jogar tudo o que tinha feito fora), mas eu pesquisei e descobri que o JOSM tem uma ferramenta ótima para tratar dos conflitos que permite analisar caso a caso, nó por nó, via por via onde houve uma diferença de versão entre os elementos modificados no arquivo .osm e os mesmos elementos no banco de dados no momento do upload (ou a qualquer momento na verdade). E essa ferramenta permite que em cada caso de conflito (quando a versão do elemento é alterada no banco de dados no momento entre o meu download e o respectivo upload das modificações), por exemplo um nó ou uma via que sofreu uma modificação após eu ter baixado no JOSM e que na minha edição eu deletei, ou o contrário o elemento foi deletado mas eu mantive no meu e fiz alguma modificação mas alguém efetuou um edit e o apagou, o JOSM gera uma lista com todos os conflitos existentes naquele momento e permite que cada caso seja tratado permitindo que se opte tanto pela versão do servidor como pela versão local, isso a nível de atributo do tag e posição geográfica, por exemplo, tiveram muitos casos em que ele havia inserido atributos de nome da rua ou outros mas havia mantido a posição e eu optei por manter o atributo que ele gerou e a posição que eu havia ajustado, dessa forma atendendo a ambos, e a maioria dos casos foram exatamente desse tipo. Provavelmente o que eu apaguei da criação dele é porque eu também já havia criado um elemento com o mesmo objetivo, tanto que a região agora se encontra praticamente completa e só mesmo um morador seria capaz de identificar detalhes que possam ser melhorados em termos de mapeamento. Claro que ainda faltam alguns nomes de ruas e outros atributos ainda podem ser explorados haja visto a infinidade de possibilidades que o OSM nos permite, mas o básico para uma pessoa usar com o objetivo de navegação. Quanto às restrições, eu estou ciente e até sei onde ficavam, todas ali na região do retorno da Av. das Américas em frente ao Barra Shopping e 1 ou 2 cruzamentos depois desse. Essa foi
Re: [Talk-br] Maratona de Edições da Copa
Resposta da moça que mapeou o de Manaus, seria bom descobrir quem é esse chefe: Olá Fernando, Agradeço suas observações, posso corrigir alguns erros sem problemas, mas como desenhei a pedido do meu chefe, e ele pediu para eu mapear o arco, isso eu não vou alterar. O que não lhe impede de editar da maneira que achar melhor. Como não sou usuária do Open a muito tempo e basicamente faço o que meu chefe pede pra eu fazer, gostaria de saber o que é canal Irc? Obrigada. Em 13 de maio de 2014 21:36, Fernando Trebien m-431753-d6b...@messages.openstreetmap.org escreveu: Olá Susan Campos, Fernando Trebien http://www.openstreetmap.org/user/Fernando%20Trebienenviou uma mensagem pelo OpenStreetMap para você com o assunto Mapa da Arena do Amazonas em Manaus: == Olá Susan, Estamos nos preparando para uma mapatona e nos deparamos com a sua obra de arte no estádio da Arena do Amazonas em Manaus ( https://lists.openstreetmap.org/pipermail/talk-br/2014-May/007572.html). Nós não costumamos mapear os arcos, nem as linhas do meio de campo. Isso é chamado mapear para o renderizador, ou mapear de uma forma a produzir um desenho específico. Ao invés disso, o correto é mapear logicamente, pensando no significado real dos elementos geométricos. Convido você a participar da discussão na lista ( http://wiki.openstreetmap.org/wiki/Pt-br:Contact) e a remapear o estádio, usando como exemplos os seguintes: - http://www.openstreetmap.org/#map=18/-22.91214/-43.23005 - http://www.openstreetmap.org/#map=18/-23.54522/-46.47375 - http://www.openstreetmap.org/#map=18/-30.06555/-51.23592 Se você tiver dúvidas, fique à vontade para perguntar à comunidade tanto na lista quanto no fórum. == Você pode também ler a mensagem em http://www.openstreetmap.org/message/read/431753 e pode responder em http://www.openstreetmap.org/message/reply/431753 -- *Susan Campos da Silva * *Cel: (92) 9239-5269* 2014-05-13 22:20 GMT-03:00 Fernando Trebien fernando.treb...@gmail.com: Natal certamente não é grave, já que o mapeamento dos telhados até poderia ser feito com building=roof. Se deveriam ser segmentos separados... provavelmente é discutível, já que literalmente não há teto. É um daqueles casos no limite entre o preciso demais e o preciso de menos, onde a decisão depende mais do fator utilidade. Eu acho que fica um pouco melhor sem separar, mas com footways representando os acessos às arquibancadas. (Se discordarem, aceito.) Já o de Manaus... alguém precisa avisar a autora que, apesar de bonitinho, não deve mapear para o renderizador. :P Vou mandar uma mensagem pra ela. 2014-05-13 21:49 GMT-03:00 Bráulio brauliobeze...@gmail.com: É, o mapeamento da Arena das Dunas está errado mesmo. Eu fiz esse mapeamento meio tosco enquanto ainda estava em construção, pra não ficar o vazio. O estádio foi inaugurado, mas ainda há muitas obras em torno dele. É necessário dar uma passada lá para mapear direitinho, principalmente os acessos. Moro a 1km de lá, mas não sei quando terei tempo :) . 2014-05-13 21:40 GMT-03:00 Wille wi...@wille.blog.br: Esse da Amazônia está todo errado. A pessoa desenhou até a linha de meio campo... A Arena das Dunas tem uma área de construção em torno dele que talvez deva ser apagada, visto que o estádio já foi inaugurado. Alguém aí de Natal pode confirmar? On 13-05-2014 21:25, Fernando Trebien wrote: Que curiosos esses estádios, especialmente o primeiro: * http://www.openstreetmap.org/#map=19/-3.08321/-60.02801 * http://www.openstreetmap.org/#map=18/-5.82665/-35.21170 2014-05-13 18:06 GMT-03:00 Alexandre Magno Brito de Medeiros alexandre@gmail.com: 2014-05-13 11:44 GMT-03:00 Wille wi...@wille.blog.br: [...] Criei a página no Wiki: http://wiki.openstreetmap.org/wiki/Mapathon_Copa_2014 A resposta (acima) para a pergunta do Lucas (abaixo). Em 13 de maio de 2014 17:13, Lucas Ferreira Mation lucasmat...@gmail.com escreveu: Já tem uma Wiki sobre a Mapatona rolando??? 2014-05-13 13:33 GMT-03:00 Wille wi...@wille.blog.br: Pode ser Mapatona sim! Coloquei a lista de cidades sede com link para o mapa da cidade e para o estádio no wiki. ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br -- Fernando Trebien +55 (51) 9962-5409 Nullius in verba. -- Fernando Trebien +55 (51) 9962-5409 Nullius in verba. ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
Re: [Talk-br] Maratona de Edições da Copa
Que curioso! :D 2014-05-14 20:02 GMT-03:00 Fernando Trebien fernando.treb...@gmail.com: Resposta da moça que mapeou o de Manaus, seria bom descobrir quem é esse chefe: Olá Fernando, Agradeço suas observações, posso corrigir alguns erros sem problemas, mas como desenhei a pedido do meu chefe, e ele pediu para eu mapear o arco, isso eu não vou alterar. O que não lhe impede de editar da maneira que achar melhor. Como não sou usuária do Open a muito tempo e basicamente faço o que meu chefe pede pra eu fazer, gostaria de saber o que é canal Irc? Obrigada. Em 13 de maio de 2014 21:36, Fernando Trebien m-431753-d6b...@messages.openstreetmap.org escreveu: Olá Susan Campos, Fernando Trebien http://www.openstreetmap.org/user/Fernando%20Trebienenviou uma mensagem pelo OpenStreetMap para você com o assunto Mapa da Arena do Amazonas em Manaus: == Olá Susan, Estamos nos preparando para uma mapatona e nos deparamos com a sua obra de arte no estádio da Arena do Amazonas em Manaus ( https://lists.openstreetmap.org/pipermail/talk-br/2014-May/007572.html). Nós não costumamos mapear os arcos, nem as linhas do meio de campo. Isso é chamado mapear para o renderizador, ou mapear de uma forma a produzir um desenho específico. Ao invés disso, o correto é mapear logicamente, pensando no significado real dos elementos geométricos. Convido você a participar da discussão na lista ( http://wiki.openstreetmap.org/wiki/Pt-br:Contact) e a remapear o estádio, usando como exemplos os seguintes: - http://www.openstreetmap.org/#map=18/-22.91214/-43.23005 - http://www.openstreetmap.org/#map=18/-23.54522/-46.47375 - http://www.openstreetmap.org/#map=18/-30.06555/-51.23592 Se você tiver dúvidas, fique à vontade para perguntar à comunidade tanto na lista quanto no fórum. == Você pode também ler a mensagem em http://www.openstreetmap.org/message/read/431753 e pode responder em http://www.openstreetmap.org/message/reply/431753 -- *Susan Campos da Silva * *Cel: (92) 9239-5269* 2014-05-13 22:20 GMT-03:00 Fernando Trebien fernando.treb...@gmail.com: Natal certamente não é grave, já que o mapeamento dos telhados até poderia ser feito com building=roof. Se deveriam ser segmentos separados... provavelmente é discutível, já que literalmente não há teto. É um daqueles casos no limite entre o preciso demais e o preciso de menos, onde a decisão depende mais do fator utilidade. Eu acho que fica um pouco melhor sem separar, mas com footways representando os acessos às arquibancadas. (Se discordarem, aceito.) Já o de Manaus... alguém precisa avisar a autora que, apesar de bonitinho, não deve mapear para o renderizador. :P Vou mandar uma mensagem pra ela. 2014-05-13 21:49 GMT-03:00 Bráulio brauliobeze...@gmail.com: É, o mapeamento da Arena das Dunas está errado mesmo. Eu fiz esse mapeamento meio tosco enquanto ainda estava em construção, pra não ficar o vazio. O estádio foi inaugurado, mas ainda há muitas obras em torno dele. É necessário dar uma passada lá para mapear direitinho, principalmente os acessos. Moro a 1km de lá, mas não sei quando terei tempo :) . 2014-05-13 21:40 GMT-03:00 Wille wi...@wille.blog.br: Esse da Amazônia está todo errado. A pessoa desenhou até a linha de meio campo... A Arena das Dunas tem uma área de construção em torno dele que talvez deva ser apagada, visto que o estádio já foi inaugurado. Alguém aí de Natal pode confirmar? On 13-05-2014 21:25, Fernando Trebien wrote: Que curiosos esses estádios, especialmente o primeiro: * http://www.openstreetmap.org/#map=19/-3.08321/-60.02801 * http://www.openstreetmap.org/#map=18/-5.82665/-35.21170 2014-05-13 18:06 GMT-03:00 Alexandre Magno Brito de Medeiros alexandre@gmail.com: 2014-05-13 11:44 GMT-03:00 Wille wi...@wille.blog.br: [...] Criei a página no Wiki: http://wiki.openstreetmap.org/wiki/Mapathon_Copa_2014 A resposta (acima) para a pergunta do Lucas (abaixo). Em 13 de maio de 2014 17:13, Lucas Ferreira Mation lucasmat...@gmail.com escreveu: Já tem uma Wiki sobre a Mapatona rolando??? 2014-05-13 13:33 GMT-03:00 Wille wi...@wille.blog.br: Pode ser Mapatona sim! Coloquei a lista de cidades sede com link para o mapa da cidade e para o estádio no wiki. ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br -- Fernando Trebien +55 (51) 9962-5409 Nullius in verba. -- Fernando Trebien +55 (51) 9962-5409 Nullius in verba.
Re: [Talk-br] Maratona de Edições da Copa
Provavelmente é um político ou alguém envolvido nas obras da Copa. Peçamos o contato dele ou da empresa. A abordagem tem de ser na linha de: sua empresa não pode interferir assim em nossos processos; contudo, pode ser firmada uma parceria onde nós ensinamos o que é permitido em nosso projeto e vocês contribuem com ele se quiserem. Alexandre Magno Em 14 de maio de 2014 20:02, Fernando Trebien fernando.treb...@gmail.comescreveu: Resposta da moça que mapeou o de Manaus, seria bom descobrir quem é esse chefe: Olá Fernando, Agradeço suas observações, posso corrigir alguns erros sem problemas, mas como desenhei a pedido do meu chefe, e ele pediu para eu mapear o arco, isso eu não vou alterar. O que não lhe impede de editar da maneira que achar melhor. Como não sou usuária do Open a muito tempo e basicamente faço o que meu chefe pede pra eu fazer, gostaria de saber o que é canal Irc? Obrigada. ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
Re: [Talk-br] Maratona de Edições da Copa
Eu não disse o contrário. Em 14 de maio de 2014 20:51, Arlindo Pereira openstreet...@arlindopereira.com escreveu: Acho que o que foi feito foi de boa intenção. ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
[Talk-br] Fwd: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM
Pessoal, Fui autorizado pelo Joaquim a encaminhar a mensagem a seguir para vocês. Por favor, leiam-na. Talvez nem todos estejam acompanhando a discussão no fórum, onde o Genulho alegou pro naoliv numa conversa privada que eram levantamentos dele mesmo. http://forum.openstreetmap.org/viewtopic.php?pid=420323#p420323 Então, devemos solicitar ao DWG a reversão de todos os changesets do Genulpho e no mínimo o bloqueio do usuário dele, talvez até mesmo bani-lo completamente por estar mentindo. Mas antes de continuar queria ouvir a opinião de vocês. -- Forwarded message -- From: Joaquim Francisco joaquim.cunha.franci...@gmail.com Date: 2014-05-13 21:31 GMT-03:00 Subject: ENC: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM To: fernando.treb...@gmail.com Caro Fernando, Chegou ao nosso conhecimento a mensagem abaixo na qual é solicitado esclarecimento ao Genulpho quanto à origem dos dados de Itabira-MG. No final de março, quando ele revelou que havia exportado dados para o OSM, já havíamos feito contato para obter explicações e na ocasião ele disse ter feito apenas de 2 pequenos municípios que tinha desenvolvido desde o início. Entretanto, agora ficou claro que ele faltou com a verdade, pois o mapa de Itabira, além de não ser um dos pequenos municípios que ele citou, há anos é desenvolvido pelo nosso colaborador Ademar Pires Guedes. Em Março, o Genulpho solicitou o desenvolvimento temporariamente ao Ademar, aproveitou a oportunidade para se apropriar indevidamente dos dados e fazer a exportação para o OSM. Em função destes fatos acabamos de bani-lo definitivamente do Tracksource e estamos lhe comunicando para que tome as medidas que julgar cabíveis no caso do OSM. Pedimos desculpas pelo ocorrido e aproveitamos a oportunidade para externar a nossa profunda admiração pelo trabalho que você e toda a equipe do OSM tem feito. Nos colocamos a sua disposição para o que precisar aqui no Tracksource. Abraços, *Chicão* *Equipe de Coordenação e Compilação do Tracksource* On Monday, May 12, 2014 6:15:46 PM UTC-3, Fernando Trebien wrote: Oi pessoal, Eu recomendo que vocês participem mais da lista talk-br e do fórum do OSM ao fazer essas importações. Olha o tamanho da discussão que causou (sem que nenhum de vocês visse/se envolvesse): http://forum.openstreetmap.org/viewtopic.php?id=25409 Quando os dados atendem os critérios legais observados pelo OSM (no mundo inteiro, não é só no Brasil), eles são muito-bem vindos, não importa que seja usando como base a tecnologia do TrackSource e o conversor do Pauo. Mas os colaboradores do OSM, preocupados com vandalismo, precisam ter certeza de que os critérios são bem entendidos por quem está importanto (especialmente se é alguém que não vem se envolvendo na lista e no fórum) e que estão sendo seguidos. Por pouco não começamos a fazer a reversão. E pelo que entendi (agora que li esta thread), não faremos, mas pra ter certeza gostaria de pedir que respondessem à mensagem (que mandei em privado para o Genulpho e depois acabei copiando pro fórum), pra ficar registrado que os dados são de uma origem juridicamente permitida. Outra coisa é que, ao se envolver com a comunidade do OSM, sempre podemos ajudar a tirar dúvidas sobre como fazer a importação da melhor forma, como mesclar os dados, etc. O que mais nos alarmou foi a forma de proceder, que sobrepôs dois conjuntos de dados, por vários dias. Se alguém tentar usar o mapa nessa região nesse instante, vai sair muito perplexo e vai achar que o OSM não presta. Além disso, toda vez que alguém nos expõe dúvidas, a gente acaba aprendendo sobre algo que falta na nossa documentação. Isso pode nos ajudar a elaborar guias e tutoriais melhores, e com isso, a evitar essas confusões no futuro. Se existe algo impedindo que vocês se comuniquem direto com a gente, por favor, nos digam. Estamos lá pra fazer amigos e fazer um bom mapa, juntos. On Friday, March 7, 2014 5:27:02 PM UTC-3, Paulo Carvalho wrote: Sim, mapas de sua própria autoria feitos a partir de fontes como IBGE, imagens do Bing, tracks de GPS, etc. são muito bem vindos ao OSM. Obrigado pelas contribuições, pois assim melhoramos o mapa no Brasil. Em 7 de março de 2014 16:13, Genulpho Mendonça Caldas genu...@gmail.com escreveu: Prezados Colegas Pegando carona nestas explicações, acabo de transferir 2 mapas que desenho para o OSM, completos, com todas as estradas rurais. Em 7 de março de 2014 12:39, Engº Agrº Daniel de Góes Nepomuceno danielne...@gmail.com escreveu: ALVISSARAS!!! Agora deu certo. Relevem minha ignorância... Atenciosamente, [image: Imagem inline 1] Em 7 de março de 2014 11:04, Erick de Oliveira Leal erickdeol...@gmail.com escreveu: É... Minha instrução pecou ai. Hehehe Em 07/03/2014 10:07, Paulo Carvalho paulo.r.m...@gmail.com escreveu: *java -jar ts2pfm.jar nomeCompletoDoMapa.gtm* Em 7 de março de 2014 10:02, Engº Agrº Daniel de Góes Nepomuceno danielne...@gmail.com escreveu: Bom dia
Re: [Talk-br] Fwd: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM
2014-05-15 0:22 GMT-03:00 Fernando Trebien fernando.treb...@gmail.com: Então, devemos solicitar ao DWG a reversão de todos os changesets do Genulpho e no mínimo o bloqueio do usuário dele, talvez até mesmo bani-lo completamente por estar mentindo. Mas antes de continuar queria ouvir a opinião de vocês. Reverter os seus changesets sim. Acho que não fica dúvidas para ninguém sobre isso. Quanto a banir, não sei se dá para chegar a este ponto (o DWG que decide). Acredito que o usuário vendo que as ações foram incorretas, passa a não mais fazer isso (ou mesmo que ocorram reincidências, existe uma progressão no nível do aviso/bloqueio). ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
Re: [Talk-br] Fwd: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM
O Trackmaker é um programa Free, tem tb a versão paga.. -Se ele coletou os dados, a obra seria dele... - O não usar o GTM no OSM, o que ele deve ter feito foi importado os GPX, o GPX até no JOSM pode se trabalhar com ele.. Por isso essa questão, Quem é dono de quem é muito vaga. Tanto lá como aqui, cada um PRODUZ e CONTRIBUI... Aqui tem que analisar a Legalidade com a Moralidade... ___ Anor C. A. de Souza Concórdia SC 49-8808-4963 Date: Thu, 15 May 2014 00:37:10 -0300 From: nao...@gmail.com To: talk-br@openstreetmap.org Subject: Re: [Talk-br] Fwd: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM 2014-05-15 0:22 GMT-03:00 Fernando Trebien fernando.treb...@gmail.com: Então, devemos solicitar ao DWG a reversão de todos os changesets do Genulpho e no mínimo o bloqueio do usuário dele, talvez até mesmo bani-lo completamente por estar mentindo. Mas antes de continuar queria ouvir a opinião de vocês. Reverter os seus changesets sim. Acho que não fica dúvidas para ninguém sobre isso. Quanto a banir, não sei se dá para chegar a este ponto (o DWG que decide). Acredito que o usuário vendo que as ações foram incorretas, passa a não mais fazer isso (ou mesmo que ocorram reincidências, existe uma progressão no nível do aviso/bloqueio). ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
Re: [Talk-br] Fwd: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM
Se a lista lá é privada, autorização teria de partir de cada um que participara da thread. Em 15 de maio de 2014 00:22, Fernando Trebien fernando.treb...@gmail.comescreveu: Fui autorizado pelo Joaquim a encaminhar a mensagem a seguir para vocês. Por favor, leiam-na. ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
Re: [Talk-br] Fwd: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM
2014-05-15 0:49 GMT-03:00 A. Carlos anorcar...@hotmail.com: -Se ele coletou os dados, a obra seria dele... Se ele usou o Bing, como ele mesmo afirmou, já não está dentro da legalidade exigida para o OSM. ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
Re: [Talk-br] Fwd: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM
2014-05-15 0:48 GMT-03:00 Alexandre Magno Brito de Medeiros alexandre@gmail.com: Se a lista lá é privada, autorização teria de partir de cada um que participara da thread. Não, a lista é aberta: https://groups.google.com/forum/#!topic/alt-tracksource/9jWCPTi10LU A mensagem que o Fernando repassou que foi enviada apenas a ele (e não para a lista deles). ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
Re: [Talk-br] Fwd: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM
Em 15 de maio de 2014 00:52, Nelson A. de Oliveira nao...@gmail.comescreveu: 2014-05-15 0:48 GMT-03:00 Alexandre Magno Brito de Medeiros alexandre@gmail.com: Se a lista lá é privada, autorização teria de partir de cada um que participara da thread. Não, a lista é aberta: https://groups.google.com/forum/#!topic/alt-tracksource/9jWCPTi10LU A mensagem que o Fernando repassou que foi enviada apenas a ele (e não para a lista deles). Embaixo veio toda uma thread. Mas OK, a lista é aberta. ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
Re: [Talk-br] Fwd: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM
2014-05-15 0:55 GMT-03:00 Alexandre Magno Brito de Medeiros alexandre@gmail.com: Embaixo veio toda uma thread. Mas OK, a lista é aberta. É a thread da lista :-) ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
Re: [Talk-br] Fwd: [ADM] Fwd: [alt-tracksource] Re: arquivo GTM no OSM
Sim, eu entendi. Se a lista fosse privada... seria o que eu falei. 2014-05-15 0:58 GMT-03:00 Nelson A. de Oliveira nao...@gmail.com: 2014-05-15 0:55 GMT-03:00 Alexandre Magno Brito de Medeiros alexandre@gmail.com: Embaixo veio toda uma thread. Mas OK, a lista é aberta. É a thread da lista :-) ___ Talk-br mailing list Talk-br@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-br
[Talk-de] Android APP notes ?
Hi, ich suche nach eine einfach APP die mir nahe/nächstgelegene Notes zeigt und am besten gleich noch Mapfactor Navigator oder ähnliches mit Zielkoordinates füttert. Damit könnte man bei schönem Wetter einfach mal eine Rundtour über die Notes machen. Flo -- Florian Lohoff f...@zz.de signature.asc Description: Digital signature ___ Talk-de mailing list Talk-de@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-de
Re: [Talk-de] Android APP notes ?
Florian Lohoff schrieb am 14.05.2014 13:32: ich suche nach eine einfach APP die mir nahe/nächstgelegene Notes zeigt und am besten gleich noch Mapfactor Navigator oder ähnliches mit Zielkoordinates füttert. OsmAnd kann die Notes anzeigen. Du musst jedoch selber eine Stelle neben dem Note als Ziel auswählen und dahin navigieren (nur osmand routing [offline, online] routing möglich. Damit könnte man bei schönem Wetter einfach mal eine Rundtour über die Notes machen. Also hier in Aachen will ich die Notes vorher per Browser einzeln angucken, ob eine vor Ortprüfung überhaupt nötig ist. hier eröffnet im Herbst ein Mexikaner will ich jetzt noch nicht vor Ort prüfen. -- Grüße Holger Jeromin ___ Talk-de mailing list Talk-de@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-de
Re: [Talk-de] Android APP notes ?
Hi, On Wed, May 14, 2014 at 01:52:40PM +0200, Holger Jeromin wrote: Florian Lohoff schrieb am 14.05.2014 13:32: ich suche nach eine einfach APP die mir nahe/nächstgelegene Notes zeigt und am besten gleich noch Mapfactor Navigator oder ähnliches mit Zielkoordinates füttert. OsmAnd kann die Notes anzeigen. Du musst jedoch selber eine Stelle neben dem Note als Ziel auswählen und dahin navigieren (nur osmand routing [offline, online] routing möglich. Habe ich gefunden - Ich finde ja OSMAnd nicht so schön - aber gut - es scheint zu funktionieren - Hab auch gleich eine Issue aufgemacht das wenn man die Note anklickt nicht gleich sagen kann Dahinrouten sondern immer daneben auf die Hintergrundkarte klicken muss. Damit könnte man bei schönem Wetter einfach mal eine Rundtour über die Notes machen. Also hier in Aachen will ich die Notes vorher per Browser einzeln angucken, ob eine vor Ortprüfung überhaupt nötig ist. hier eröffnet im Herbst ein Mexikaner will ich jetzt noch nicht vor Ort prüfen. Wenn es Winter ist schon ;) OsmAnd zeigt ja auch den content an :) Flo -- Florian Lohoff f...@zz.de signature.asc Description: Digital signature ___ Talk-de mailing list Talk-de@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-de
Re: [Talk-de] Android APP notes ?
Hi In Locus kann man auch das herunterladen von Notes in der Nähe aktivieren. Man kann neue erstellen und auch kommentierenschließen hab ich nicht gefunden. Was jedoch sehr gut geht ist die Übergabe an andere Kartenanwendungen (Mapfactor...). mfg Christian Am 14. Mai 2014 17:38 schrieb Florian Lohoff f...@zz.de: Hi, On Wed, May 14, 2014 at 01:52:40PM +0200, Holger Jeromin wrote: Florian Lohoff schrieb am 14.05.2014 13:32: ich suche nach eine einfach APP die mir nahe/nächstgelegene Notes zeigt und am besten gleich noch Mapfactor Navigator oder ähnliches mit Zielkoordinates füttert. OsmAnd kann die Notes anzeigen. Du musst jedoch selber eine Stelle neben dem Note als Ziel auswählen und dahin navigieren (nur osmand routing [offline, online] routing möglich. Habe ich gefunden - Ich finde ja OSMAnd nicht so schön - aber gut - es scheint zu funktionieren - Hab auch gleich eine Issue aufgemacht das wenn man die Note anklickt nicht gleich sagen kann Dahinrouten sondern immer daneben auf die Hintergrundkarte klicken muss. Damit könnte man bei schönem Wetter einfach mal eine Rundtour über die Notes machen. Also hier in Aachen will ich die Notes vorher per Browser einzeln angucken, ob eine vor Ortprüfung überhaupt nötig ist. hier eröffnet im Herbst ein Mexikaner will ich jetzt noch nicht vor Ort prüfen. Wenn es Winter ist schon ;) OsmAnd zeigt ja auch den content an :) Flo -- Florian Lohoff f...@zz.de -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.10 (GNU/Linux) iQIVAwUBU3OODJDdQSDLCfIvAQijXxAAj6XqjRYqS5ar2dVlQxWgIcRnbNIKpEiM jQh2n16QaAetVHnkIBck/noebG+/wupXhVgtVNEgWoGiDR4QBGfHjaRE+oEuW07W n+cLLMxaOK/GKenCyblIIJo5AHS7Cz3OTQA1fGfiwMCnb/4Utw1KmG6U1YPPNQxz jQLjZAIlen7w/v4h8aB3CRmlziJfd/Wfw4DD295a7I4QykMN/I+9L0lKk4IkhVps /NjeAu3EfxzMdkPUM9Sk3FqpeMuPJSI3FQxRboTuPoyo5jRBFFMAPgxPlTx//7OJ Z+K7Yvts3M8Eufa5nHi2lzw5h+I6kqotPxl1BbhhvtYZYYtPdRx1KTDYRshEPtva XnSWzNa4KHWmpGFcGmW3mLssQk61fGfafVN8+IRTUEWZQsUeVRtDOhd/V/MQdmUb vB/lBlwbwtbtwp5yApjurX2mJZvsvd3gA9dAmJ/jRg85EOV//FUv0VYpYFvsMJFA 4bU9xIQzHS75xRvtvLEKMVZ2Cjpa0IVDpPqAr8vILoRCZU2LVOA4J9LvWjKmVx7q A/IPsPxThcyfccTiP3JE1mtJkCR4WTL0GRoKZdIzJHflDifsTSXJDg8WSOqn1x1P vixtw+W2+kV7Z4I5OL7Aogl0Pnfq75ZXcA2uqpK+ZRmXnxB98SJ9ZjKOlEZ9drIX VxhpWsrCaig= =KXA/ -END PGP SIGNATURE- ___ Talk-de mailing list Talk-de@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-de ___ Talk-de mailing list Talk-de@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-de
[Talk-in] Planning Architecture students contributing to OSM
http://hydlab.in/blog/interns/planning-architecture-students-contributing-to-osm/ -Satya Satyaakam.net | fossevents.in | fossacademy.org ___ Talk-in mailing list Talk-in@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-in
[Talk-it] Fwd: [OSM-talk] OSM France BANO project... openaddresses in France
anche in Francia...openaddress. -- Forwarded message -- From: Christian Quest cqu...@openstreetmap.fr Date: Wed, May 14, 2014 at 11:16 AM Subject: [OSM-talk] OSM France BANO project... openaddresses in France To: OpenStreetMap t...@openstreetmap.org OSM-FR has initiated a project to create an open database of addresses in France. Data sources will be OSM + available opendata + address data collected from the cadastre. The resulting dataset will be under ODbL (because of OSM data and also many opendata sets are also under ODbL). In the past weeks, we've been coding the collector scripts which are also matching non OSM data with OSM data, to get better street names (Avenue des Champs-Élysées instead of AV DES CHAMPS ELYSEES). We've starting collecting data a few days ago and the progress can be seen here: http://openstreetmap.fr/outils/bano/status A BANO overlay rendering is also available to view the coverage and density: http://tile.openstreetmap.fr/~cquest/leaflet/bano.html#6/46.988/0.912 green = OSM data orange = opendata blue = cadastre + OSM enhancements red = cadastre only The resulting dataset will not be imported as is in OSM as the french community considers it needs to be manually reviewed. As a first step, we prefer to concentrate on making sure we reduce the number of missing streets and street names and this is simplified by the automatic matching that is done between OSM and non-OSM data. When no match is found, we make it visible on the BANO rendered layer at the higher zoom levels, and showing the expected name and street code. Here is an example: http://tile.openstreetmap.fr/~cquest/leaflet/bano.html#17/48.82902/2.31073 We're about to publish the first datasets, one for each departement (roughtly 100.000 to 1.000.000 addresses depending on the area) which be of course listed on openaddresses.io We expect to have a coverage of 80%-90% (roughtly 20 million addresses). For more info, check the wiki page (in French only, translations welcome): The BANO team is also reachable on IRC #bano (irc://irc.oftc.net:6667/bano) and on twitter https://twitter.com/ProjetBANO -- Christian Quest - OpenStreetMap France ___ talk mailing list t...@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk -- -S ___ Talk-it mailing list Talk-it@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-it
Re: [Talk-it] openstreetmapper card
dopo praticamente due settimane (ultimo voto inserito è stato il 4/5/2014) chiudo ufficialmente le votazioni per il background del retro del biglietto da visita. Sorprendentemente ha vinto la proposta iniziale, infatti con 5 voti a favore (pari al 33%) il vecchio background è la proposta che ha raccolto più voti seguito da Oceani grigio chiari sfumati con 3 voti (pari al 20%) i voti totali sono stati 15 e come sempre io non ho espresso il voto. provvedo subito ad aggiornare l'immagine presente sul wiki. Da oggi OSM italia ha un biglietto da visita ufficiale. - Ciao, Aury -- View this message in context: http://gis.19327.n5.nabble.com/openstreetmapper-card-tp5795902p5806264.html Sent from the Italy General mailing list archive at Nabble.com. ___ Talk-it mailing list Talk-it@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-it
[Talk-it] OSM su Marcopolo TV
Ciao *, oggi mi sono imbattuto nel programma Diario di viaggio - Catalogna nel canale televisivo a tema viaggi del DTT denominato Marco Polo e ho visto, con molto piacere, che il programma mostrava un tour per tutta Barcellona e come mappa mostrava la nostra cara mappa OSM con stile Standard. Non ho potuto verificare se fossero in regola con la licenza d'uso, ma sicuramente è stato molto soddisfacente, anche perché la mappa in quell'area è estremamente dettagliata. Se riesco a recuperare il video lo posto. Buon mapping :-) -- E' assurdo impiegare gli uomini di intelligenza eccellente per fare calcoli che potrebbero essere affidati a chiunque se si usassero delle macchine Gottfried Wilhelm von Leibnitz, Filosofo e Matematico (1646-1716) Internet è la più grande biblioteca del mondo. Ma il problema è che i libri sono tutti sparsi sul pavimento John Allen Paulos, Matematico (1945-vivente) Luca 'remix_tj' Lorenzetto, http://www.remixtj.net , lorenzetto.l...@gmail.com ___ Talk-it mailing list Talk-it@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-it
Re: [Talk-it] Traduzione di Mapgive
Ciao, alcune note: * ho modificato il footer del sito per aggiornarlo rispetto alla traduzione e per eliminare i riferimenti alla privacy policy, al FOIA (Freedom of Information Act) ed alle info di copyright del Dipartimento di Stato USA. Ho invece aggiunto un link al sito originale (http://mapgive.state.gov) e spostato tutte le info in una (estremamente prolissa) pagina separata: http://mapgive.openstreetmap.it/ulteriori-informazioni/ * Ho aggiunto delle info sul progetto Mapgive originale, su FBK e sulla comunità italiana di OSM in questa pagina: http://mapgive.openstreetmap.it/a-proposito-open-mapping/ * Ho aggiunto anche delle immagini, con le quali però ho qualche problema di visualizzazione, in particolare non si vedono alcune immagini nelle seguenti pagine: ** http://mapgive.openstreetmap.it/perche-mappare/ ** http://mapgive.openstreetmap.it/ulteriori-informazioni/ ** http://mapgive.openstreetmap.it/a-proposito-open-mapping/ Qualcuno ha qualche idea sul perché? Grazie. C ___ Talk-it mailing list Talk-it@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-it
[Talk-it] [mapgive] Storie italiane per mapgive
Ciao a tutti, (cambio thread per chiarezza) Chi di voi ha guardato la pagina storie su Mapgive: http://mapgive.openstreetmap.it/storie/ avrà potuto apprezzare quanto questi esempi siano molto chiari. Tuttavia pensavo che sarebbe ancora meglìo aggiungere (non sostituire!) anche degli esempi relativi all'Italia e sempre legati all'idea che le mappe siano utili durante le emergenze ma non solo. Per questo mi venivano in mente subito: * SardSOS, e la mappatura HOT della Sardegna * Matera, (alla voce empowering communities) Qualcuno di voi vuole darmi una mano a scrivere una pagina per queste storie d'uso di OpenStreetMap? Avete altri esempi? Ciao, Cristian ___ Talk-it mailing list Talk-it@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-it
Re: [Talk-it] Traduzione di Mapgive
quali sono le immagini che non vedi? Il giorno 14/mag/2014, alle ore 21:31, Cristian Consonni kikkocrist...@gmail.com ha scritto: * Ho aggiunto anche delle immagini, con le quali però ho qualche problema di visualizzazione, in particolare non si vedono alcune immagini nelle seguenti pagine: ** http://mapgive.openstreetmap.it/perche-mappare/ ** http://mapgive.openstreetmap.it/ulteriori-informazioni/ ** http://mapgive.openstreetmap.it/a-proposito-open-mapping/ Qualcuno ha qualche idea sul perché? Francesco Piero Paolicelli TW: @piersoft STORE: GooglePlay/AppStore WWW: www.apposta.biz Sorry for typos, sent by mobile. ___ Talk-it mailing list Talk-it@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-it
Re: [Talk-it] Traduzione di Mapgive
Il 14 maggio 2014 21:50, Francesco Piero Paolicelli pierso...@gmail.com ha scritto: quali sono le immagini che non vedi? * http://mapgive.openstreetmap.it/perche-mappare/ Tutte e tre le icone, confronta: http://mapgive.openstreetmap.it/perche-mappare/ con: http://mapgive.state.gov/why-map/ * http://mapgive.openstreetmap.it/ulteriori-informazioni/ tutte le immagini della pagina: https://github.com/osmItalia/MapgiveIT/blob/master/_assets/img/src-images/cc-zero.png https://raw.githubusercontent.com/osmItalia/MapgiveIT/master/_assets/img/src-images/kml_document.svg * http://mapgive.openstreetmap.it/a-proposito-open-mapping/ Il logo FBK ed logo OSM-IT. Ciao, C ___ Talk-it mailing list Talk-it@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-it
Re: [Talk-se] Registrering av medgivande för släppt material
Ursäkta det sena svaret. Vi löste preliminärt medgivandefrågan genom att arkivera mailet internt. På så sätt finns åtminstone ett diarienr som man kan referera till även externt. Ett konto på openstreetmap.se skulle dock kunna vara rätt bra för att ge ökad tillgång till datan (dvs. all hjälp/användning är uppskattad). Lämpligt kontonamn är då lokal-profil för att hålla ihop det med OSM/OHM användarnamn. Förhoppningsvis får jag även tillgång till fler fria dataset i framtiden. Den som är intresserad av att nysta i denna GIS-skatt alternativt ge mig tips om hur man går till väga för att få in den i OHM skulle verkligen vara en ängel. /André André Costa | GLAM-tekniker, Wikimedia Sverige | andre.co...@wikimedia.se | +46 (0)733-964574 Stöd fri kunskap, bli medlem i Wikimedia Sverige. Läs mer på blimedlem.wikimedia.se 2014-04-25 14:27 GMT+02:00 Karl Wettin karl.wet...@kodapan.se: On 24 Apr 2014, at 15:43, André Costa andre.co...@wikimedia.se wrote: Får ursäkta att detta egentligen rör OHM men tänkte att processen lär vara samma som på OSM. Jag fick nyligen ut data från Riksarkivet över alla historiska gränser för kommuner/län/socknar etc. tillsammans med ett medgivande om att detta var ok att använda materialet under en ODbL-licens. Vad jag nu funderar på är hur/var jag kan registrera att så är fallet? Datan ligger vanligtvis inte ute publikt. Liknande måste ha skett tidigare med något av de många dataset som inkluderas i OSM. Det närmsta vi har tror jag är http://wiki.openstreetmap.org/wiki/WikiProject_Sweden/Myndighetskontakter. Om du vill så kan vi fixa ett konto åt dig på openstreetmap.se, där kan du få lägga ut datamängderna under en url som blir något åt hållet http://andre.users.openstreetmap.se/riskarkivet/. kalle ___ Talk-se mailing list Talk-se@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-se ___ Talk-se mailing list Talk-se@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-se
[Talk-se] OSM/Wikipub
Om en vecka är det åter wikipub i Stockholm http://sv.wikipedia.org/wiki/Wikipedia:WT#Wikipub.2C_Stockholm.2C_onsdag_21_majwikipub Jag tänkte delta, wikipedianer är ett trevligt gäng att dricka sockerdricka med. Vore kul om även andra aktiva (eller nybörjare) inom OSM hittade dit för att filosofera om allt mellan himmel och jord. /Markus ___ Talk-se mailing list Talk-se@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-se
Re: [Talk-es] Cerca de Arnedo (era Nuestros Toponimos de Parajes han sido elegidos como lo Peor de OSM)
2014-05-13 11:12 GMT+02:00 Roberto Pla p...@aire.org: Entiendo que hay muchas cosas criticables, pero, ¿Cual es el motivo de critica en esta parte del mapa (Arnedo)? http://www.openstreetmap.org/?lat=42.22779lon=-2.10184zoom=17 Hola, echando un vistazo a ese mapa, he visto que aquí [1], cerca de Las Tres Tetas :P, está el pico Raposal con una altura de 707.664978027344 metros, y con fuente el IDEE. ¿Sabéis si esta altura está bien? Me refiero, ¿el IDEE da estos datos de altura, o puede ser que algún script de importación o algo haya hecho algo raro con el redondeo? Me parece mucha precisión, pero soy lego en estos temas... [1] http://osm.org/go/b~iBJ0LB Muchas gracias, Un saludo, -- Jonás Andradas ___ Talk-es mailing list Talk-es@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-es
Re: [Talk-es] Cerca de Arnedo (era Nuestros Toponimos de Parajes han sido elegidos como lo Peor de OSM)
2014-05-14 10:02 GMT+02:00 Jonas Andradas j.andra...@gmail.com: 2014-05-13 11:12 GMT+02:00 Roberto Pla p...@aire.org: Entiendo que hay muchas cosas criticables, pero, ¿Cual es el motivo de critica en esta parte del mapa (Arnedo)? http://www.openstreetmap.org/?lat=42.22779lon=-2.10184zoom=17 Hola, echando un vistazo a ese mapa, he visto que aquí [1], cerca de Las Tres Tetas :P, está el pico Raposal con una altura de 707.664978027344 metros, y con fuente el IDEE. ¿Sabéis si esta altura está bien? Me refiero, ¿el IDEE da estos datos de altura, o puede ser que algún script de importación o algo haya hecho algo raro con el redondeo? Me parece mucha precisión, pero soy lego en estos temas... [1] http://osm.org/go/b~iBJ0LB Muchas gracias, Un saludo, Se me había olvidado añadir el enlace el nodo: http://www.openstreetmap.org/node/32968746 -- Jonás Andradas ___ Talk-es mailing list Talk-es@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-es
Re: [Talk-es] Cerca de Arnedo (era Nuestros Toponimos de Parajes han sido elegidos como lo Peor de OSM)
Hola. Esa altura no se refiere a la del pico, sino a un vértice geodésico que debe haber ahí. Ese nodo procede de una importación de vértices geodésicos de la Red Regentehttp://www.ign.es/ign/layoutIn/actividadesGeodesiaRedgd.do(que hizo hace años Iván) e igual que él hay 1029 repartidos por toda España. Evidentemente a tal precisión no se llega y menos en altura, no se de donde debe proceder el error, el tema de las alturas y los programas o sistemas con los que los hayan calculados pueden generar esas disparidades de precisiones que no son tales. En la página del IGN existe un buscador de estos vértices[1]; poniendo Raposal obtienes un pdf con las características del mismo ftp://ftp.geodesia.ign.es/Red_Geodesica/Hoja0243/024334.pdf. Ahí si está la altura redondeada al milímetro.. y con eso es mas que suficiente : [1] http://www.ign.es/ign/layoutIn/geodesiaVertices.do Saludos El 14 de mayo de 2014, 10:04, Jonas Andradas j.andra...@gmail.comescribió: 2014-05-14 10:02 GMT+02:00 Jonas Andradas j.andra...@gmail.com: 2014-05-13 11:12 GMT+02:00 Roberto Pla p...@aire.org: Entiendo que hay muchas cosas criticables, pero, ¿Cual es el motivo de critica en esta parte del mapa (Arnedo)? http://www.openstreetmap.org/?lat=42.22779lon=-2.10184zoom=17 Hola, echando un vistazo a ese mapa, he visto que aquí [1], cerca de Las Tres Tetas :P, está el pico Raposal con una altura de 707.664978027344 metros, y con fuente el IDEE. ¿Sabéis si esta altura está bien? Me refiero, ¿el IDEE da estos datos de altura, o puede ser que algún script de importación o algo haya hecho algo raro con el redondeo? Me parece mucha precisión, pero soy lego en estos temas... [1] http://osm.org/go/b~iBJ0LB Muchas gracias, Un saludo, Se me había olvidado añadir el enlace el nodo: http://www.openstreetmap.org/node/32968746 -- Jonás Andradas ___ Talk-es mailing list Talk-es@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-es ___ Talk-es mailing list Talk-es@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-es
[Talk-at] JOSM: route zu route_master hinzufügen
Ich hab's zwar schon einmal gemacht/geschafft, aber das war wohl eher Zufall. Zumindest hab ich keine Ahnung mehr, wie das geht. Also, hier ist meine Frage: Wie fügt man im JOSM eine route zu einem route_master hinzu? LG, Christian ___ Talk-at mailing list Talk-at@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-at
Re: [Talk-at] JOSM: route zu route_master hinzufügen
Hallo! On 14/05/14 08:03, Christian Aigner wrote: Wie fügt man im JOSM eine route zu einem route_master hinzu? route auswählen, auf der rechten Seite in der Relationsliste die route_master rechts anklicken und Add selection to 1 relation auswählen (ganz unten). Alternativ die route_master im Relationseditor öffnen (in Relationsliste den Bearbeitungsbutton wählen), dann die route auswählen und mit dem Links-Pfeil im Dialog der Relation hinzufügen und dann den Relationseditor mit OK schließen. Grüße Simon ___ Talk-at mailing list Talk-at@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-at
Re: [Talk-at] JOSM: route zu route_master hinzufügen
Am 14.05.2014 08:22, schrieb Simon Legner: Hallo! On 14/05/14 08:03, Christian Aigner wrote: Wie fügt man im JOSM eine route zu einem route_master hinzu? route auswählen, auf der rechten Seite in der Relationsliste die route_master rechts anklicken und Add selection to 1 relation auswählen (ganz unten). Alternativ die route_master im Relationseditor öffnen (in Relationsliste den Bearbeitungsbutton wählen), dann die route auswählen und mit dem Links-Pfeil im Dialog der Relation hinzufügen und dann den Relationseditor mit OK schließen. Grüße Simon Danke Simon, hat geklappt. :-) Auf welcher Wiki-Seite könnte ich das verewigen? Ich nehme an, vor diesem Problem stehen auch andere. LG, Christian ___ Talk-at mailing list Talk-at@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-at
[Talk-at] NÖ-Öffi-Seite im Wiki aktualisiert
Da ich gerade eine Buslinie eingezeichnet habe, hab ich auch die NÖ-Öffi-Seite im Wiki aktualisiert: http://wiki.openstreetmap.org/wiki/Lower_Austria_Public_Transport Da ist noch einiges zu tun. ;-) LG, Christian ___ Talk-at mailing list Talk-at@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-at
Re: [Talk-at] JOSM: route zu route_master hinzufügen
Hallo! On 14/05/14 08:37, Christian Aigner wrote: Auf welcher Wiki-Seite könnte ich das verewigen? Relevante Wiki-Seiten sind wohl: http://wiki.openstreetmap.org/wiki/JOSM/Advanced_editing#Add_new_members_to_a_relation http://josm.openstreetmap.de/wiki/Help/Dialog/RelationEditor http://josm.openstreetmap.de/wiki/Help/Dialog/RelationList Die unteren beiden sind wohl nicht mehr ganz am aktuellen Stand ;-) Grüße Simon ___ Talk-at mailing list Talk-at@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-at
Re: [Talk-cz] Hlášení chyb RUIAN přímo z JOSM
On Tue 13-05-14 21:05:19, Petr Vejsada wrote: V OSM tyto dvojité adresy buď byly a nebo nebyly, ale vždy u nich byla poznámka 'Nekonzistence KM a DB' či tak nějak podobně. Neznám přesně historii těch importů do OSM a tak se k tomu nemohu přesněji vyjádřit. K Nekonzistence KM a DB historická poznámka na okraj. Takto označená jsou adresní místa, která byla nalezena v katastrální mapě, avšak chyběly doplňující údaje v UIR-ADR. Poloha byla tedy známa, přesná adresa však nikoli. Pak jsou adresní místa v UIR-ADR, která neměla svůj protějšek v katastrální mapě. Tedy adresa známa, ale místo ne. Z pochopitelných důvodů tyto body v OSM chybí. Například na Praze východ je takových záznamů 9880, Mladá Boleslav 2498 nebo Česká Lípa 3403. Pak můžeme narazit na body, které mají tag fixme. Typicky u adres, kde je špatně přiřazena městská část, případně i ulice. V ideálním případě přispivatel, který import prováděl, tyto výskyty ošetřil, než výsledek nahrál na server. Ze však zkušenosti vím, že se tak ne vždy stalo. Libor ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz
Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže?
Mně to přijde taky nekompletní, asi by to chtělo chtělo sehnat ten CD-ROM, o kterém autor ve své práci píše, že je přílohou. :-) Proto jsem chtěl zkusit tu cestu přes GDAL. Udělal jsem bodove_pole.txt - https://drive.google.com/file/d/0B5J67rBdf34NbVhtWTJHN1hiWUE/edit?usp=sharing - souřadnice bodů TB, ZhB a CZEPOS v S-JTSK a WGS84 Chybu transformačního klíče lze vizualizovat tak, že se vezmou ty S-JTSK souřadnice, natransformují se v PostGISu pomocí +towgs84=570.8,85.7,462.8,4.998,1.587,5.261,3.56 do WGS84 jako jeden bod úsečky a ty WGS84 souřadnice jako druhý konec úsečky. Nebo si lze vyhrát s GRASSem: http://freegis.fsv.cvut.cz/gwiki/S-JTSK_/_Chyba_p%C5%99i_transformaci_z_WGS84_do_S-JTSK MK Petr Vejsada píše v diskusním příspěvku news:1703919.yriTVijfGK@mrnous... Dobrá, tak jsem se mírně znemožnil ;-). Počítat potřebujeme ten vlastní grid. Vstupem je databáze bodů a ta se bude interpolovat. Kromě Postgisu tedy potřebujeme ještě R a PL/R, k tomu možná něco z CRANu. Zacházet s R celkem zvládám, i když v jiné oblasti než jsou geo-úlohy. Zdá se mi, že k pochopení, dostatečnému ke zprovoznění úlohy, mi chybí porozumění všem parametrům funkce dataz_tps_raster. Máme tam: CREATE OR REPLACE FUNCTION dataz_tps_raster ( tab character varying, col character varying, numx integer, numy integer, srid integer, zoom numeric ) RETURNS raster AS $$ ... parametry: tab - jméno tabulky s body col - jméno sloupce, jakého? numx - nevím - používá se jako parametr st_makeemptyraster numy - nevím - dtto srid - čeho SRID? nemohu nalézt tyto funkce: - dataz_create_vector - dataz_return_value struktura tabulky tab: - the_geom - asi geometrie bodu, kterého? Křovákova nebo WGS84? - 'col' - ??? Je to někomu jasnější? -- Petr Dne Út 13. května 2014 22:49:17, Petr Vejsada napsal(a): Zdravím, tato problematika mi není úplně jasná, tak se třeba zeptám blbě - co je na tom k počítání? IMO jde o to, tu databázi, co máte k dispozici, převést do zdrojové formy pro nad2bin, pustit na to nad2bin a pak už jen upravit definici +proj ... a od té doby to Postgis bude umět, ne? Netuším formát té zdrojové formy ani nevím, jak vypadá ta databáze bodů. Je někde k dispozici k nahlédnutí či sosnutí? -- Petr Dne Út 13. května 2014 21:56:26, Martin Kokes napsal(a): Mám celou databázi polohového bodového pole TB, ZhB a CZEPOS v ETRS89(ETRF2000) S-JTSK, nechtěl by mi někdo, kdo se víc orientuje v GDAL (Jáchym Čepický?) pomoci vytvoři nový nadgrid ntv2? http://lists.maptools.org/pipermail/proj/2013-January/006539.html Docela by nám to všem pomohlo v souvislosti s propojením s RÚIAN, protože přiznejme si, sedmiprvková transformace není v pohraničí to pravé ořechové. viz. http://freegis.fsv.cvut.cz/gwiki/S-JTSK_/_Grid http://www.kma.zcu.cz/main.php?KMAfile=./STRUCTURE/05_ebooks/04_Zaverecne_ pr ace/zav_prace.phpDRC=./STRUCTURE/05_ebooks/04_Zaverecne_prace/DRL=CZDR OF= 0osCislo=52920 Psal jsem si na to téma s p. Ježkem, zkoušel jsem kontaktovat p. Chlupa, ale bez výsledku. Jsem na to ochotný dát volnou výpočetní kapacitu na postgisu na Xeon E5 3.6 GHz s SSD. MK ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz
Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže?
Mám celou databázi polohového bodového pole TB, ZhB a CZEPOS v ETRS89(ETRF2000) S-JTSK, nechtěl by mi někdo, kdo se víc orientuje v GDAL (Jáchym Čepický?) pomoci vytvoři nový nadgrid ntv2? *** ten grid je už spočtený, jen je třeba ho převést do správného formátu pro GDAL. http://www.cuzk.cz/Zememerictvi/Geodeticke-zaklady-na-uzemi-CR/GNSS/Nova-realizace-systemu-ETRS89-v-CR.aspx ha hanoj ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz
Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže?
Mno, děkuji pěkně za důvěru. GDAL sice obsluhovat umím (nebo jsem si to do teď myslel), ale pro tuhle oblast aplikace si tedy nejsem jistý ... Na GDAL je po největší zvíře (i když tak nevypadá) Martin Landa - a kromě toho ví, o čem mluvíš :-) Honza Ježek pokud vím se v tom problému fakt vyzná, i když je spíš přes Javu (což je detail), ale počítám, že teď má celkem plno rodinných a pracovních povinností. Můžete mi zkusit když tak po lopatě, jako někomu kdo sice držel Theo 20 v ruce a chápe z rychlíku problém transformace mezi souř. systémy vysvětlit, co potřebujete a proč nestačí standardní 7prvká transformace (alias +towgs84) a k čemu je tenhle grid vlasně dobre do OSM? Jestli by to bylo moc vysvětlování, tak se omlouvám, ..., holt jsem studoval špatnou školu dík J On Wed, May 14, 2014 at 09:32:15AM +0200, hanoj wrote: Mám celou databázi polohového bodového pole TB, ZhB a CZEPOS v ETRS89(ETRF2000) S-JTSK, nechtěl by mi někdo, kdo se víc orientuje v GDAL (Jáchym Čepický?) pomoci vytvoři nový nadgrid ntv2? *** ten grid je už spočtený, jen je třeba ho převést do správného formátu pro GDAL. http://www.cuzk.cz/Zememerictvi/Geodeticke-zaklady-na-uzemi-CR/GNSS/Nova-realizace-systemu-ETRS89-v-CR.aspx ha hanoj ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz -- Jachym Cepicky URL: http://geosense.cz e-mail: jachym.cepicky at geosense cz PGP: http://les-ejk.cz/pgp/JachymCepicky.pgp @jachymc ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz
Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže?
Velice stručně - dotransformace pomocí gridu se dělá kvůli tomu, že S-JTSK není homogenní a má lokální deformace (dané už historicky od jeho vzniku). Jednoznačný matematický vztah je mezi WSG84 a S-JTSK/05 a grid právě vyjadřuje rozdíly mezi hladkým S-JTSK/05 a lokálně deformovaným S-JTSK. Takže jeho použitím se zvýší přesnost transformace a lze používat globální transformační klíč pro celou republiku s přesností, vyhovující katastru. J. Veselý -- Původní zpráva -- Od: Jachym Cepicky jachym.cepi...@geosense.cz Komu: OpenStreetMap Czech Republic talk-cz@openstreetmap.org Datum: 14. 5. 2014 10:56:12 Předmět: Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže? Mno, děkuji pěkně za důvěru. GDAL sice obsluhovat umím (nebo jsem si to do teď myslel), ale pro tuhle oblast aplikace si tedy nejsem jistý ... Na GDAL je po největší zvíře (i když tak nevypadá) Martin Landa - a kromě toho ví, o čem mluvíš :-) Honza Ježek pokud vím se v tom problému fakt vyzná, i když je spíš přes Javu (což je detail), ale počítám, že teď má celkem plno rodinných a pracovních povinností. Můžete mi zkusit když tak po lopatě, jako někomu kdo sice držel Theo 20 v ruce a chápe z rychlíku problém transformace mezi souř. systémy vysvětlit, co potřebujete a proč nestačí standardní 7prvká transformace (alias +towgs84) a k čemu je tenhle grid vlasně dobre do OSM? Jestli by to bylo moc vysvětlování, tak se omlouvám, ..., holt jsem studoval špatnou školu dík J On Wed, May 14, 2014 at 09:32:15AM +0200, hanoj wrote: Mám celou databázi polohového bodového pole TB, ZhB a CZEPOS v ETRS89(ETRF2000) S-JTSK, nechtěl by mi někdo, kdo se víc orientuje v GDAL (Jáchym Čepický?) pomoci vytvoři nový nadgrid ntv2? *** ten grid je už spočtený, jen je třeba ho převést do správného formátu pro GDAL. http://www.cuzk.cz/Zememerictvi/Geodeticke-zaklady-na-uzemi-CR/GNSS/Nova- realizace-systemu-ETRS89-v-CR.aspx ha hanoj ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz -- Jachym Cepicky URL: http://geosense.cz e-mail: jachym.cepicky at geosense cz PGP: http://les-ejk.cz/pgp/JachymCepicky.pgp @jachymc ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz;___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz
Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže?
Nebudu předstírat, že tomu rozumím. Jediné co vím je, že u nás v Beskydech to způsobuje docela výrazný posun RUAN vs. KM. Půl metru je už docela dost. Viz tato mapka: http://freegis.fsv.cvut.cz/gwiki/S-JTSK_/_Chyba_p%C5%99i_ transformaci_z_WGS84_do_S-JTSK Marián -- Původní zpráva -- Od: JV j@seznam.cz Komu: OpenStreetMap Czech Republic talk-cz@openstreetmap.org Datum: 14. 5. 2014 11:19:34 Předmět: Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže? Velice stručně - dotransformace pomocí gridu se dělá kvůli tomu, že S-JTSK není homogenní a má lokální deformace (dané už historicky od jeho vzniku). Jednoznačný matematický vztah je mezi WSG84 a S-JTSK/05 a grid právě vyjadřuje rozdíly mezi hladkým S-JTSK/05 a lokálně deformovaným S-JTSK. Takže jeho použitím se zvýší přesnost transformace a lze používat globální transformační klíč pro celou republiku s přesností, vyhovující katastru. J. Veselý -- Původní zpráva -- Od: Jachym Cepicky jachym.cepi...@geosense.cz Komu: OpenStreetMap Czech Republic talk-cz@openstreetmap.org Datum: 14. 5. 2014 10:56:12 Předmět: Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže? Mno, děkuji pěkně za důvěru. GDAL sice obsluhovat umím (nebo jsem si to do teď myslel), ale pro tuhle oblast aplikace si tedy nejsem jistý ... Na GDAL je po největší zvíře (i když tak nevypadá) Martin Landa - a kromě toho ví, o čem mluvíš :-) Honza Ježek pokud vím se v tom problému fakt vyzná, i když je spíš přes Javu (což je detail), ale počítám, že teď má celkem plno rodinných a pracovních povinností. Můžete mi zkusit když tak po lopatě, jako někomu kdo sice držel Theo 20 v ruce a chápe z rychlíku problém transformace mezi souř. systémy vysvětlit, co potřebujete a proč nestačí standardní 7prvká transformace (alias +towgs84) a k čemu je tenhle grid vlasně dobre do OSM? Jestli by to bylo moc vysvětlování, tak se omlouvám, ..., holt jsem studoval špatnou školu dík J On Wed, May 14, 2014 at 09:32:15AM +0200, hanoj wrote: Mám celou databázi polohového bodového pole TB, ZhB a CZEPOS v ETRS89(ETRF2000) S-JTSK, nechtěl by mi někdo, kdo se víc orientuje v GDAL (Jáchym Čepický?) pomoci vytvoři nový nadgrid ntv2? *** ten grid je už spočtený, jen je třeba ho převést do správného formátu pro GDAL. http://www.cuzk.cz/Zememerictvi/Geodeticke-zaklady-na-uzemi-CR/GNSS/Nova- realizace-systemu-ETRS89-v-CR.aspx ha hanoj ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz -- Jachym Cepicky URL: http://geosense.cz e-mail: jachym.cepicky at geosense cz PGP: http://les-ejk.cz/pgp/JachymCepicky.pgp @jachymc ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz; ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz
Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže?
Jo, to chápu. A co chceme od GDAL? Nažrat nějaký soubor z www.cuzk.cz/Zememerictvi/Geodeticke-zaklady-na-uzemi-CR/GNSS/Nova-realizace-systemu-ETRS89-v-CR.aspx ? to není problém, jako co to chcete? GeoTIFF? Nebo do GRASSu? Nebo jako nějaký vektor? A jaký soubor případně J On Wed, May 14, 2014 at 11:31:01AM +0200, Marián Kyral wrote: Nebudu předstírat, že tomu rozumím. Jediné co vím je, že u nás v Beskydech to způsobuje docela výrazný posun RUAN vs. KM. Půl metru je už docela dost. Viz tato mapka: http://freegis.fsv.cvut.cz/gwiki/S-JTSK_/_Chyba_p%C5%99i_transformaci_z_WGS84_do_S-JTSK Marián -- Původní zpráva -- Od: JV j@seznam.cz Komu: OpenStreetMap Czech Republic talk-cz@openstreetmap.org Datum: 14. 5. 2014 11:19:34 Předmět: Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže? Velice stručně - dotransformace pomocí gridu se dělá kvůli tomu, že S-JTSK není homogenní a má lokální deformace (dané už historicky od jeho vzniku). Jednoznačný matematický vztah je mezi WSG84 a S-JTSK/05 a grid právě vyjadřuje rozdíly mezi hladkým S-JTSK/05 a lokálně deformovaným S-JTSK. Takže jeho použitím se zvýší přesnost transformace a lze používat globální transformační klíč pro celou republiku s přesností, vyhovující katastru. J. Veselý -- Původní zpráva -- Od: Jachym Cepicky jachym.cepi...@geosense.cz Komu: OpenStreetMap Czech Republic talk-cz@openstreetmap.org Datum: 14. 5. 2014 10:56:12 Předmět: Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže? Mno, děkuji pěkně za důvěru. GDAL sice obsluhovat umím (nebo jsem si to do teď myslel), ale pro tuhle oblast aplikace si tedy nejsem jistý ... Na GDAL je po největší zvíře (i když tak nevypadá) Martin Landa - a kromě toho ví, o čem mluvíš :-) Honza Ježek pokud vím se v tom problému fakt vyzná, i když je spíš přes Javu (což je detail), ale počítám, že teď má celkem plno rodinných a pracovních povinností. Můžete mi zkusit když tak po lopatě, jako někomu kdo sice držel Theo 20 v ruce a chápe z rychlíku problém transformace mezi souř. systémy vysvětlit, co potřebujete a proč nestačí standardní 7prvká transformace (alias +towgs84) a k čemu je tenhle grid vlasně dobre do OSM? Jestli by to bylo moc vysvětlování, tak se omlouvám, ..., holt jsem studoval špatnou školu dík J On Wed, May 14, 2014 at 09:32:15AM +0200, hanoj wrote: Mám celou databázi polohového bodového pole TB, ZhB a CZEPOS v ETRS89(ETRF2000) S-JTSK, nechtěl by mi někdo, kdo se víc orientuje v GDAL (Jáchym Čepický?) pomoci vytvoři nový nadgrid ntv2? *** ten grid je už spočtený, jen je třeba ho převést do správného formátu pro GDAL. http://www.cuzk.cz/Zememerictvi/Geodeticke-zaklady-na-uzemi-CR/GNSS/Nova-realizace-systemu-ETRS89-v-CR.aspx ha hanoj ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz -- Jachym Cepicky URL: http://geosense.cz e-mail: jachym.cepicky at geosense cz PGP: http://les-ejk.cz/pgp/JachymCepicky.pgp @jachymc ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz -- Jachym Cepicky URL: http://les-ejk.cz e-mail: jachym.cepicky at gmail com PGP: http://les-ejk.cz/pgp/JachymCepicky.pgp @jachymc ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz
Re: [Talk-cz] Nový S-JTSK grid, kdo pomůže?
Zdravim, Dne 14. května 2014 10:55 Jachym Cepicky jachym.cepi...@geosense.cz napsal(a): Na GDAL je po největší zvíře (i když tak nevypadá) Martin Landa - a kromě toho ví, o čem mluvíš :-) no, nejake zarezy tam mam (VFK, VFR), ale tohle se tyka spise knihovny Proj.4. Grid generoval v ramci DS Jan Jezek [1], jak psal Jachym. Bohuzel ty linky ted nefunguji, zkusil jsem oslovit JJ, zda je nekde jeste nenajde. Honza Ježek pokud vím se v tom problému fakt vyzná, i když je spíš přes Javu (což je detail), ale počítám, že teď má celkem plno rodinných a pracovních povinností. Az bude nejaky cas, tak se snad k tomu dostanu. ML [1] http://freegis.fsv.cvut.cz/gwiki/S-JTSK_/_Grid -- Martin Landa * http://geo.fsv.cvut.cz/gwiki/Landa ___ Talk-cz mailing list Talk-cz@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-cz
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. Quant à area je ne trouve pas spécialement de doc ici, et je suis preneur si plus d'infos : http://wiki.openstreetmap.org/wiki/Overpass_API/Language_Guide Sinon, effectivement si je priorise la requête : node[name~^Toulouse](area.zone); mais comment lui dire ensuite d'appliquer un filtre area Michel Le 14 mai 2014 04:46, Philippe Verdy verd...@wanadoo.fr a écrit : Je pense plutôt que c'est le premier filtre de la seconde requête (node(zone)à qui n'est pas assez sélectif. La zone (le polygone de la France adminstrative de niveaus 2) est gigantesque et contient des millions de noeuds. La première requête (vers la variable zone est relativelent sélective car elle recherche des polygones ayant deux attributs très sélectifs (name=France déjà, affiné par admin_level=2 pour les homonymes peu nombreux): à priori zone ne contient qu'un polygone (assez complexe malgré tout avec quelques milliers de sommets essentiellement sur les frontières terrestres, car en mer les limites des eaux territoriales ne sont pas très compliquées, mais son on choisissait la limite côtière alors là ce sont près de 200 000 noeuds, peut-etre plus maintenant avec les cotes de plus en plus affinées et les ilots). Il vaut mieux faire des requêtes en commençant par le filtre le plus sélectif; celui sur le name élimine quasiment tout, et ne crée donc pas une table temporaire énorme, le filtre sur la zone France est à appliquer après sur ce qui reste (s'il reste quelquechose). Overpass n'a toujours pas d'optimiseur statistique des plans d'exécutions qu'il crée en interne (il ne sait pas évaluer la sélectivité des requêtes: même s'il y a des index primaires utilisables, ça peut être encore inefficace si la sélectivité de la requête est faible, et c'est pire si cela doit passer par des index secondaires n'incluant pas d'autres données nécessaires sur des éléments non sélectifs) et en stockant des tonne de données temporaires. Assez vite il tombe sur les limites de quota (en volume, ou encore plus en nombre d'I/O qui sollicite beaucoup les disques avec des caches peu efficaces, et en fin de compte aboutissant à dépasser le quota de temps d'exécution mais avec cette requête-là on doit tomber sur tous ces quotas en même temps, mais impossible ici de dire lequel tombe en premier). De fait on doit penser à une optimisation manuelle de ses requêtes en évaluant soi-même quelles quantités de données sont séletionnées à chaqué étape. Donc ici il suffit d'inverser les deux requêtes : filtrer d'abord sur les noeuds ayant ce nom puis sélectionner les points restants dans la zone retenue. La solution sera inversée si la zone est assez petite (ne dépasse pas la limite raisonnable de taille d'une zone restangulaire de téléchargement de JOSM par exemple ; toute la France c'est beaucoup trop gros si on n'a pas déjà effectué une première sélection très sélective). Le 13 mai 2014 21:46, Christian Quest cqu...@openstreetmap.fr a écrit : Je pense que c'est la premier area qui cause un dépassement côté serveur... et pas que le serveur t'a envoyé 513MB de data ;) Sur une plus petite zone (Monaco) ça répond bien qu'il n'a rien trouvé. Le 13 mai 2014 20:55, Mides mides@gmail.com a écrit : Bonsoir, À cette requête, pour test, sensée trouver toutes les tags name débutants par cette chaine, trzetgsqgdfgqsdaze cette réponse m’est retournée, et pourtant je ne pense pas qu’il y ait beaucoup de tags name où l'on trouve ce mot. Assez étonnant comme résultat. *Requête :* area [name=France][admin_level=2]-.zone; ( node(area.zone) [name~^trzetgsqgdfgqsdaze ]; ); out skel; *Réponse : * Une erreur est survenue lors de l'exécution de la requête overpass ! Voici ce que l'API overpass a retourné : runtime error: Query run out of memory in query at line 4 using about 513 MB of RAM. Michel ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
Le 14/05/2014 09:04, Mides a écrit : J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. La France administrative de niveau 2 correspond à la France métropolitaine + DOM/TOM et autres territoires. Autant dire que si le polygone d'emprise est un simple rectangle, il couvre quasiment la terre entière ! et mieux vaut donc s'en passer. Si tu ne veut que la France métropolitaine, fait bien attention a ne prendre qu'elle : http://nominatim.openstreetmap.org/details.php?place_id=98156803 http://www.openstreetmap.org/relation/1403916 name = France métropolitaine admin_level = 3 Quant à area je ne trouve pas spécialement de doc ici, et je suis preneur si plus d'infos : http://wiki.openstreetmap.org/wiki/Overpass_API/Language_Guide Sinon, effectivement si je priorise la requête : node[name~^Toulouse](area.zone); mais comment lui dire ensuite d'appliquer un filtre area Michel Le 14 mai 2014 04:46, Philippe Verdy verd...@wanadoo.fr mailto:verd...@wanadoo.fr a écrit : Je pense plutôt que c'est le premier filtre de la seconde requête (node(zone)à qui n'est pas assez sélectif. La zone (le polygone de la France adminstrative de niveaus 2) est gigantesque et contient des millions de noeuds. La première requête (vers la variable zone est relativelent sélective car elle recherche des polygones ayant deux attributs très sélectifs (name=France déjà, affiné par admin_level=2 pour les homonymes peu nombreux): à priori zone ne contient qu'un polygone (assez complexe malgré tout avec quelques milliers de sommets essentiellement sur les frontières terrestres, car en mer les limites des eaux territoriales ne sont pas très compliquées, mais son on choisissait la limite côtière alors là ce sont près de 200 000 noeuds, peut-etre plus maintenant avec les cotes de plus en plus affinées et les ilots). Il vaut mieux faire des requêtes en commençant par le filtre le plus sélectif; celui sur le name élimine quasiment tout, et ne crée donc pas une table temporaire énorme, le filtre sur la zone France est à appliquer après sur ce qui reste (s'il reste quelquechose). Overpass n'a toujours pas d'optimiseur statistique des plans d'exécutions qu'il crée en interne (il ne sait pas évaluer la sélectivité des requêtes: même s'il y a des index primaires utilisables, ça peut être encore inefficace si la sélectivité de la requête est faible, et c'est pire si cela doit passer par des index secondaires n'incluant pas d'autres données nécessaires sur des éléments non sélectifs) et en stockant des tonne de données temporaires. Assez vite il tombe sur les limites de quota (en volume, ou encore plus en nombre d'I/O qui sollicite beaucoup les disques avec des caches peu efficaces, et en fin de compte aboutissant à dépasser le quota de temps d'exécution mais avec cette requête-là on doit tomber sur tous ces quotas en même temps, mais impossible ici de dire lequel tombe en premier). De fait on doit penser à une optimisation manuelle de ses requêtes en évaluant soi-même quelles quantités de données sont séletionnées à chaqué étape. Donc ici il suffit d'inverser les deux requêtes : filtrer d'abord sur les noeuds ayant ce nom puis sélectionner les points restants dans la zone retenue. La solution sera inversée si la zone est assez petite (ne dépasse pas la limite raisonnable de taille d'une zone restangulaire de téléchargement de JOSM par exemple ; toute la France c'est beaucoup trop gros si on n'a pas déjà effectué une première sélection très sélective). Le 13 mai 2014 21:46, Christian Quest cqu...@openstreetmap.fr mailto:cqu...@openstreetmap.fr a écrit : Je pense que c'est la premier area qui cause un dépassement côté serveur... et pas que le serveur t'a envoyé 513MB de data ;) Sur une plus petite zone (Monaco) ça répond bien qu'il n'a rien trouvé. Le 13 mai 2014 20:55, Mides mides@gmail.com mailto:mides@gmail.com a écrit : Bonsoir, À cette requête, pour test, sensée trouver toutes les tags name débutants par cette chaine, trzetgsqgdfgqsdaze cette réponse m’est retournée, et pourtant je ne pense pas qu’il y ait beaucoup de tags name où l'on trouve ce mot. Assez étonnant comme résultat. _Requête :_ _ _ area [name=France][admin_level=2]-.zone; ( node(area.zone) [name~^trzetgsqgdfgqsdaze ]; ); out skel; _Réponse : _
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
Le 14 mai 2014 09:04, Mides mides@gmail.com a écrit : J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. Ça dépend ce que tu veux dire par remonter. Si par là tu entends que tu va recevoir une grosse quantité de données, effectivement ce n'est pas le cas (exemple de Monaco). En fait, l'overpass va commencer par sélectionner tout les noeuds dans l'area... et remonter ça en RAM, sauf que là, sur la France entière... ça dépasse les 512Mo qui sont sa limite et c'est pour ça que tu as une erreur. Ca n'a rien à voir avec ce qui sera transféré au final. -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Un MOOC OSM ?
Le 13/05/2014 17:47, Jean-Guilhem Cailton a écrit : Super. Qui pourrait être intéressé pour participer ? Je suis des vôtres. J'ai pas encore suivi le premier cours mais çà ne saurait tarder. Je pense que pour un premier MOOC autant viser le grand public. Montrer ce qu'on peut faire avec les outils OSM (permalink, routage, uMap) et amener tout un chacun à devenir contributeur débutant avec iD. Antoine. ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Parcelle=yes
2014-05-13 21:35 GMT+02:00 Jean-Marc Liotier j...@liotier.org: A ma droite, parcelle=no - utilisé dans les parages d'Orange pour qualifier 9355 adresses: Comme le tag n'a été ni discuté, ni formalisé sur le wiki (ne parlons même pas d' approuvé), il n'a aucune valeur pour la communauté. En plus, il n'est même pas explicable par lui-même (mais j'ai bien ma petite idée). On ne pourra donc blamer celui ou celle qui déciderait de modifier, altérer ou supprimer ces parcelle=no sur des adresses. - Du côté de Tivaouane, l'aspect cadastral est probablement l'origine du terme, mais il désigne de nos jours en Afrique de l'Ouest une unité d'habitation le plus souvent enceinte d'un mur et rassemblant divers bâtiments autour d'une cour. Je n'ai pas retrouvé cette définition dans wikipédia. Le terme utilisé me gêne beaucoup pour deux raisons : 1. les tags sont normalement en anglais 2. le risque de confusion avec son sens occidental est grand. Je pense qu'il faudrait chercher du côté du vocabulaire anglais un terme qui corresponde mieux à cet usage mais qui s'éloigne du mot parcel comme lot ou plot. Une discussion sur un thème assez proche est encore disponible ici : http://wiki.openstreetmap.org/wiki/Talk:Proposed_features/Tag:allotments%3Dplot Pieren ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
Je pensais que l'on pouvait travailler sur une emprise du style inside_polygon. (admin_level = 2) sans pour cela remonter toutes les données existantes dans ce polygone. Je peux biaiser le problème en définissant une bbox mais ce n'est pas le top non plus (résultats en UK) node[name~^Police](42.33194,-4.79556,51.07167,8.230); Michel Le 14 mai 2014 09:55, Christian Quest cqu...@openstreetmap.fr a écrit : Le 14 mai 2014 09:04, Mides mides@gmail.com a écrit : J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. Ça dépend ce que tu veux dire par remonter. Si par là tu entends que tu va recevoir une grosse quantité de données, effectivement ce n'est pas le cas (exemple de Monaco). En fait, l'overpass va commencer par sélectionner tout les noeuds dans l'area... et remonter ça en RAM, sauf que là, sur la France entière... ça dépasse les 512Mo qui sont sa limite et c'est pour ça que tu as une erreur. Ca n'a rien à voir avec ce qui sera transféré au final. -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
overpass n'est tout simplement pas conçu pour faire des requêtes sur des zones aussi grandes. Le 14 mai 2014 10:36, Mides mides@gmail.com a écrit : Je pensais que l'on pouvait travailler sur une emprise du style inside_polygon. (admin_level = 2) sans pour cela remonter toutes les données existantes dans ce polygone. Je peux biaiser le problème en définissant une bbox mais ce n'est pas le top non plus (résultats en UK) node[name~^Police](42.33194,-4.79556,51.07167,8.230); Michel Le 14 mai 2014 09:55, Christian Quest cqu...@openstreetmap.fr a écrit : Le 14 mai 2014 09:04, Mides mides@gmail.com a écrit : J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. Ça dépend ce que tu veux dire par remonter. Si par là tu entends que tu va recevoir une grosse quantité de données, effectivement ce n'est pas le cas (exemple de Monaco). En fait, l'overpass va commencer par sélectionner tout les noeuds dans l'area... et remonter ça en RAM, sauf que là, sur la France entière... ça dépasse les 512Mo qui sont sa limite et c'est pour ça que tu as une erreur. Ca n'a rien à voir avec ce qui sera transféré au final. -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
[OSM-talk-fr] @osmthis : tout simplement génial !
Les notes OSM ont 2 défauts principaux selon moi: - trop souvent anonymes, donc pas possible de joindre l'auteur dans bien des cas - uniquement textuelles... joindre une photo serait génial Mais ça c'était avant @osmthis ! Il suffit d'envoyer un tweet géolocalisé au compte @osmthis qui se chargera de créer la note qui comprte le lien vers le tweet, donc la photo et le contact avec l'auteur. Il suffisait d'y penser, bravo à sashazykov pour avoir mis en place cette passerelle: http://www.openstreetmap.org/user/sashazykov/diary/21824 Contribuer à OSM (même indirectement) n'a jamais été aussi simple ! -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
Peut être effectivement que ce n'est pas conçu pour cela mais partant donc du principe que c'est le area qui pose problème, en englobant une zone trop important, je serai curieux de savoir pourquoi avec ces deux requêtes, une fonctionne très bien alors que l’autre lève une erreur. L’approche est certes différente mais le area reste identique pour les deux et le résultat renvoyé est normalement le même. area [name=France][admin_level=2]-.zone; ( node(area.zone) [name=Conseil Général]; ); out meta; //-- area [name=France][admin_level=2]-.zone; ( node(area.zone) [name~^Conseil Général$]; ); out meta; Pour info, ce problème est très récent. Michel Le 14 mai 2014 10:48, Christian Quest cqu...@openstreetmap.fr a écrit : overpass n'est tout simplement pas conçu pour faire des requêtes sur des zones aussi grandes. Le 14 mai 2014 10:36, Mides mides@gmail.com a écrit : Je pensais que l'on pouvait travailler sur une emprise du style inside_polygon. (admin_level = 2) sans pour cela remonter toutes les données existantes dans ce polygone. Je peux biaiser le problème en définissant une bbox mais ce n'est pas le top non plus (résultats en UK) node[name~^Police](42.33194,-4.79556,51.07167,8.230); Michel Le 14 mai 2014 09:55, Christian Quest cqu...@openstreetmap.fr a écrit : Le 14 mai 2014 09:04, Mides mides@gmail.com a écrit : J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. Ça dépend ce que tu veux dire par remonter. Si par là tu entends que tu va recevoir une grosse quantité de données, effectivement ce n'est pas le cas (exemple de Monaco). En fait, l'overpass va commencer par sélectionner tout les noeuds dans l'area... et remonter ça en RAM, sauf que là, sur la France entière... ça dépasse les 512Mo qui sont sa limite et c'est pour ça que tu as une erreur. Ca n'a rien à voir avec ce qui sera transféré au final. -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
Visiblement le regv prend toutes les données, mais pas le v normal et ça produit un dépassement de capacité. Sûrement une question d'optimisation de la requête. As-tu essayé de croiser les filtres ? A voir aussi : [maxsize:1073741824] (je parle de la requête en version xml). Le 14 mai 2014 13:49, Mides mides@gmail.com a écrit : Peut être effectivement que ce n'est pas conçu pour cela mais partant donc du principe que c'est le area qui pose problème, en englobant une zone trop important, je serai curieux de savoir pourquoi avec ces deux requêtes, une fonctionne très bien alors que l'autre lève une erreur. L'approche est certes différente mais le area reste identique pour les deux et le résultat renvoyé est normalement le même. area [name=France][admin_level=2]-.zone; ( node(area.zone) [name=Conseil Général]; ); out meta; //-- area [name=France][admin_level=2]-.zone; ( node(area.zone) [name~^Conseil Général$]; ); out meta; Pour info, ce problème est très récent. Michel Le 14 mai 2014 10:48, Christian Quest cqu...@openstreetmap.fr a écrit : overpass n'est tout simplement pas conçu pour faire des requêtes sur des zones aussi grandes. Le 14 mai 2014 10:36, Mides mides@gmail.com a écrit : Je pensais que l'on pouvait travailler sur une emprise du style inside_polygon. (admin_level = 2) sans pour cela remonter toutes les données existantes dans ce polygone. Je peux biaiser le problème en définissant une bbox mais ce n'est pas le top non plus (résultats en UK) node[name~^Police](42.33194,-4.79556,51.07167,8.230); Michel Le 14 mai 2014 09:55, Christian Quest cqu...@openstreetmap.fr a écrit : Le 14 mai 2014 09:04, Mides mides@gmail.com a écrit : J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. Ça dépend ce que tu veux dire par remonter. Si par là tu entends que tu va recevoir une grosse quantité de données, effectivement ce n'est pas le cas (exemple de Monaco). En fait, l'overpass va commencer par sélectionner tout les noeuds dans l'area... et remonter ça en RAM, sauf que là, sur la France entière... ça dépasse les 512Mo qui sont sa limite et c'est pour ça que tu as une erreur. Ca n'a rien à voir avec ce qui sera transféré au final. -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Marc Sibert m...@sibert.fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
A dire vrai, je vais avoir quelques difficultés à optimiser quoique ce soit. J'utilise cette syntaxe, area [name=France][admin_level= 2]-.zone; donc je comprends le fonctionnement mais je ne trouve pas de doc concernant le area, du moins ici : http://wiki.openstreetmap.org/wiki/Overpass_API/Language_Guide D'ailleurs si tu as un point de chute doc, je suis preneur. Michel Le 14 mai 2014 14:02, Marc SIBERT m...@sibert.fr a écrit : Visiblement le regv prend toutes les données, mais pas le v normal et ça produit un dépassement de capacité. Sûrement une question d'optimisation de la requête. As-tu essayé de croiser les filtres ? A voir aussi : [maxsize:1073741824] (je parle de la requête en version xml). Le 14 mai 2014 13:49, Mides mides@gmail.com a écrit : Peut être effectivement que ce n'est pas conçu pour cela mais partant donc du principe que c'est le area qui pose problème, en englobant une zone trop important, je serai curieux de savoir pourquoi avec ces deux requêtes, une fonctionne très bien alors que l’autre lève une erreur. L’approche est certes différente mais le area reste identique pour les deux et le résultat renvoyé est normalement le même. area [name=France][admin_level=2]-.zone; ( node(area.zone) [name=Conseil Général]; ); out meta; //-- area [name=France][admin_level=2]-.zone; ( node(area.zone) [name~^Conseil Général$]; ); out meta; Pour info, ce problème est très récent. Michel Le 14 mai 2014 10:48, Christian Quest cqu...@openstreetmap.fr a écrit : overpass n'est tout simplement pas conçu pour faire des requêtes sur des zones aussi grandes. Le 14 mai 2014 10:36, Mides mides@gmail.com a écrit : Je pensais que l'on pouvait travailler sur une emprise du style inside_polygon. (admin_level = 2) sans pour cela remonter toutes les données existantes dans ce polygone. Je peux biaiser le problème en définissant une bbox mais ce n'est pas le top non plus (résultats en UK) node[name~^Police](42.33194,-4.79556,51.07167,8.230); Michel Le 14 mai 2014 09:55, Christian Quest cqu...@openstreetmap.fr a écrit : Le 14 mai 2014 09:04, Mides mides@gmail.com a écrit : J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. Ça dépend ce que tu veux dire par remonter. Si par là tu entends que tu va recevoir une grosse quantité de données, effectivement ce n'est pas le cas (exemple de Monaco). En fait, l'overpass va commencer par sélectionner tout les noeuds dans l'area... et remonter ça en RAM, sauf que là, sur la France entière... ça dépasse les 512Mo qui sont sa limite et c'est pour ça que tu as une erreur. Ca n'a rien à voir avec ce qui sera transféré au final. -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Marc Sibert m...@sibert.fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
[OSM-talk-fr] Quelqu'un pour arrêter Super-Map ?
Bonjour, Il applique systématiquement les qualifications de la Route500 et supprime les espaces des références de voies. et en plus il ajoute un petit commentaire : comment=Please, don't use Bing for the roads classification, it's prohibited! ***secondary road*** A+ PS : j'arrive pas à ma connecter sur la messagerie OSM où je suis. -- Marc Sibert m...@sibert.fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
Tu pourrais aussi taper sur l'overpass-FR... ça éviterai le recourt à l'area ;) Le 14 mai 2014 14:29, Mides mides@gmail.com a écrit : A dire vrai, je vais avoir quelques difficultés à optimiser quoique ce soit. J'utilise cette syntaxe, area [name=France][admin_level= 2]-.zone; donc je comprends le fonctionnement mais je ne trouve pas de doc concernant le area, du moins ici : http://wiki.openstreetmap.org/wiki/Overpass_API/Language_Guide D'ailleurs si tu as un point de chute doc, je suis preneur. Michel Le 14 mai 2014 14:02, Marc SIBERT m...@sibert.fr a écrit : Visiblement le regv prend toutes les données, mais pas le v normal et ça produit un dépassement de capacité. Sûrement une question d'optimisation de la requête. As-tu essayé de croiser les filtres ? A voir aussi : [maxsize:1073741824] (je parle de la requête en version xml). Le 14 mai 2014 13:49, Mides mides@gmail.com a écrit : Peut être effectivement que ce n'est pas conçu pour cela mais partant donc du principe que c'est le area qui pose problème, en englobant une zone trop important, je serai curieux de savoir pourquoi avec ces deux requêtes, une fonctionne très bien alors que l'autre lève une erreur. L'approche est certes différente mais le area reste identique pour les deux et le résultat renvoyé est normalement le même. area [name=France][admin_level=2]-.zone; ( node(area.zone) [name=Conseil Général]; ); out meta; //-- area [name=France][admin_level=2]-.zone; ( node(area.zone) [name~^Conseil Général$]; ); out meta; Pour info, ce problème est très récent. Michel Le 14 mai 2014 10:48, Christian Quest cqu...@openstreetmap.fr a écrit : overpass n'est tout simplement pas conçu pour faire des requêtes sur des zones aussi grandes. Le 14 mai 2014 10:36, Mides mides@gmail.com a écrit : Je pensais que l'on pouvait travailler sur une emprise du style inside_polygon. (admin_level = 2) sans pour cela remonter toutes les données existantes dans ce polygone. Je peux biaiser le problème en définissant une bbox mais ce n'est pas le top non plus (résultats en UK) node[name~^Police](42.33194,-4.79556,51.07167,8.230); Michel Le 14 mai 2014 09:55, Christian Quest cqu...@openstreetmap.fr a écrit : Le 14 mai 2014 09:04, Mides mides@gmail.com a écrit : J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. Ça dépend ce que tu veux dire par remonter. Si par là tu entends que tu va recevoir une grosse quantité de données, effectivement ce n'est pas le cas (exemple de Monaco). En fait, l'overpass va commencer par sélectionner tout les noeuds dans l'area... et remonter ça en RAM, sauf que là, sur la France entière... ça dépasse les 512Mo qui sont sa limite et c'est pour ça que tu as une erreur. Ca n'a rien à voir avec ce qui sera transféré au final. -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Marc Sibert m...@sibert.fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
Tout à fait dans le mesure où c'est la France qui m’intéresse, mais où trouve t-on cet outil ou se serveur. J'ai bien essayé de changer l'adresse sur overpass turbo eu , mais aucun résultat. Michel Le 14 mai 2014 15:35, Christian Quest cqu...@openstreetmap.fr a écrit : Tu pourrais aussi taper sur l'overpass-FR... ça éviterai le recourt à l'area ;) Le 14 mai 2014 14:29, Mides mides@gmail.com a écrit : A dire vrai, je vais avoir quelques difficultés à optimiser quoique ce soit. J'utilise cette syntaxe, area [name=France][admin_level= 2]-.zone; donc je comprends le fonctionnement mais je ne trouve pas de doc concernant le area, du moins ici : http://wiki.openstreetmap.org/wiki/Overpass_API/Language_Guide D'ailleurs si tu as un point de chute doc, je suis preneur. Michel Le 14 mai 2014 14:02, Marc SIBERT m...@sibert.fr a écrit : Visiblement le regv prend toutes les données, mais pas le v normal et ça produit un dépassement de capacité. Sûrement une question d'optimisation de la requête. As-tu essayé de croiser les filtres ? A voir aussi : [maxsize:1073741824] (je parle de la requête en version xml). Le 14 mai 2014 13:49, Mides mides@gmail.com a écrit : Peut être effectivement que ce n'est pas conçu pour cela mais partant donc du principe que c'est le area qui pose problème, en englobant une zone trop important, je serai curieux de savoir pourquoi avec ces deux requêtes, une fonctionne très bien alors que l’autre lève une erreur. L’approche est certes différente mais le area reste identique pour les deux et le résultat renvoyé est normalement le même. area [name=France][admin_level=2]-.zone; ( node(area.zone) [name=Conseil Général]; ); out meta; //-- area [name=France][admin_level=2]-.zone; ( node(area.zone) [name~^Conseil Général$]; ); out meta; Pour info, ce problème est très récent. Michel Le 14 mai 2014 10:48, Christian Quest cqu...@openstreetmap.fr a écrit : overpass n'est tout simplement pas conçu pour faire des requêtes sur des zones aussi grandes. Le 14 mai 2014 10:36, Mides mides@gmail.com a écrit : Je pensais que l'on pouvait travailler sur une emprise du style inside_polygon. (admin_level = 2) sans pour cela remonter toutes les données existantes dans ce polygone. Je peux biaiser le problème en définissant une bbox mais ce n'est pas le top non plus (résultats en UK) node[name~^Police](42.33194,-4.79556,51.07167,8.230); Michel Le 14 mai 2014 09:55, Christian Quest cqu...@openstreetmap.fr a écrit : Le 14 mai 2014 09:04, Mides mides@gmail.com a écrit : J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. Ça dépend ce que tu veux dire par remonter. Si par là tu entends que tu va recevoir une grosse quantité de données, effectivement ce n'est pas le cas (exemple de Monaco). En fait, l'overpass va commencer par sélectionner tout les noeuds dans l'area... et remonter ça en RAM, sauf que là, sur la France entière... ça dépasse les 512Mo qui sont sa limite et c'est pour ça que tu as une erreur. Ca n'a rien à voir avec ce qui sera transféré au final. -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Marc Sibert m...@sibert.fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Parcelle=yes
On 05/14/2014 10:26 AM, Pieren wrote:
- Du côté de Tivaouane, l'aspect cadastral est probablement l'origine du
terme, mais il désigne de nos jours en Afrique de l'Ouest une unité
d'habitation le plus souvent enceinte d'un mur et rassemblant divers
bâtiments autour d'une cour.
Je n'ai pas retrouvé cette définition dans wikipédia. Le terme utilisé
me gêne beaucoup pour deux raisons : 1. les tags sont normalement en
anglais 2. le risque de confusion avec son sens occidental est
grand.
Je pense qu'il faudrait chercher du côté du vocabulaire anglais un
terme qui corresponde mieux à cet usage mais qui s'éloigne du mot
parcel comme lot ou plot. Une discussion sur un thème assez
proche est encore disponible ici :
http://wiki.openstreetmap.org/wiki/Talk:Proposed_features/Tag:allotments%3Dplot
Effectivement, dans http://www.liutprand.it/articoliMondo.asp?id=89 qui
parle en Anglais de l'histoire immobilière de Dakar, c'est le terme
'allotment' qui est retenu.
Si on regarde par exemple http://www.lefaso.net/spip.php?article48398 on
trouve dans le processus de lotissement une synonymie entre 'parcelle'
et 'lot' - ce qui renforce la pertinence de 'allotment' dont 'lot' est
la traduction.
En France, d'après
http://www.developpement-durable.gouv.fr/IMG/Fiche%20les%20lotissements%20%C3%A0%20partir%20du%201er%20mars%202012.pdf
: un lot est une parcelle issue d'une division foncière réalisée en vue
de construire sur celle-ci... Ca correspond bien.
Dans mes recherches, je tombe par hasard sur
http://www.ambaburkina-fr.org/afrique/ où se trouve l'adresse de
l'ambassade du Burkina à Tripoli: Route de Gargeresh, Lot n° 7 BP 81
902 Tripoli-Libye - le lot y est aussi l'unité d'habitation.
De même, l'adresse du consulat du Congo Brazzaville au Bénin ('Lot 34-E,
PK 6, Tokplégbé, Cotonou') ou celle de la Guinée Conakry ('Lot 106 B,
Sodjèatimè, Cotonou') mentionnées à
http://www.gouv.bj/contacts/consulats-au-b-nin
Une relation hiérarchique existe-t-elle entre lot et parcelle ?
http://www.diarrasec.com/index.php/contact/burkina-faso mentionnant le
contact Burkinabé d'une société d'expertise comptable suggère que oui:
l'adresse est 'parcelle 05, lot 11, section 474' à Ouaga 2000... Il y a
donc une relation hiérarchique entre parcelle et lot, les parcelles
étant subdivisées en lots destinés à la construction lors du lotissement
- ce qui est cohérent avec la définition Française du lot et ce que j'ai
lu et entendu du processus Burkinabé de lotissement.
Il semble donc bien qu'on converge vers l'utilisation du concept de
'lot' reconnu internationalement et donc une étiquette Openstreetmap
'allotment' pour désigner le concept ouest-Africain de parcelle (faux
ami du terme légal Français 'parcelle') en tant qu'unité d'habitation
porteuse d'adresse, souvent enceinte d'un mur, rassemblant plusieurs
bâtiments souvent autour d'une cour.
Dans Openstreetmap, 'allotments' (au pluriel) semble actuellement être
fortement lié aux potagers urbains mais 'allotment' (au singulier) me
semble idéal. http://taginfo.openstreetmap.org/keys/allotment#values
indique que 'allotment=plot' est utilisé pour décrire les lots d'un
potager urbain à Birmingham: http://overpass-turbo.eu/s/3nQ
Alors, 'allotment=plot' ? Une autre valeur d'allotment=* ? Réponse à
http://forum.wordreference.com/showthread.php?t=541552 qui décrit la
différence entre 'lot' et 'plot' : 'lot' est la subdivision
constructible d'une parcelle cadastrale tandis que 'plot' est
grosso-modo un bout de potager sans définition légale spécifique.
Je propose donc 'allotment=plot'
Des commentaires ?
___
Talk-fr mailing list
Talk-fr@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] [Talk-sn] Parcelle=yes
On 05/14/2014 04:49 PM, Jean-Marc Liotier wrote: Je propose donc 'allotment=plot' Raaah - je dis exactement le contraire de ce que je voulais dire... Reprenons... La proposition est, comme vous l'aviez certainement compris, 'allotment=lot' ... ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Parcelle=yes
Je me permets d'intervenir à propos du cas malgache. A Madagascar, un système d'adresse utilisant le terme lot est utilisé. J'ai documenté ça dans le wiki: http://wiki.openstreetmap.org/wiki/FR:WikiProject_Madagascar#Adresses Néanmoins, je n'ai pas parlé du cas de la ville de Mahajanga où les adresses comprennent un Lot et une Parcelle. Un lot comporte plusieurs parcelles. - Du côté de Tivaouane, l'aspect cadastral est probablement l'origine du terme, mais il désigne de nos jours en Afrique de l'Ouest une unité d'habitation le plus souvent enceinte d'un mur et rassemblant divers bâtiments autour d'une cour. Ca ressemble bien aux lots malgaches (ou aux parcelles majungaises). Une relation hiérarchique existe-t-elle entre lot et parcelle ? http://www.diarrasec.com/index.php/contact/burkina-faso mentionnant le contact Burkinabé d'une société d'expertise comptable suggère que oui: l'adresse est 'parcelle 05, lot 11, section 474' à Ouaga 2000... Il y a donc une relation hiérarchique entre parcelle et lot, les parcelles étant subdivisées en lots L'inverse du cas de Majajanga... Je propose donc 'allotment=lot' Je trouve qu'on va dans un truc très spécifique avec pratiquement clé et valeur dédiées. En plus, c'est limite tracer chaque propriété privée. Un peu comme si on traçait les parcelles cadastrales en France... Et pour les adresses? addr:allotment ou addr:lot Sur un noeud au milieu? Sur le polygone? Eric ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr
Re: [OSM-talk-fr] Overpass.Turbo.eu résultat requête
Effectivement sacrée optimisation des requêtes. donc, si je lance cette requête sur une zone bien définie qui concerne la France et DOM/TOM, on peut effectivement supposer qu'il y a une surcharge des données au niveau AREA --- area [name=France][admin_level=2]-.zone; ( node(area.zone) [name~^Conseil Général$]; ); out meta; -- Maintenant, si je lance la même requête mais sur le monde entier, sans AREA, ça passe. Il y a quelque chose qui m'échappe. node [name~^Conseil Général$]; out meta; Michel Le 14 mai 2014 15:35, Christian Quest cqu...@openstreetmap.fr a écrit : Tu pourrais aussi taper sur l'overpass-FR... ça éviterai le recourt à l'area ;) Le 14 mai 2014 14:29, Mides mides@gmail.com a écrit : A dire vrai, je vais avoir quelques difficultés à optimiser quoique ce soit. J'utilise cette syntaxe, area [name=France][admin_level= 2]-.zone; donc je comprends le fonctionnement mais je ne trouve pas de doc concernant le area, du moins ici : http://wiki.openstreetmap.org/wiki/Overpass_API/Language_Guide D'ailleurs si tu as un point de chute doc, je suis preneur. Michel Le 14 mai 2014 14:02, Marc SIBERT m...@sibert.fr a écrit : Visiblement le regv prend toutes les données, mais pas le v normal et ça produit un dépassement de capacité. Sûrement une question d'optimisation de la requête. As-tu essayé de croiser les filtres ? A voir aussi : [maxsize:1073741824] (je parle de la requête en version xml). Le 14 mai 2014 13:49, Mides mides@gmail.com a écrit : Peut être effectivement que ce n'est pas conçu pour cela mais partant donc du principe que c'est le area qui pose problème, en englobant une zone trop important, je serai curieux de savoir pourquoi avec ces deux requêtes, une fonctionne très bien alors que l’autre lève une erreur. L’approche est certes différente mais le area reste identique pour les deux et le résultat renvoyé est normalement le même. area [name=France][admin_level=2]-.zone; ( node(area.zone) [name=Conseil Général]; ); out meta; //-- area [name=France][admin_level=2]-.zone; ( node(area.zone) [name~^Conseil Général$]; ); out meta; Pour info, ce problème est très récent. Michel Le 14 mai 2014 10:48, Christian Quest cqu...@openstreetmap.fr a écrit : overpass n'est tout simplement pas conçu pour faire des requêtes sur des zones aussi grandes. Le 14 mai 2014 10:36, Mides mides@gmail.com a écrit : Je pensais que l'on pouvait travailler sur une emprise du style inside_polygon. (admin_level = 2) sans pour cela remonter toutes les données existantes dans ce polygone. Je peux biaiser le problème en définissant une bbox mais ce n'est pas le top non plus (résultats en UK) node[name~^Police](42.33194,-4.79556,51.07167,8.230); Michel Le 14 mai 2014 09:55, Christian Quest cqu...@openstreetmap.fr a écrit : Le 14 mai 2014 09:04, Mides mides@gmail.com a écrit : J'ai un peu de mal à appréhender cette API, comme par exemple cette syntaxe : area [name=France][admin_level=2]-.zone; je pensais qu'à ce niveau là, je ne remontais pas une quantité phénoménale de données mais juste le polygone d'emprise. Ça dépend ce que tu veux dire par remonter. Si par là tu entends que tu va recevoir une grosse quantité de données, effectivement ce n'est pas le cas (exemple de Monaco). En fait, l'overpass va commencer par sélectionner tout les noeuds dans l'area... et remonter ça en RAM, sauf que là, sur la France entière... ça dépasse les 512Mo qui sont sa limite et c'est pour ça que tu as une erreur. Ca n'a rien à voir avec ce qui sera transféré au final. -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Marc Sibert m...@sibert.fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr -- Christian Quest - OpenStreetMap France ___ Talk-fr mailing list Talk-fr@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-fr ___ Talk-fr
