Re: [OSM-talk] [Talk-us] plea for bot quality control
On Mon, May 3, 2010 at 8:26 AM, Frederik Ramm wrote: > I think it would be a good idea to create a repository of such tagging > standards where everyone can create new standards to his heart's > content, and the user of an editor can then select one of the existing > standards (or create a new one). - That would most likely create a > situation in which a few popular tagging standards emerge, and a certain > competition among these will help create something really good in the > long run without ever having to have something "authoritative". > > I like this idea a lot, Frederik. I imagine something like a presets file that all editors could parse to display commonly-used key/value pairs. Each would have a description (optionally i18nized (so that "motorway" would describe a "highway" in the US and an actual "motorway" in the UK)) and perhaps could be auto-updated weekly from the planet file. That way JOSM and Potlatch could present suggestions that are broader than just the currently-loaded OSM file (in JOSM's case) and some preset file (in Potlatch's case). Perhaps some simplified editor could only show the top X values for a given tag (ranked by use in the wild) while a more advanced editor could show a whole lot more. ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] [Talk-us] plea for bot quality control
Hi, Chris Hunter wrote: > Perhaps a bot to auto-ban accounts that are running unauthorized bots ;) I'm with you on implementing measures to allow a stricter control of automated edits. > One of the other things that OSM desperately needs is a way to > coordinate tagging across editors and renderers. Yes I know that the > traditional answer is "tag for reality, not for the renderer", but in > all seriousness, the data is useless without renderers and vice-versa. As long as that coordination happens voluntarily and by evolution and good information rather than be vote and decree, yes that would be a good idea. > I haven't had the time lately to check-in with the US SOTM group, but > I'd like to see a schema file for tags that have gone through the > approval process that JOSM, mkgmap, Mapnik, Potlatch, etc... can all read. While it is certainly possible to provide such a file and certainly possible for editors to support it, I am 99% sure that no sane editor will limit itself to tags that have gone through one approval process or the other. > Obviously, the file would need to be kept centralized and only be > writable by a limited number of people. No, that is not at all obivous. If you intend to create a file that has a certain "authority" and writability only for a select elite then I can assure you that it will be sidelined. I think it would be a good idea to create a repository of such tagging standards where everyone can create new standards to his heart's content, and the user of an editor can then select one of the existing standards (or create a new one). - That would most likely create a situation in which a few popular tagging standards emerge, and a certain competition among these will help create something really good in the long run without ever having to have something "authoritative". Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09" E008°23'33" ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] [Talk-us] plea for bot quality control
Perhaps a bot to auto-ban accounts that are running unauthorized bots ;) Seriously though, I've all-but stopped contributing to the map until after the SOTMs because of issues like this. One of the other things that OSM desperately needs is a way to coordinate tagging across editors and renderers. Yes I know that the traditional answer is "tag for reality, not for the renderer", but in all seriousness, the data is useless without renderers and vice-versa. I haven't had the time lately to check-in with the US SOTM group, but I'd like to see a schema file for tags that have gone through the approval process that JOSM, mkgmap, Mapnik, Potlatch, etc... can all read. Obviously, the file would need to be kept centralized and only be writable by a limited number of people. CVS, GIT, and MediaWIKI all have user-rights functions to accomplish this. Any takers? On Sun, May 2, 2010 at 10:52 AM, Nakor wrote: > On 5/2/2010 9:59 AM, Richard Welty wrote: > > as i work my way west along the PA/NY border fixing connectivity > > problems, i'm > > suitably impressed by the patches west of Waverly/NY 17 where the ways > > are missing, > > only the nodes have been left behind. > > > > please, bots are powerful tools. don't run bots that do things like > > remove dups unless > > you have a quality control plan. please. > > > > > > Richard, > > You might want to extend your message to talk as there are quite a few > bots running that are not maintained by US based contributors. In PA I > have seen a lot of duplicate nodes fixed but with duplicate ways left > behind. I messaged some of the users doing this but did not get any > response so far. > > Thanks, > > N. > > ___ > Talk-us mailing list > talk...@openstreetmap.org > http://lists.openstreetmap.org/listinfo/talk-us > ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
