[OSM-talk] quality insurance tool osmose
Hello, Osmose [1], a french quality insurance tool has now some analysers analysing data out of France. * One of the analyser [2] is raising errors on relations boundary=adminitrative wich : - are not closed - have some end nodes more than two times (ie that some way is included more tha 2 times). This backend is runned once a day. It also analyses recursive relations (relations including other relations such as relation 11980). * Another backend [3] is analysing self-intersecting polygons. It only raises errors on Europe and is runned twice a week (because of postgis import duration). This quality insurance tool is multi backend, so people wanting to analyse data may report them on the main map. Update procedure is (lighly) described on the wiki [4]. Sources of my backends analysing data on france are available on [5]. [1] http://osmose.openstreetmap.fr/map/cgi-bin/index.py [2] http://osmose.openstreetmap.fr/map/cgi-bin/index.py?source=3 [3] http://osmose.openstreetmap.fr/map/cgi-bin/index.py?source=28 [4] http://wiki.openstreetmap.org/wiki/Osmose [5] http://osm1.crans.org/src/osmose/ -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] North Pole
Peter Childs a écrit : > Why is the North Pole at 0,0? > http://www.openstreetmap.org/?lat=0&lon=0&zoom=16&layers=000BFTF No idea... is it hard coded ? > and why at higher zoom levels does it appear as a Car Park. I can't > see any data but it seams to appear in all the different renderer s. Because there was a car park ;-) I deleted it : http://www.openstreetmap.org/api/0.6/changeset/1563349/download > Last Time I checked you could not park your car in the middle of the Atlantic. Sure ? Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Python API
Etienne Chové a écrit : > Dears, > > I wrote a python class to communicate with OSM API (read, write, > update). For interested users, informations are here [1]. > > May I put sources on the dev server ? This is now hosted on the main osm svn repositrory. I appologize for people wanting it on an external repository but I prefer to put it on an osm server. It's now at : http://svn.openstreetmap.org/applications/utils/python_lib/ Thanks to Eddy Petrișor for some bugfix and ideas proposed on his repository. -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Python API
Etienne Chové a écrit : > Dears, > > I wrote a python class to communicate with OSM API (read, write, > update). For interested users, informations are here [1]. > > May I put sources on the dev server ? Here it is : http://osmose.openstreetmap.fr/src/modules/OsmApi.py -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Python API
Eddy Petrișor a écrit : > 2009/4/29 Etienne Chové : >> [1] http://wiki.openstreetmap.org/wiki/PythonOsmApi > > BTW, under which license are publishing your changes? May I suggest > one of 'LGPL', 'GPL v2 or later' or 'GPL v3 or later'? I would be fine > with either of those, but I think LGPL would be the best choice for > such code as a "library". Until now, all my publcations are GPLv3, and this one yet have it licence header. I don't know differences with LGPL. -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Python API
Thomas Wood a écrit : > See also the osmparser.py class, that may also be useful. > http://trac.openstreetmap.org/browser/applications/editors/django/osmeditor/lib/osmparser.py Interesting... does it work with API 0.6 and changesets ? (see 0.5 inside). -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Python API
Eddy Petrișor a écrit : > 2009/4/29 Etienne Chové : >> Dears, >> >> I wrote a python class to communicate with OSM API (read, write, >> update). For interested users, informations are here [1]. > > COOL! I was thinking of writing such a class myself, but it's great > you did it :) . > >> May I put sources on the dev server ? > > May I suggest keeping it in a git repository so other people than the > ones with commit access to the openstreetmap svn can contribute? > > I have used repo.or.cz which is a free service offered by Petr Baudis, > author of many improvements in the git user interface for the > osm-helpers improved code. > > I am probably going to be one of your early adopters :-) I can put it where ever people want it. I think svn.openstreetmap.org is the best. Where to ask to get svn account ? -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Python API
Shaun McDonald a écrit : > > On 29 Apr 2009, at 17:00, Etienne Chové wrote: > >> Dears, >> >> I wrote a python class to communicate with OSM API (read, write, >> update). For interested users, informations are here [1]. >> >> May I put sources on the dev server ? >> > > Wouldn't it be better to put them into SVN? It's what I'd like to say ;-) -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
[OSM-talk] Python API
Dears, I wrote a python class to communicate with OSM API (read, write, update). For interested users, informations are here [1]. May I put sources on the dev server ? -- Etienne [1] http://wiki.openstreetmap.org/wiki/PythonOsmApi ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] 500 Internal Server Error during upload
Jacek Konieczny a écrit : > On Tue, Apr 28, 2009 at 02:29:12AM -0500, Ian Dees wrote: >> Or! It could be all the Germans racing to get to the 100th changeset: > > 100th reached, then the load should drop now :) > And it should be quite sane, until near 1234567th, I guess ;) Congratulation to french contributor nin2jardin who reached the 100th changeset, and to all osm contributors who have made 99 previous changesets so he could reached this one. -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] planet? geofabrik download...
Frederik Ramm a écrit : > Hi, > > Etienne Chové wrote: >> I think today france extract still have the same problem : > > It should now be fixed, finally... I hope ;-) Yes, all right. Thanks for that. Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] planet? geofabrik download...
Frederik Ramm a écrit : > Hi, > > Gary68 wrote: >> just wanted to use a planet slice from geofabrik and noticed that in >> bremen.osm.bz2 at least 4 referenced nodes are missing: > > Osmosis now exports incomplete objects by default, rather than clipping > them. One has to explicitly request clipping which is what I've done for > tonight's job, so expect old-style clipped files tomorrow. I think today france extract still have the same problem : % grep "\"12631399\"" france.osm This node is used in a way but not declared. -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] planet? geofabrik download...
Gary68 a écrit : > there are situations where this is needed. for instance when i want to > have a rectangle exactly, if i want to clip sharp (maps/pictures). > > on the other hand if i have a file with referenced nodes where the > coordinates are missing i will have a problem when i want to use the > coordinates. I think the best would be keeping unclipped way and adding missing nodes in the osm file. Is that possible with osmosis ? -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] planet? geofabrik download...
Gary68 a écrit : > hi, > > just wanted to use a planet slice from geofabrik and noticed that in > bremen.osm.bz2 at least 4 referenced nodes are missing: > > node 31088130 not found > node 257415561 not found > node 285091039 not found > node 303607968 not found > > those are referenced by ways in that file. I have exactly the same problem. It appears when a way contains some nodes inside the bounding polygon and some outside. In the osm file you have : - nodes inside bounding polygon - ways containing at leat one node inside Last week we got inside nodes and truncated ways, but theses ways where not real beacause in fact they contains more nodes, now nodes are cited but not includes. Better or worse ? I have no idea but it makes my own scripts crashed... I'm patching them. -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] [tactile_paving=yes/(no)/irritating] Feature Proposal - RFC - (tactile paving)"
Claudius a écrit : > Wenn du jetzt noch bitte im Proposal erklären könntest, wofür der Wert > "irritating" steht... Der irritiert mich nämlich, da er weder dort noch > in Wikipedia erklärt wird. Anybody to translate it in English as it's on talk@ ? Thanks -- Etienne ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] more changeset question
Shaun McDonald a écrit : > On 22 Apr 2009, at 10:56, maning sambale wrote: > >> limits to the number if node/way edits/deletion/addition per changeset >> > > There is a maximum of 50,000 nodes, ways or relations per changeset. > There is a maximum of 2000 nodes per way. Is-it possible to give max change count in [1] ? As other limits are in this page. -- Etienne [1] http://www.openstreetmap.org/api/capabilities ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
