Re: [OSM-talk] How to contribute new tag and symbol?
On May 22, 2009, at 12:06, Ingo Lantschner wrote: Beside of the tag-name: I still have no idea, how new developed tags, rules and symbols can be fed back into the project. Some possibilities -- different people have different opinions on what's useful or required: * discuss it (mailing list, forum, wiki) * document on the wiki * use it * work it through the feature proposal process on the wiki * file tickets with the renderers (http://trac.openstreetmap.org/) * ... Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] my etrex died?
On Mar 22, 2009, at 03:01, Maning Sambale wrote: For some reasons I can't explain, my etrex couldn't start anymore. At first I thought it's the battery but plugging it to my usb doesn't work either. I see no physical damage in the unit and it's still working yesterday. Any idea why? Or how do repair it. In addition to what Karl said, also try removing the batteries, and possibly try turning it on without batteries or USB attached. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] rights of way and designation=*
On Feb 26, 2009, at 11:57, Mike Harris wrote: I support Richard's logic 100% but am unsure whether I want to put the effort in to go back and add the tags to all those ways I have done! (;) - at least until there had been enough discussion that this was well established as a new standard. Is the proposal for a new key designation (afaik there isn't such a key yet in (common) use??) with the various values - footpath, bridleway, restricted_byway, BOAT and - perhaps - ORPA, adopted, unadopted? I've had a look at tagwatch (unfortunately not terribly up-to-date) and documented this suggestion and current use at http://wiki.openstreetmap.org/wiki/Key:designation . Please flesh the page out! It'd be nice to have a list of sensible values there; also, should there be a :uk or uk: in the tag or value? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] rights of way and designation=*
On Feb 26, 2009, at 12:53, Ed Loach wrote: Robert wrote: values there; also, should there be a :uk or uk: in the tag or value? I wouldn't have thought the uk: was needed, as you can presumably tell that from where the path is. Also, I think the various statuses may vary in the different constituent countries of the UK so the uk: tag would also be inappropriate. And after all, it would be a UK public footpath if it's situated in the UK, so there's really no need to track that. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Render strangeness
On Feb 22, 2009, at 16:04, Thomas Wagner wrote: I experienced the same problem with some of my roads before. Unfortunately I did not get a helpful response on the list. But I found, that editing the road again (also it is ok in edit mode), especially splitting and moving one node of each segment, had the road rendered correctly next times. How was Dirk's reply at http://lists.openstreetmap.org/pipermail/talk/2009-January/033002.html not helpful? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Walking Routes - wiki needs some work?
On Feb 20, 2009, at 11:14, Ed Loach wrote: In the wiki, Relation:route[1] suggests network of uk_ldp for the UK long distance path network, but Walking_Routes[2] suggests iwn/nwn/rwn/lwn for network types. It looks like the uk_ldp goes back over a year to October 2007, so there are probably a number of these already in existence. The contradiction between the two pages has also led to a relation I created based on the Walking_Routes page being amended to that on the Relation:Route one (which is understandable if people are already used to the Relation:Route definitions). I don’t know how to find out how many relations already exist tagged with network=uk_ldp - perhaps someone could find out? And perhaps someone could decide what to do about the wiki contradictions. Also on the Relation:route page the Cambridge citibus network is still mentioned in the network description, but the value in the network column has been removed (browsing the page history). If you follow the tagwatch links from Relation:route, you can get at the numbers. network GB Europe uk_ldp 21 21 lwn 1 191 rwn 4 354 nwn 0 22 iwn 0 0 Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Render strangeness
On Feb 21, 2009, at 07:44, Matt White wrote: I was just pottering around checking some of the mapping I had done, and noticed some strangeness in the rendering of a road I mapped about two months ago: http://www.openstreetmap.org/?lat=-37.5138lon=144.4427zoom=14layers=B000FTF The road in question is Blue Gum Track, and it is definitely not straight as per the map. All the renders are showing it as a straight line, yet I recall it being particuarly windy (and I just about broke my neck when I hit a washout on it at 40km/h). The straight bit is a separate way (there's a fair few nodes on these windy roads, so I tend to chop them into a couple of pieces) to the rest of the Blue Gum track way to the south. Potlatch is showing the way is it should be. Anyway, can someone with more smarts than me have a look at it and see if there is anything obviously wrong here. http://www.openstreetmap.org/browse/way/29308957 It's got lots of nodes, but shows as a straight line since the intermediate nodes have been deleted (follow some of the node links). Probably, a minor edit in Potlatch (say changing a tag) will restore it. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
[OSM-talk] near longitude 180
Hello, there's some strange coastline data near longitude 180. See eg http://openstreetmap.org/browse/way/24020654 I thought I'd ask before trying to fix this, in case I'd flood the world otherwise. I haven't found an editor that works well in this area -- they all seem to think the world ends at ±180 degrees. Here's a challenge: Which editor will be first to support seamless editing all around the world? (Let's leave the poles out for now.) For testing, I traced a small lake crossing the line as a single closed way with natural=water: http://openstreetmap.org/browse/way/31254026 . That's good data as far as I understand the data model. It'd be nice to have a reasonably well-mapped area across this meridian as a testing ground for tools. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] near longitude 180
On Feb 19, 2009, at 15:38, andrzej zaborowski wrote: 2009/2/19 Robert (Jamie) Munro rjmu...@arjam.net: The right solution here is to map 0-360 degrees to the unsigned integers 0-2^32. When you get an overflow, the right thing happens. It also makes the most efficient use of the resolution available. Yes, that sounds good. It doesn't really solve the main problem which is likely that a way segment connecting two nodes on the opposite sides of 180 deg can be interpreted as either crossing the meridian 180 or not. And there doesn't seem to be a solution to that, other than duplicating part of surface, i.e. allowing nodes so be stored as having longitude of 181 deg or -179 deg. No, the interpretation of that segment not crossing the meridian is wrong. Viewing the earth as a sphere, the segment from node 1 to node 2 is the shorter part of the great circle going through node 1 and node 2. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Announce: OSM2Go map editor 0.6.13 released for Maemo, Debian, and Ubuntu
On Feb 18, 2009, at 02:02, Andrew Chadwick (mailing lists) wrote: The OSM2go mobile map editor has been updated, please update your copies if you're following what we do :) This version adds full editability of relations, and we'd really love your feedback. So it's available as binaries for a number of popular platforms. Finally got around to giving it a try: it compiled fine on OS X, though installation required the following patch, as install -D is a GNUism. --- data/Makefile 2009-02-16 21:30:10.0 +0100 +++ ../osm2go-0.6.13-mod/data/Makefile 2009-02-18 13:17:01.0 +0100 @@ -19,7 +19,8 @@ install -m 644 *.xml $(DESTDIR)$(PREFIX)/share/$(APP) install -m 644 *.style $(DESTDIR)$(PREFIX)/share/$(APP) for f in `find icons -name *.png`; do \ - install -D -m 644 $$f $(DESTDIR)$(PREFIX)/share/$(APP)/$$f ; \ + install -d $(DESTDIR)$(PREFIX)/share/$(APP)/`dirname $$f` ;\ + install -m 644 $$f $(DESTDIR)$(PREFIX)/share/$(APP)/$$f ; \ done; I didn't do more than change a couple of tags so far, but the first impression was very good! Very smooth interface, congratulations! I'll try using it for a bit at least. The very first impression was slightly less good: it's quite difficult to set up a new project -- did I miss some way to paste a slippy map URL, or is there a slippy map chooser available or planned? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Oxbridges of Konigsberg
On Feb 12, 2009, at 12:06, Stephen Gower wrote: On Thu, Feb 12, 2009 at 10:47:10AM +, Andrew Chadwick (email lists) wrote: Stephen Gower wrote: What's the most efficient route for visiting all Oxford's colleges? So, since the data for Oxford is pretty much there, is this a challenge any of the routing engines can help with? Not purely based on OSM data, you'll need OXPOINTS information for the lodge locations too. Given who the friend is, I suspect the reason for asking is that they want to improve the OXPOINTS data, which, if this is the case will also have the advantage that it won't have been derived from GoogleMaps like the current dataset is. But, can it be done? I'm not sure this is documented or encouraged, thus I'm not sure the URL will stay accessible, but $ curl -d 'Start = -1.267,51.75034 End = -1.260884,51.753598 Via = lang = de distunit = KM routepref=BicycleavoidAreas=useTMC=noMotorways=instructions=true' http://data.giub.uni-bonn.de/openrouteservice/php/OpenLSRS_DetermineRoute.php | grep TotalTime yields the time that ORS calculates for the bicycle route from Start to End. Do this for each pair of college locations, and you've got yourself a complete graph you can solve traveling salesman for. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] live editing and conflict management
Hi Richard, please don't mistake me for one of the German Potlatch haters. I rather like it, and acknowledge the great work you've done. But: On Dec 16, 2008, at 10:24, Richard Fairhurst wrote: management. Conflict management isn't really an issue when you're redownloading from the server every minute or so; it is when you're clicking 'Upload' after an hour's editing. I don't think that's true: It's a problem as soon as two people are editing simultaneously in the same area. I ran into the problem a bit recently when giving phone support to somebody new to OSM, both editing the same area in Potlatch. Mostly, I took care to use Play mode, but when not, it was quite hard to resist fixing one or the other small error. Another situation where this is quite likely to show up is when two people try to fix OpenStreetBugs as soon as they appear. With API 0.6, we'll notice when this happens, right? By the way, has anyone successfully disabled the Potlatch welcome dialog? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] mkgmap makes routable garmin maps
On Dec 12, 2008, at 16:43, Andy Allan wrote: On Sat, Dec 6, 2008 at 4:12 PM, Robert Vollmert rvollmert-li...@gmx.net wrote: Hi all, there seem to be a few Garmin users around here. If you'd like to give routable OSM-derived maps a try, there's some instructions on the wiki at http://wiki.openstreetmap.org/wiki/Mkgmap/routing . The support is still quite incomplete both because the Garmin format isn't completely understood and because of bugs. More the latter, probably. Help on either topic would be appreciated. Awesome. Are there any docs anywhere being made for the routing format similar to the venerable http://sourceforge.net/projects/garmin-img ? There's some documentation on the NOD file in http://svn.parabola.me.uk/display/trunk/doc/nod.txt . That covers the routing graph. I don't think the NET file is currently documented, though i may be missing something. Apart from that, it's in the code for the moment. There's also libgarmin.sourceforge.net (IMG reader for navit) and its wiki. Also, when I was putting it together yesterday I saw osm2mp.pl suggesting it was dealing with turn restrictions (although on reading the code, it's not exactly sophisticated logic!). But when I was playing with it last night my eTrex it was ignoring them. Am I expecting too much, or would you expect turn restrictions to work? mkgmap doesn't write restrictions for the moment. For simple from-via- to restrictions, this is a matter of writing the code, since that part is understood. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Updated view of 'A year of edits on OSM' and also Santa's Routes!
On Dec 9, 2008, at 21:58, Ed Loach wrote: I think the way heading north from Finland may actually pass straight through Finland and start somewhere near Riga. But I can’t find it using Mapnik or Potlatch. http://openstreetmap.org/browse/way/27611977 Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Updated view of 'A year of edits on OSM' and also Santa's Routes!
(to the list also) On Dec 9, 2008, at 22:29, Scott Atwood wrote: On Tue, Dec 9, 2008 at 1:19 PM, Robert Vollmert [EMAIL PROTECTED] wrote: On Dec 9, 2008, at 21:58, Ed Loach wrote: I think the way heading north from Finland may actually pass straight through Finland and start somewhere near Riga. But I can't find it using Mapnik or Potlatch. http://openstreetmap.org/browse/way/27611977 Did you find this with a brute force search? Or did you have a more elegant method? The high-res version of the Europe image pointed somewhere west of Riga. Then the data layer. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] mkgmap makes routable garmin maps
On Dec 7, 2008, at 17:39, Andy Street wrote: I had a go at producing a routable map for my local area but whenever I transfer it to my eTrex Vista HCX it always routes on the in-built basemap no matter what I try. It seems you're doing everything right. The gmapsupp.img you generated looks fine to me. Some routing data was generated, at least. Probably the data we're writing is just wrong in your case. One thing you could try is to run mkgmap with assertions and logging: java -enableassertions -Dlog.config=resources/logging.properties -jar mkgmap.jar is what I use, but it's possible you need to be in the dist directory for java to find the logging.properties file. Thanks for your feedback! Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
[OSM-talk] mkgmap makes routable garmin maps
Hi all, there seem to be a few Garmin users around here. If you'd like to give routable OSM-derived maps a try, there's some instructions on the wiki at http://wiki.openstreetmap.org/wiki/Mkgmap/routing . The support is still quite incomplete both because the Garmin format isn't completely understood and because of bugs. More the latter, probably. Help on either topic would be appreciated. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit war on the wiki map features
On Dec 1, 2008, at 11:02, Bernhard Zwischenbrugger wrote: In Vienna we have an event called Friday Night Skating. Every week about 1000 Inline Skater meet at 10pm and skate on normal roads. The police blocks all the roads an it is possible to skate on roads that are for normal for cars only. The route is about 15 to 25 km. To plan an event like this is not easy. It should be a different route every week. If it's combined with sightseeing it's optimal. There are similar events in many cities like Paris, Munich,... sometimes with much more skaters. For beginners the road surface is very important. It should be possible to plan a Friday Night Skate route with data from OSM. If we have a tag skate:xy bicycle:xy people think it's allowed to go by bike or inline skates on this roads - but it isn't. Since smoothness=good/excellent handles this fine, I'd suggest to just use it. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit war on the wiki map features
2008/12/1 Richard Fairhurst [EMAIL PROTECTED] Personally I believe the easiest and most flexible thing is just to extend the access tags: bicycle=no|yes|difficult|unsuitable so you'd get highway=bridleway foot=yes (permitted, no problem) bicycle:racer=unsuitable (permitted but not practical) bicycle:hybrid=difficult (permitted but challenging) bicycle:mtb=yes (permitted, no problem) The obvious problem with this is the massive redundancy. You need to tag for every possible form of transport, or infer suitability for something exotic from the provided suitabilities. On Dec 1, 2008, at 11:09, Douglas Furlong wrote: This feels like a far more suitable solution, than smoothness (and Ice rink is smooth, but I doubt a racing bike would have much fun on it!). Hurray for absurd arguments. Obviously, 'slippery=yes' is implied on ice rinks. I do wonder why people are always jumping on the corner cases to discredit smoothness=*. Would one of you that think smoothness is worse than nothing care to comment on the definition by example I proposed in http://lists.openstreetmap.org/pipermail/talk/2008-November/031779.html ? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit war on the wiki map features
On Dec 1, 2008, at 11:15, Douglas Furlong wrote: If this is an argument in favour of smoothness, then you would run in to exactly the same problem (just not as fine grained). If a user see's a road as being tagged as smooth, then they'd think that they could roller blade on it, which apparently they are not allowed to. Here, we run in to a problem where suitability and permissibility are not going along with each other. Thus, we put permissibility in one key (skate=yes/permissive/no/...), physical suitability in others (surface=*, smoothness=*, steepness=*, slippery=yes/no). And we don't mix them all in one key. Smoothness is completely independent of access rights. What gave you the idea it wasn't? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] My data got deleted
On Nov 30, 2008, at 17:36, [EMAIL PROTECTED] wrote: One of the ways that I saw had removed a street is here. http://www.openstreetmap.org/?lat=28.57944lon=77.2855zoom=16layers=B000FTF You're right, they were mapped and deleted: http://openstreetmap.org/browse/way/25524666/history Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] how to map speed breakers/bumper
On Nov 30, 2008, at 06:22, ビカス ヤダワ (vikas yadav) wrote: Here in the streets of New Delhi and Gurgaon, we have lots of speed breakers (or bumps - These are small, raised portions on the road to slow down a fast vehicle, eg, passing through a living street entering a faster highway.) Some are standard height while some are ugly and speed has to be reduced to dead slow to prevent it hitting the vehicle. I searched the wiki but could not find any markers for these. How do I mark these on the roads? Does this fit? http://wiki.openstreetmap.org/index.php/Key%3atraffic_calming Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] My data got deleted
On Nov 30, 2008, at 06:16, ビカス ヤダワ (vikas yadav) wrote: I checked one of the sections that I had surveyed, mapped and uploaded two months back. Suddenly when I went through that street yesterday, I could not see my edits anymore. It was a plain single street without a one way property while I had neatly placed both one way streets paralell and also given the right points of U turns and traffic signals. I wish to know who delete this data from the server or why is it gone. Please suggest. I'd start by looking at the current version in the data browser (view the street on the main map, and check the Data box in the layer selection thing). Or you can try hitting 'u' in Potlatch to make it show deleted ways, then look at the history of a deleted way using 'h'. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] smoothness
This turned out rather long. Summary: smoothness is a useful tag, though the wiki definition may be lacking. Thanks for reading. On Nov 27, 2008, at 11:27, Dave Stubbs wrote: The table is full of such subjective assessments: can I roller blade on it. No, I can't. It doesn't help that I can't roller blade at all. So sure, if the tag was, is it possible to roller blade down this road assuming a skill level of a Grade III Roller Blading Proficiency Award? then it might have a point, but it's not. How about: Would an average roller blader like to use this? Personally, I couldn't care less for an absolutely precise and objective definition for a tag. If the description gives a good idea of how to use the tag in most situations, that's perfect. There'll always be corner cases. It's quite possible that people have tried too hard to define smoothness objectively (and have claimed too strongly that it's even possible to define it 100% precisely). Here's how I see smoothness (on the smooth side of things, I don't care about things beyond bad). If the people that formulated the smoothness proposal disagree, I guess that proves your point. excellent: this is what well paved new cycle ways tend to be like; some fine type of asphalt; good for roller-skating, a pleasure on a road bike good: your typical road in good state; a cycleway like above but with some small bumps from tree roots because they didn't care to put a proper foundation (?) underneath; a high-quality non-paved footway in a park intermediate: a road the has been worn down and could use a new cover, some unevenness from heavy traffic; motorway made of slabs of concrete with annoying bumps when passing to a new slab (you'd really want to use the fast lane exclusively if that's recently been repaved); lots of tree root induced bumps on a cycleway; a footway in a park with coarser gravel or uneven enough that there'll be puddles when it rains; high-quality cobblestoned road (small stones with flat surface, or perhaps some filling of the gaps); the average motorist wouldn't mind, the average cyclist wouldn't complain (at least not loudly), you wouldn't want to skate here. Anything worse, I'd tag bad for now and put a note/fixme in so someone else can say how bad it really is. I think smoothness fits the above distinctions quite well. Together with surface=paved/unpaved, it should provide most information about a way's surface that users of wheeled (on-road) vehicles would like to have when deciding which road to choose. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Edit war on the wiki map features
On Nov 25, 2008, at 17:16, Andy Allan wrote: Go ChrisCF is all I can say - I'd rather that the wiki was a meritocracy With those in charge that show most determination in an edit war? than ochlocracy and I'm flabbergasted that such ill-conceived tagging is now an acceptable norm. Personally, I don't see what's wrong with distinguishing between a normal paved road and one that's suitable for inline skating with smoothness=good vs. smoothness=excellent. Or between roughly cobblestoned road and the one most people wouldn't mind riding a bicycle on with smoothness=bad vs. smoothness=intermediate. What's your way to tag surface quality? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] [tagging] Feature Proposal - Voting - (service=parking aisle)
On Aug 5, 2008, at 08:53, Thorsten Feles wrote: Lennard voor den Dag schrieb: That earlier proposal was highway=parking_aisle, not service=parking_aisle (with highway=service) as it stands now, IIRC. But its not getting better, the service key is already in use by the railways guys. Even a path is called highway in osm, why change it ? I don't see the conflict: For railways, service=* distinguishes between different types of service lines, for highways it distinguishes between different type of service roads. It's not like you're going to tag a way with both, is it? By the way, highway:service serves it well ... And railway:service for the other use? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] House numbers... One more suggestion
On Jul 29, 2008, at 00:40, Karl Newman wrote: Don't overestimate the usage of the current data scheme, though. The Germans are prolific mappers, but I would be surprised if there are even a few thousand addresses entered in the current format, if that. According to tagwatch, around 2 in Europe, of which 7200 in Germany. http://tagwatch.stoecker.eu/Europe/En/top_undocumented_keys.html http://tagwatch.stoecker.eu/Germany/En/top_undocumented_keys.html I wonder where the other 1 are? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Tagwatch and osmxapi-links
On Jul 29, 2008, at 14:25, Stefan Neufeind wrote: in tagwatch, e.g. at http://tagwatch.stoecker.eu/Europe/En/tags.html I see entries that link to adresses like http://osmxapi.hypercube.telascience.org/api/0.5/*%5Bvalue=LIDL%5D That page takes quite a while to load but then returns with an empty document. Is that a timeout in the api or so? How do people manage to use tagwatch to find specific tagged entries (e.g. which need fixing)? Those links point to XML documents that your browser may (Firefox) or may not (Safari) display. Best to download link location and save to whatever.osm. Would make sense to provide some kind of style information to make any browser that can handle XML display the data nicely? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Namefinder priorities
On Jul 25, 2008, at 13:13, David Earl wrote: I've thought about not tagging for the rendering (and name finder is a kind of renderer), but there isn't a simple algorithmic solution. While it might be possible to do some analysis of connections to try to determine when two things are part of the same, this is a very expensive calculation in what is already a slow process, which makes practicality difficult, though not impossible; but in the case of indicating the preferred branch, I think this is a matter of judgement and can't be done algorithmically from the information that's already there (I could do some guesses based on length of number of nodes perhaps, but it would always be heuristic). Doing that might be useful anyway, but I don't think it will provide a perfect solution in all cases. An alternative to search=yes that might be more generally useful is to group parts of a street into a relation (see http://wiki.openstreetmap.org/index.php?title=Relation:street) . That would be an object you could point to. It would also be possible to mark spurs with role=spur if that would help. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] aerialway cable_car not available in mapnik
On Jul 22, 2008, at 09:41, mariner wrote: The map feature aerialway=cable_car isn't rendered in mapnik. Could someone fix this? Here a link to the problem: http://openstreetmap.org/?lat=46.89994lon=8.50536zoom=15layers=B00FTF If you'd like something to be rendered, I think it's best to file a bug report on http://trac.openstreetmap.org. Your normal login/ password will work. Component slippy_map in this case. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] OSM maps in 3D
On Jul 18, 2008, at 09:27, elvin ibbotson wrote: Very nice but it needs DirectX. I cut my map programming teeth on a viewer for British OS maps which uses Java 3D (http://britain.poco.org.uk/desktop.html ). I can’t share it because of copyright restrictions on the maps, but the principle would apply to any map source including OSM. Why not use Java instead of Microsoft stuff then it would run on anything. There’s an awful lot of us using Linux or Macs - anything but Windows!. I like the idea of Kosmos but - MS .net!! If you had tried, you'd know that Kosmos does in fact run on Linux with Mono. This one doesn't since there's no DirectX support in Mono, but I'm sure it'd be possible to port it to something like http://axiomengine.sourceforge.net/ . In my opinion, a Windows only project with source available is worth a lot more than some closed Java thing. On that note, how can the license on the maps keep you from releasing the source code to your viewer? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] User seams to add trash
On Jul 8, 2008, at 02:19, wer-ist-roger wrote: So befor comming to the wrong conclusions I like to contact that person but I have no idea how to search for a user so that I can write a message (people search on OSM is realy bad, sorry) Just in case it is a violating person I like to ask if somebody know MapDeliverer. http://www.openstreetmap.org/user/MapDeliverer Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] www.OpenRouteService.org now supports Bicycle Routing with OSM Data
Hello, very cool indeed. Very much so. - work in progress: tracktype As a cyclist I'd appreciate it if track with no additional info, track with surface=paced, track with surface=gravel, and tracks of type1, 2 and probably 3 would be taken into account. This may be unrealistic, but maybe you can include tracktype, surface and the proposed key smoothness in your database, and allow the user to select allowed values for each? Or even weights for each combination? There could be default profiles for road bike or city bike or mountain bike. Of course there are other combinations and possbilities and we are evaluating those, but this shall serve as a first start ... What immediately comes to my mind as a feature request for (much :) later versions are downloading the caclulated route as gpx track or (optional) route and via-points I agree. I just used the service to create a route for me and was a little disappointed I couldn't download it as gpx. One other problem I had is that it seems to have a preference for routing against the oneway direction on dual carriage ways. Or does it just disregard oneway? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Walking routes and OSM (again)
On Jun 27, 2008, at 16:55, [EMAIL PROTECTED] wrote: a) Get the user to click each way in turn on the slippy map. Each way then gets highlighted (possible via OpenLayers Vector layer). When finished, user clicks Done and can add any further comments. This should be fairly easy to implement - and some of the code (detecting the closest way to a mouse click) has already been implemented, both client and server side. I prefer this approach. My only concern is that when making relations and routes, often one needs to split a way. Is there any good UI for doing this for someone who doesn't know OSM? There is no reason as to why the overlay couldn't automatically split the way on it's own vector overlay and then when the route is confirmed that the split be transfered to OSM along with the relation information. An alternative would be to specify a part of a way by starting and end node (a subway, compare http://wiki.openstreetmap.org/index.php/Relations/Proposed/Segmented_Tag) . If the routes are to be stored in OSM as relations, this wouldn't be a such good idea, but assuming you're storing them externally and didn't want to write to the OSM database, it's an option. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] [tagging] Graves?
On Jun 24, 2008, at 09:26, Matthias Urlichs wrote: Quick question: How do you tag individual graves (either part of a cemetery, or not)? Unless I'm missing obvious ideas, not many people have tagged single graves according to tagwatch. Obvious choices: historic=grave, historic=tomb or if you want something for not-necessarily-historically-significant graves, perhaps burial=grave, burial=mass_grave, burial=mausoleum, burial=pyramid Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] pronunciation tag
Disclaimer: based on a little of web research; I have no particular knowledge of linguistics or speech synthesis. On Jun 24, 2008, at 03:54, SteveC wrote: On 23 Jun 2008, at 18:52, Lauri Hahne wrote: I think some standard form should be used if we ever want to do something like this. Although IPA is the official standard, it isn't very computer or user friendly. Therefore I think something like SAMPA, MRPA or X-SAMPA should be used. These are used to some extend among linguistics and are all based on ASCII. These would also relieve the pain of trying to figure out what something would be in phonetic pseudo-english. can you summarise these with examples? supercalifragilisticexpialidocious IPA: /ˌsuːpɚˌkælɪˌfrædʒəlˌɪstɪkˌɛkspiːˌælɪ ˈdoʊʃəs/ CXS: /su:[EMAIL PROTECTED]klIfr[EMAIL PROTECTED]IstIkEkspi:lI'[EMAIL PROTECTED]/ CXS is basically X-SAMPA, which is basically an ASCII-encoding of IPA. Since we do unicode, I'd think we should rather go with IPA. See http://www.theiling.de/ipa/ for an online converter. I didn't find any speech synthesis package that does IPA directly, though. Festival's Sable markup language http://www.cstr.ed.ac.uk/projects/festival/manual/festival_10.html#SEC33 provides for IPA, though festival doesn't implement this. It does allow e.g. PRON SUB=toe maa toetomato/PRON. A possible alternative is the free-as-in-beer mbrola http://tcts.fpms.ac.be/synthesis/mbrola/ . It's a speech synthesis backend based on diphones (two halves of phones). Its input format appears to be SAMPA plus additional data. There's still some language dependency in there, though. Espeak http://espeak.sourceforge.net/ can target mbrola, perhaps IPA could be added as a language? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] pronunciation tag
On Jun 24, 2008, at 11:02, Robert Vollmert wrote: A possible alternative is the free-as-in-beer mbrola http://tcts.fpms.ac.be/synthesis/mbrola/ . It's a speech synthesis backend based on diphones (two halves of phones). Its input format appears to be SAMPA plus additional data. http://www.acm.org/crossroads/xrds9-3/french.html is a nice introduction. But I really like the idea of local contributors recording the names. It seems to be feasible in terms of storage, even: At around 50kB a name, all names in Germany would take around 50GB at the moment. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] [tagging] - Alpine Hut
On Jun 13, 2008, at 18:00, Alexander Zatko wrote: I found out that there is a tag tourism=chalet which seems to be quite a similar concept to alpine hut. Given that the chalet tag is already approved, I will start using it, but am not sure what to do with the proposal I created for Alpine Hut. To me, tourism=chalet is a building that you'd rent with a group or family for holidays, usually in or close to a village. An alpine hut on the other hand is usually higher up in the mountains and further from civilization, which operates more like a youth hostel. So I think your proposal should stand. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
[OSM-talk] KML tiles
Hello, the attached KML file links to a very rough proof-of-concept set of KML tiles of OSM road data. There's several issues remaining (more on that below). What do people think? Is this worth doing on a larger scale? Is it evil? Obviously, this shouldn't be used for mapping. The aim is to use the data -- for example, I like to measure distances of runs using Google Earth with OSM data (which requires tracks and footways, not roads). It should also be a good advertisement for OSM to the Google Earth community. Perhaps this approach would also be useful to the Marble guys for displaying vector data -- I hear Marble supports KML. The main issues at the moment are gaps at tile boundaries. This should be easy to fix using a better tile cutting method (osmosis' completeWays or the tile data server, once it's running). Furthermore, there's some network links missing between adjacent tiles which may cause problems when panning. The choice of tiling isn't useful for a 3d earth. Low zoom data should be simplified. Some ways disappear from the middle if you zoom in. It's too slow and not smooth enough. How it works: For each standard z/x/y-tile, cut out the relevant OSM- data and convert to KML using osmexport from osmlib, then massage the result to contain region information and links to the four subtiles. Cheers Robert tiles.kml Description: application/vnd.google-earth.kml ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] JOSM cut and paste between layers results in position change
On Jun 11, 2008, at 12:05, David Earl wrote: Maybe there should be an option to move or copy to another layer in the same location? That would certainly be pretty straightforward to do. Would people prefer 1. an additional paste operation (Paste in same place / Duplicate in same place) 2. a Duplicate Layer option on the context menu of the layer 3. A Preferences option to change the behaviour of paste Either 1 or 3 would be great. 2 would make it hard to take just a part of a layer (duplicate, cut, delete rest, duplicate). Please close http://josm.openstreetmap.de/ticket/587 if you implement this, and thanks in advance! Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] KML tiles
On Jun 11, 2008, at 17:30, Nic Roets wrote: The current Google Earth maps for South Africa are supplied by Tracks4Africa and AND. Both of them only show major roads and are useless for routing. In contrast, OSM has all the roads and many footways for Pretoria (Cape Town and Johannesburg nearly done). If I post a KML to a 4x4 newsgroup, chances are that a few T4A contributors will be impressed and start sharing their tracklogs with us. For just one city, you should be able to create a single usable KML with osmexport right now. It shouldn't get larger than the corresponding planet extract. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Nested areas
On Jun 6, 2008, at 09:09, spaetz wrote: But if you tag a river universally over quite a bit with layer=-1 just for the fun of it, as was in the original example, then this looks weird. And osmarender is right to make it look weird, isn't it? I think this can be correct, if say a river runs in a river bed that's dug into to the landscape. So the surface of the river is always say 2m below the surrounding areas. Bridges over the river are flat. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] osmarenderer issues
On May 31, 2008, at 13:46, Sven Grüner wrote: The curving doesn't happen in Osmarender but is applied to the final SVG, that's why it's not possible to differentiate between streets, rivers and buildings, they're all just lines. The responsible script is lines2curves.pl: http://trac.openstreetmap.org/browser/applications/rendering/tilesAtHome/lines2curves.pl#L421 Ah, thanks! At least the comments are strange, since usually 180 degrees is pi, and 90 degrees is 0.5*pi. So maybe it's not working quite as intended. If nobody else gets to it first, I'll have a closer look later. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Mulltipolygons and Mapnik
On May 31, 2008, at 19:27, Chris Hill wrote: I've already changed the wiki to match this situation. This is the way it was, so we are just back to the same position before it was changed (in my view) erroneously. I also thing this makes logical sense too. The outer marks the edge of the water and so does the inner. There is a short discuusion on the wiki talk page. As far as I can tell, the main argument for this reversion is that's what the renderers want. I wonder why those who claim the change was wrong didn't speak up in the lengthy discussions in March? (Starting at http://lists.openstreetmap.org/pipermail/talk/2008-March/023876.html and http://lists.openstreetmap.org/pipermail/dev/2008-March/009345.html , and I do remember one opposing opinion which I can't find.) Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] How can wide intersections be mapped?
On May 23, 2008, at 02:01, David Muir Sharnoff wrote: I'm trying to figure out how to represent an intersection that is very wide: a traffic circle could placed in the middle without moving the edges. I've tried adding extra ways for various ways across the expanse but it doesn't look right. Suggestions? http://www.openstreetmap.org/?lat=37.845598lon=-122.236367zoom=18layers=B0FT The area=yes highway=residential that's there now looks like a good solution. It's not clear it will render well, though. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Lakes and relations, what did I break?
On Apr 17, 2008, at 01:26, Dermot McNally wrote: To anyone who can show me what I broke: http://geo.topf.org/comparison/index.html?mt0=mapnikmt1=tahx=971y=657z=11 [...] So I decided to fix them. Rather than follow my usual practice of representing islands in lakes as land at layer 1 I decided to use relations to model them as holes in the water. Something has gone wrong, though, as can be seen from the rendered output, but my data looks clean to me in the editor and I think I've followed the guidelines on polygons with holes. (Though the lakes do have quite a few nodes, so maybe I'm just putting too much load on the [EMAIL PROTECTED] clients). Perhaps it's because the holes aren't oriented counter-clockwise? It's possible that osmarender still relies on orientation for rendering multipolygons. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Changes in Relations do not upload: 500 Internal Server Errror
On Apr 12, 2008, at 18:57, Rainer Dorsch wrote: I can upload normal changes to OSM, but I cannot upload changes in relations. I keep getting Error while parsing: An error occured: 500 Internal Server Error Is that a know problem? I've had this error message in JOSM when I tried uploading a relation with duplicate members. Not sure if this has been fixed, consider updating your version of JOSM. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Cycleway byway
On Apr 11, 2008, at 13:08, Chris Hill wrote: The national Byway cycle route passes close to my home, so I'd like to add it to the map. The Wiki [1] suggests that I add to the relation 9327. How do I do this when the existing parts of the relation are far away so I cannot get the existing plus the new on either JOSM or Potlatch. Once I have one local way in the relation it should be easy to add others. What goes wrong if you download a small area containing an existing part of the byway in addition to your home area in JOSM? Alternatively, you could save http://api.openstreetmap.org/api/0.5/ relation/9327/full as byway.osm and try loading that file into JOSM. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] linz dataset for nz - attribution methods summary
On Apr 2, 2008, at 13:08, Cartinus wrote: On Wednesday 02 April 2008 12:02:49 Robin Paulson wrote: true, but as i suggested in a previous mail, i'm not sure why someone would need to do this. if a user is importing another dataset which needs attribution, they would likely be someone responsible/trusted, and given temporary rights to add to the 'attributions' tag. who else would need to edit it? i can't see a situation where anyone would feel a need to delete and then re-create? what would they gain, in real terms? You probably can't see the need because you are currently not editing in an area that is peppered with nodes that contain a source=AND tag. Why would a node that I moved a bit and changed into a highway=bus_stop name=* still be attributed to wherever it was imported from first? Why would a lake where I changed the east bank according to Yahoo aerial images or my own GPS tracks still be attributed to wherever it was imported from first? Etc., etc., etc. I may be missing something, but why would we need to introduce a read- only attribution tag if we already have it? It's the source tag of the first version of an object, in http://api.openstreetmap.org/api/0.5/objtype/id/history Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Mapping Mottram and Tintwistle proposed bypass
On Mar 29, 2008, at 09:39, Andy Robinson wrote: On 28/03/2008, Peter Miller [EMAIL PROTECTED] wrote: 1) How does one tag something that is being considered seriously (such as the Mottram Tintwistle bypass), but which may well never get built? I think I will just put the estimated build date given by the highways agency for now. (I will also continue to use the tunnel trick to get it to render in the mean time). On the basis that we only put data into the project relating to physical objects I dont think we should put any items in that might get built. This should really go into a separate layer... If consensus is that this data should not go in the database, I'd suggest creating it in JOSM and saving to a .osm file. Then if you want to create a map that shows the proposed highway, you could include this file in the input to the renderer. Personally, I don't see any harm in adding the data to the OSM database for now -- it certainly shouldn't get rendered on the main maps, but that can be achieved by appropriate tagging. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] [tagging] RFC: railway=incline
On Mar 27, 2008, at 00:56, Alex L. Mauer wrote: Sven Geggus wrote: 1.) adding railway=funicular and rag=yes for non funicular incline railways 2.) adding railway=incline and an additional tags for types of incline railways (funicular,rag, ...) I think it is important to be able to mark a railway as an incline railway of some sort, without having to specify what drive mechanism it uses. If that can fit into the first suggestion, I'm all for it. incline=yes ? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] [tagging] RFC: railway=incline
On Mar 26, 2008, at 00:24, Alex Mauer wrote: Sven Geggus wrote: Alex Mauer [EMAIL PROTECTED] wrote: I've written up a proposal here: http://wiki.openstreetmap.org/index.php/Proposed_features/Incline_railway I don't like this! It is often impossible to differeciate between incline railways and ordinary railways, esecially in countries like Switzerland. If it's impossible to differentiate, then how do you know how to tag it? So far, different railway tags have described different networks. A railway=rail doesn't suddenly change into a a railway=subway. But a part of a railway might go over a bridge (bridge=yes) or have a rack (rack=yes for that segment). Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Tourist/Leisure Trails
Hello, On Mar 17, 2008, at 19:28, [EMAIL PROTECTED] wrote: Is there any way of marking a 'trail', where a marked route which may exist on other ways in part or as a whole? This should probably be done using a route relation: http://wiki.openstreetmap.org/index.php/Relations/Routes Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] highway=living_street not rendered?
On Mar 15, 2008, at 05:04, Inge Wallin wrote: I am a very new user that has started to map up the village in Sweden where I live. However, I have found something strange. Look at the slippy map, and search for Ljungsbro. Notice the to streets Kohagsvägen and Ugglebovägen southeast of the center. Then press the 'edit'. What you will see is that there are two small stumps of streets with tags highway=living_street. But they are not rendered in the real map, neither with Mapnik, nor with Osmarender. Is this a bug or am I missing something? I just filed bug reports against Mapnik and Osmarender yesterday (tickets 740, 741). It seems this was forgotten when the tag was moved to Map features. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] area_with_holes as alternative to multipolygon relation
On Mar 13, 2008, at 23:27, Frederik Ramm wrote: True, but since there can only be one circumference of a polygon, could we not specify that if more than one outer ways exist in a multipoly relation, these will be merged to make the circumference? That would be very confusing. I'd expect outer and inner ways to be the same kind of way (closed). Several outer_part ways instead of one outer would be better, but that's still not good: What if an inner way is also so long you'd like to split it up? Basically, I think using several open outer ways like this is abusing the fact that there's just one outer way to start with. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] area_with_holes as alternative to multipolygon relation
On Mar 12, 2008, at 11:47, Dirk-Lüder Kreie wrote: Robert Vollmert schrieb: Certainly the multipolygons which are just a polygon with several ways making up the border are broken and should be fixed. I hope to get a handle on these. What do we do with ways that get excessively long if we combine the polygon border? (some really big forests and lakes come to mind) I don't know -- create a new relation type specifically for this? As far as I can tell, the fact that putting multiple border ways in a multipolygon relation may have worked is due to the implementation of multipolygons in the renderers. But this isn't what the relation is meant to do. There could be a general purpose relation for collecting several ways that meet at the ends into one virtual longer way. Say type=long_way, with roles part_1,...,part_n (relation members aren't ordered). And if this forest has a hole, put the long_way relation in a multipolygon relation. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] areas without holes
On Mar 11, 2008, at 23:02, Jon Burgess wrote: I've just fixed another 248 which were shown up by locating all polygons output by the old osm2pgsql algorithm whose outer ring had more than a single way. I've gone through relations with id below 3000 now, joining up the easy outer rings and setting inner/outer roles where possible (but not removing tags from inner ways yet). There's 1122 multipolygons left, of which 956 now have a member of role=outer. That should be enough to switch to role-based rendering. What do you think? I think the remaining ones are mostly on the Thames (multipolygons of two opposing waterway=riverbank's that are also natural=coastline -- I'm not sure how the coastline checker would handle closing these up). There's also a couple of broken multipolygons from the AND imports, e.g. 1445. I've fixed a few, the rest could be left for JOSM +validator. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] area_with_holes as alternative to multipolygon relation
On Mar 10, 2008, at 22:51, Igor Brejc wrote: I too am a little bit confused: now the whole issue basically comes down to renaming the relation from multipolygon to area_with_holes. But the inital proposal had some other features, like using the inner polygons' tags to render the inner polygons content directly (not as holes) + abandoning clockwise/anticlockwise approach. I still want all of those features. Is the renaming really necessary? I think the existing renderers already know how to draw multipolygons even without outer/inner specification. I can't speak for osmarender, but as I understand it, osm2pgsql only does this correctly on the assumption that ways are oriented in a specific way, and inner ways are tagged the same as outer ways. This is a prime example of tagging for the renderer. That may be what we want but a significant proportion of the existing data also has: 1) an outer ring which is not a single closed way 2) the same tags on all of the ways Shouldn't this be handled just by fixing the inconsistent data? Certainly the multipolygons which are just a polygon with several ways making up the border are broken and should be fixed. I hope to get a handle on these. I'm not so sure what to do about pairs of opposite riverbanks. These are also clearly not multipolygons and just abuse the way the renderers work... Probably best to just close them up and tag the fact, so they can be fixed properly once people agree on [[Relations/ Proposed/Rivers]] or similar. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] what value for access in parks etc that you have to pay to enter?
On Mar 11, 2008, at 15:39, David Ebling wrote: This question has been bugging me for a while. What access= tag should you usefor footpaths and other ways that are within a park that you must pay to enter. They are not really private, but neither are they free public access. If there's some kind of restriction like this, I tend towards access=restricted. Perhaps access=fee? From the perspective of having to tag all ways in a park like this, we should really be able to say all footways within this park default to access=whatever and surface=gravel... Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] area_with_holes as alternative to multipolygon relation
On Mar 10, 2008, at 22:43, Jon Burgess wrote: The original multipolygons created by the conversion above all had the same tags and no defined roles. Does osm2pgsql really require the same tags on all ways? The comments in the code seem to say it's collecting tags from all member ways, in which case removing tags from the inner ways shouldn't break things. As I mentioned before, I am not against change. I'd be really happy to see this rationalisation occur. It will eventually result in a cleaner data model and less code in osm2pgsql. The transition period does risk breaking what we have already and this is what I'm trying to avoid. I should confess at this point that before the discussion resumed this afternoon I fixed multipolygon relations with id 1000. So my question above will be answered soon enough... Would it be difficult to handle relations differently based on whether or not there are members with non-empty roles? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
[OSM-talk] fixing multipolygons
Hello,
I've attached an ugly python script that does some manipulations to
relations in an OSM file it reads from stdin. In particular, for
relations that aren't degenerate, it puts role=outer on the largest
polygon and role=inner on all others. It also removes tags from inner
ways that are the same as on the outer way.
It only works on multipolygons that consist of closed ways. There's a
lot of multipolygons with two opposing riverbanks in the database.
Also some worse ones like http://api.openstreetmap.org/api/0.5/
relation/240. These really aren't multipolygons and should be
relations of a different type -- should I modify them to
type=multi_way_area or something?
There's both relations with less than two members, and ways with less
than two nodes in here. Ok if I delete these?
Here's some statistics for the relations with id up to 500:
393 relations
235 with non-closed ways
53 with less than two members
32 containing ways with less than two nodes
56 modified relations
122 modified ways
Cheers
Robert
from osmxml import *
nodes = {}
ways = {}
relations = {}
def area(way):
global nodes
nds = way.nodes
assert len(nds) 2, less than two nodes in way %d % way.id
assert nds[0] == nds[-1], trying to compute area for non-closed way: %d % way.id
coords = [ (nodes[id].lat,nodes[id].lon) for id in nds ]
area = 0
for i in range(1, len(coords)):
area += coords[i][0] * coords[i-1][1] - coords[i-1][0] * coords[i][1]
return 0.5 * abs(area)
from xml.dom import minidom
def splitosm(dom):
assert dom.documentElement.nodeName == u'osm'
global nodes
global ways
global relations
nodes = {}
ways = {}
relations = {}
for elem in dom.documentElement.childNodes:
if elem.nodeType == minidom.Element.nodeType:
id = int(elem.getAttribute(id))
if elem.nodeName == 'node':
obj = Node(elem)
nodes[obj.id] = obj
elif elem.nodeName == 'way':
obj = Way(elem)
ways[obj.id] = obj
elif elem.nodeName == 'relation':
obj = Relation(elem)
relations[obj.id] = obj
def fix(filein, fileout):
dom = minidom.parse(filein)
splitosm(dom)
global deleted_tags
deleted_tags = {}
for r in relations.values():
try:
type = r[type]
except KeyError:
sys.stderr.write(rel %d: no type, delete?\n % r.id)
continue
if r[type] == multipolygon:
fixmultipolygon(r)
for r in relations.values():
if r._modified:
r.action = modify
for w in ways.values():
if w._modified:
w.action = modify
sys.stderr.write(%s\n % str(deleted_tags))
fileout.write(dom.toxml(utf-8))
# some more things we could check:
# * tags on the relation
# * member ways with roles other than , inner, outer
# tags that should not be removed from inner polygons
ignored_tags = [created_by]
def fixmultipolygon(rel):
assert rel[type] == multipolygon
if len(rel.members) 2:
sys.stderr.write(rel %d: multipolygon with less than two members, delete?\n % rel.id)
return
mnodes = [ m for m in rel.members if m.type == 'node' ]
if mnodes:
sys.stderr.write(rel %d: multipolygon with nodes: %s\n % (rel.id, mnodes))
return
mrels = [ m for m in rel.members if m.type == 'relation' ]
if mrels:
sys.stderr.write(rel %d: multipolygon with relations: %s\n % (rel.id, mrels))
return
mways = [ m for m in rel.members if m.type == 'way' ]
rempty = [ m for m in mways if m.role == '' ]
rinner = [ m for m in mways if m.role == 'inner' ]
router = [ m for m in mways if m.role == 'outer' ]
if len(router) 1:
sys.stderr.write(rel %d: more than one outer way\n % rel.id)
return
if len(router) == 1:
if rempty:
log(one outer, some empty)
for m in rempty:
log('setting role: inner')
m.role = 'inner'
if len(router) == 0:
try:
byarea = [ (m, area(ways[m.ref])) for m in rempty ]
except AssertionError, e:
sys.stderr.write(rel %d: %s\n % (rel.id,str(e)))
return
byarea.sort(lambda x, y: cmp(y[1], x[1]))
rempty = [ m for (m,s) in byarea ]
rempty[0].role = 'outer'
for m in rempty[1:]:
m.role = 'inner'
# now delete outer tags from inner ways
rempty = [ m for m in mways if m.role == '' ]
rinner = [ m for m in mways if m.role == 'inner' ]
router = [ m for m in mways if m.role == 'outer' ]
assert not rempty
assert len(router) == 1
outer = ways[router[0].ref]
rkeys = [ k for k in outer.keys() if k not in ignored_tags ]
for m in rinner:
inner = ways[m.ref]
for k in rkeys:
if inner.get(k, ) == outer[k]:
del inner[k]
try:
deleted_tags[k] += 1
except:
deleted_tags[k] = 1
import sys
if __name__ == __main__:
fix(sys.stdin, sys.stdout)
class Element(object):
def __init__(self, elem):
object.__init__(self)
self._elem = elem
self._modified = False
self._tags = {}
for t in elem.getElementsByTagName(tag):
self._tags[t.attributes[k].value] = t
def _getattr(self, name):
returns an attribute's value (unicode)
return self._elem.attributes[name].value
def _setattr(self, name, value):
sets an attribute's value
Re: [OSM-talk] area_with_holes as alternative to multipolygon relation
On Mar 4, 2008, at 22:10, Jon Burgess wrote: Thanks for the information! It's becoming clear why things are the way they are currently. How about we define this as a new relation type and depreciate the multipolygon type. Which would take us right back to the beginning of the thread :). We could only remove this suppression if we removed the tags from all of the ways. This would require all the multipolygon tags to be moved to the relation instead of the ways. The current number of multipolygon relations in the database isn't that big. A couple of thousand, it looks like. I'll try making a script that applies inner/outer roles heuristically, based on polygon area or similar, and that handles the tags sensibly. If that works out well, we could switch over without creating a new relation. But I don't mind either way. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
[OSM-talk] area_with_holes as alternative to multipolygon relation
Hello, http://wiki.openstreetmap.org/index.php/Relations/Proposed/ Area_with_holes contains a proposal for an alternative to the current multipolygon relation. Please tell me what you think about it. Am I missing something? In short, the proposed changes are the following: type=multipolygon - type=area_with_holes role=outer, way, clockwise - role=boundary, way, clockwise role=inner, way, anti-clockwise - role=hole, way, clockwise same tags on inner and outer ways; these describe the multipolygon - tags on boundary describe the area, tags on hole describe the hole This would also solve the naming ambiguity, though my aim is mostly to make things easier to tag. In particular, with the proposed relation, it's not necessary to create duplicate ways (clockwise and anti-clockwise) for holes with properties. I'm actually in favour of dropping orientation restrictions completely: The area corresponding to a closed way should be the closed area it bounds in, say, the hemisphere all nodes lie in. However, as it stands, the proposal should be straightforward to implement for any renderer that can deal with the multipolygon relation. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] area_with_holes as alternative to multipolygon relation
Hello, On Mar 3, 2008, at 18:03, Frederik Ramm wrote: Until now I was unaware that we currently require the outer/inner ways of polygons to be clockwise/anticlockwise. It seems that some renderers work better if that is the case but nowhere is it a requirement. That's how I read http://wiki.openstreetmap.org/index.php/Relations/ Multipolygon. And tagging multipolygons as in my proposal (i.e., tag the hole as a separate area) didn't seem to work well with either mapnik or osmarender. From what I remember of reading the code, both renderers skip areas with role=inner when rendering. In particular, with the proposed relation, it's not necessary to create duplicate ways (clockwise and anti-clockwise) for holes with properties. Who said this was necessary today? The wiki. And I've seen other people tag like that. I'll update the wiki later if nobody objects, and see what I can do about the renderers. Thanks! Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] area_with_holes as alternative to multipolygon relation
Hello, On Mar 3, 2008, at 21:25, Jukka Rahkonen wrote: mapnik or osmarender. From what I remember of reading the code, both renderers skip areas with role=inner when rendering. I can see that both renderers do draw the inner roles as holes, see: http://www.openstreetmap.org/? lat=60.16334lon=24.94202zoom=16layers=B0FT I think the problem I'm trying to address clearly: To map a lake in a forest area, I should currently create both a closed way with natural=water, and a closed way that is an 'inner' member of the forest's multipolygon relation. Both of these ways share the same nodes (and may or may not need to be oriented in special ways). This seems redundant. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] area_with_holes as alternative to multipolygon relation
On Mar 3, 2008, at 22:29, Martijn van Oosterhout wrote: Hmm, I thought you could use the inner natural=water as the boundary of the forest? Does this not work? I think not. I'm pretty sure for osmarender, but may not have had enough patience to test out the various combinations with mapnik. FWIW, I'd just had a layer=1 to the lake and forget about the multipolygon. I may just do that. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] area_with_holes as alternative to multipolygon relation
On Mar 3, 2008, at 22:45, Sven Grüner wrote: Robert Vollmert schrieb: I think not. I'm pretty sure for osmarender, but may not have had enough patience to test out the various combinations with mapnik. I was wrong about osmarender: What I remembered is actually code from mapnik (or rather, osm2pgsql): static int pgsql_out_way [...] const char *multipolygon = getItem(tags, multipolygon); // If the way has been part of a multipolygon then skip if (multipolygon) return 0; The multipolygon tag is set in pgsql_out_relation for ways with role=inner. Furthermore, pgsql_out_relation appears to aggregate tags of all ways involved, so you should end up with landuse=forest,natural=water on the multipolygon. I wonder what happens when both have differing natural= tags? Or am I reading this all wrong? I haven't really been able to test because so far, my mapnik install doesn't appear to render any areas... Nah, that island is just natural=water and role=inner in a multipolygon. The green comes from the park encasing the lake. But it only works when drawn counter-clockwise. http://www.informationfreeway.org/?lat=52.249lon=10.52656zoom=17 And it also works in mapnik! Maybe you're just lucky with the drawing order? :) I'll keep experimenting. Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Straw Poll - Disk space used by OSM photos/audio
Hello, On Feb 13, 2008 9:41 AM, Gregory [EMAIL PROTECTED] wrote: Maybe (I possibly thought this when I started) so nobody knows where I live/start, but I could easily not worry about that. On Feb 13, 2008, at 19:14, Karl Newman wrote: Yeah, I've been collecting traces, too, but not yet uploading them because I don't really want the cluster of points around my house. I found a graphical editor that can edit traces (GPSTrackMaker) but I haven't had time to edit them for uploading yet. I believe gpsbabel supports this kind of operation: gpsbabel -i gpx -f raw.gpx -x radius,exclude,distance=1.5M,lat=30.0,lon=-90.0 -o gpx -F filtered.gpx should remove all points within 1 1/2 miles of the specified lat/lon pair from raw.gpx, saving the output in filtered.gpx. (I haven't tested this.) See also http://www.gpsbabel.org/htmldoc-development/ filter_radius.html . Would it make sense to add these kind of filters to the GPX upload facility to encourage more people to upload their traces? I'd also like some automatic time shift (to say the first of the month, at 0:00)... Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] [tagging] grouping simple tag proposals
Hello, On Feb 3, 2008, at 20:20, Robin Paulson wrote: taking on board one of Frederic's comments from last week: there are a lot of proposed 'shop' tags on the proposals page, something which is overwhelming and time-consuming to solve using our current method of tag proposal/ratifying i think it would be a good idea to group a lot of these tags together in one proposal, [...] any tags that get even the slightest objection/comments, etc can be spun out to their own proposal and discussed in more depth as usual sounds good, but why not go a step further: There could be a page on the wiki for simple proposals, and anything that gets no objections for a week is automatically approved and added to Map Features. Cheers Rob ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Mapping canals
On Jan 23, 2008, at 11:44, David Earl wrote: I had in mind (and it'll probably stay in mind!) a renderer which showed you a ground level view of the street you were moving along with upcoming turnings and so on, like a satnav display, which showed signposts - no right turn, this way to Amarillo - and a 30 sign as you entered a 30 area (not a 52.12345 sign). How about an extra key maxspeed:source_value_with_unit=30mph and a cron job that updates maxspeed from maxspeed:source_value_with_unit? Cheers Rob ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] walking routes?
On Jan 22, 2008, at 13:07, [EMAIL PROTECTED] wrote: Nick Whitelegg wrote: TBH I would be fairly dubious about tagging any non-waymarked walks/cycle rides as routes, let alone ones of my own devising. This is interpretation which should be kept out of the largely factual OSM. The data might not fit into the OSM but its still useful. Many websites live from it, e.g. http://www.gps-tour.info/. IMO an easy way to maintain these routes would be to define special tags to the OSM GPS traces database (http://openstreetmap.org/traces). E.g. in this case the tag could be 'recommended_walking_tour'. The trace should contain only the tour and nothing else in this case. A routing application that is aware of these tags could notify the user about nearby recommended tours. But for that purpose, people can use gps-tour.info, right? OSM would be interesting by allowing to present recommended walks, etc. as a sequence of OSM ways. But this data probably would better go into a separate database. I'm sure there's an opportunity for a nice project here: A walks/ rides database that allows to construct such walks by selecting way segments. Perhaps you could also offer a program that approximates uploaded GPX tracks using existing ways, and offer the ability to upload missing ways (or refer to OSM for the last part). I've also been thinking TrailRunner (even better: a free alternative) should allow creating routes from OSM vector data. Cheers Rob ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
Re: [OSM-talk] Render icons for parking areas
On Jan 19, 2008, at 23:10, Lukasz Stelmach wrote: Ok. Forgive me my sarcasm earlier in this thread but I really think automatic placement of icons at the areas is not so good. Let me then propose different approach. Let's use the new algorithm to create nodes that would be rendered as icons. Such nodes should have a tag that attaches them to the area and the created_by tag containing the automaton name so they can be automatically moved if the conditions change. If for whatever reaseon the automatic placement is not reasonable the node can be moved by hand and have the created_by changed without removing its attachment to the area. Then the algorithm knows not to move it. Why not place nodes automatically, unless the area is member of the labelled_area-relation (with two members: the area and the position of the icon/label), as was suggested before? Cheers Robert ___ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
