Re: [Tagging] Tagging of topographic areas with a name
On 08/19/2013 11:51 PM, Masi Master wrote: This is a bit away from the new valley mountain discus, but has a connection to the first mail. Tagging should be thought-out with possible examples, if we don't want to change the tagging or live with a bad tagging. Another example I had just yesterday was a lake called Seebergsee in the alps. The lake itself is comprised of a very small persistent lake which is well delimited, and a marsh which is filled with water during 1/3 of the year as the snow melts. Independently of the tagging (which is well delimited in this case), the name refers to the lake _and_ the marsh. Maybe there's some waterway relation magic for this specific case, but I'd rather use some consistent topological naming of areas also for these cases. ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Tagging of topographic areas with a name
On 08/17/2013 05:47 PM, Wolfgang Zenker wrote: * fly lowfligh...@googlemail.com [130817 17:13]: On 16.08.2013 19:05, Masi Master wrote: Hmm, I'm not sure that boundary is the right tag. Isn't it a border, and not an area? Boundaries describe an area but you are right that they are not really boundaries, especially if the border lines are not clearly defined [..] I'm under the impression this discussion is leading to ever more complicated ideas, due to the problem that the features we want to name on the map are not really clearly defined areas. Just for clarity, I was really hoping to find an already-established tagging scheme for these features (named topological areas, valleys), and bringing up the schemes I found in several other places rather than trying to overcomplicate things. While I agree that rendering should follow tagging, I also go by the idea that as long as the scheme is consistent, one could switch to an improved one later quickly enough. Also, the people involved with rendering should have a pretty decent overview of how the tags are actually used, corner cases and the limitations involved. This is as important as tagging itself IMHO (I was rendering navteq data in the past, so I value a lot input from software implementations). I'm not sure if this list is followed by people involved with styling/rendering OSM data itself? (please tell me if some other list might be more appropriate). ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Tagging of topographic areas with a name
On 08/08/2013 11:54 PM, Martin Koppenhoefer wrote: I guess in this case I can simply re-use the geometry in a new relation with the proper valley name with type=multipolygon, place=region, region:type=valley? I'd use type=multipolygon natural=valley I'm still not satisfied with type=multipolygon: http://wiki.openstreetmap.org/wiki/Multipolygon#Detailed_tagging specifically: * The relation has tags: Use the relation tagging. Ignore anything on the ways. However, this is not what should happen for a lake group where each lake name is independent (ie, the group is just a topological feature). And, as I said before, unnamed lakes should not inherit the name of the group. After re-reading the whole thread, I tend to agree with fly more, as a boundary type seem to be much more appropriate: type=boundary boundary=topologic natural=water name=lake group name the boundary relation has the advantage of not requiring a fake polygon (as opposed to place=locality). I have two examples of type=multipolygon which I introduced: http://www.openstreetmap.org/browse/relation/3126464 http://www.openstreetmap.org/browse/relation/3126459 whole type=multipolygon relation simply broke the rendering (but renderers here seem to be compliant). ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Tagging of topographic areas with a name
On 08/07/2013 10:19 PM, Friedrich Volkmann wrote: On 06.08.2013 15:51, Yuri D'Elia wrote: http://www.openstreetmap.org/#map=14/45.2466/6.0866 which has been tagged with a multipoligon relation. Unfortunately, the relation has some problems: - not rendered anywhere? This is a super-relation, with other relations as members. This is not allowed for multipolygon relations. It should rather be a type=collection relation. This is how water areas such as riverbanks use to be joined, and I use collection relations for sets of rocks etc. too. Don't expect dumb renderers like Mapnik to render superrelations, though. Very good explaination. It seems to me that the closest tagging scheme might be a loose area with place=locality. Would that be a good idea? That depends on what the name belongs to. If it's the name of a lake, forest, or other physical feature, place=* would be just wrong. After reading all the replies, it seems that if a group of lakes has a name, I would probably use either a multipolygon (if feasible) or a super-relation, with the appropriate natural tag. Though for places without actual physical attributes, place=location sounds reasonable. It also looks like that the ThunderForest maps are correctly rendering the place=location tag: http://www.opencyclemap.org/?zoom=11lat=46.5215lon=11.37205layers=000B I will now convert this group to a super-relation. My issue with normal multipolygons is also that smaller, unnamed lakes inherit the name of the relation, which is incorrect. These proposals are somewhat obsolete, as natural=* has widely been accepted as the key for all geomorphological features. See http://wiki.openstreetmap.org/wiki/Key:natural, group 3. A valley is just the complement of a ridge or arete. Just draw a line along the valley and tag it with natural=valley. I still have doubts about this. For the valley I'm speaking about the whole region, which is an area. By looking at your next pointer (about mountain_range), it looks like I can follow the same scheme and use region_type=valley as a subtype. Similarly, we have areas for entire mountain groups, which are fundamental for a topographic map in the alps. Again, the boundaries of such areas are not so important, but it's mostly used as an indication for the name placement. natural=mountain_range is already in use for the Alps. The mountain groups within the Eastern Alps are tagged place=region, see the members of relation 2113486. This has been incredibly helpful! I assume this is the data that is being used to render the topographic map at dianacht.de? (http://geo.dianacht.de/topo/) ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Tagging of topographic areas with a name
On 08/08/2013 07:15 PM, Martin Koppenhoefer wrote: Though for places without actual physical attributes, place=location sounds reasonable. thing is that place=locality is very generic, you don't get additional information what the name refers to, especially if tagged on a node Understood. ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Tagging of topographic areas with a name
On 08/08/2013 08:56 AM, Friedrich Volkmann wrote: On 08.08.2013 01:24, Pieren wrote: On Wed, Aug 7, 2013 at 10:19 PM, Friedrich Volkmannb...@volki.at wrote: It should rather be a type=collection relation. I really hate type=collection. One of the worst idea in OSM. All relations are collections. At least it is semantically correct, while type=site relations are often used for features on multiple sites. You can think of type=collection as an abbreviation of type=bare_and_general_collection. All other relations have special members (e.g. inner/outer in multipolygons) or at least special meanings (type=route). type=cluster has also been suggested. I would be ok with it, but it would require a proposal to make it more popular. What about type=site with the appropriate natural tag? https://wiki.openstreetmap.org/wiki/Relations/Proposed/Site I was just looking at the wiki, and type=collection seems to be pretty frowned upon. ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Tagging of topographic areas with a name
On 08/07/2013 10:19 PM, Friedrich Volkmann wrote: Similarly, we have areas for entire mountain groups, which are fundamental for a topographic map in the alps. Again, the boundaries of such areas are not so important, but it's mostly used as an indication for the name placement. natural=mountain_range is already in use for the Alps. The mountain groups within the Eastern Alps are tagged place=region, see the members of relation 2113486. So, I was looking about using place=region for valleys. At least for the valleys I was looking into, it seems that Italy already has a boundary=administrative multipolygon for most of them, although in rare cases some natural features are more detailed than the administrative boundary. Let's take this for example: http://www.openstreetmap.org/browse/relation/47143 where the administrative boundary matches exactly with the actual valley. I guess in this case I can simply re-use the geometry in a new relation with the proper valley name with type=multipolygon, place=region, region:type=valley? ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
[Tagging] Tagging of topographic areas with a name
Hi everyone. I'm in the alps, and I've been mapping some areas in the region. I have two questions regarding tagging where I couldn't find a decent consensus on the wiki. There are many areas in the region that go by a specific name. I have two cases where a group of lakes (as a whole) is known by a name, but then each single lake has also his own lake. I found an existing example in France, Les 7 Eaux: http://www.openstreetmap.org/#map=14/45.2466/6.0866 which has been tagged with a multipoligon relation. Unfortunately, the relation has some problems: - not rendered anywhere? I would expect that when the scale is high enough, and there's no place to render the lake names, the name of the relation is shown. But it's not. On the contrary, unnamed lakes simply take the name of the relation. - sometimes I not only have lakes, but I might have other features inside that area, that are logically part of the same known spot. Is a relation still a good idea in that case? It seems to me that the closest tagging scheme might be a loose area with place=locality. Would that be a good idea? I did a test, here: http://www.openstreetmap.org/#map=15/46.4696/10.7590 but again, no renderers seem to pick up this important information (the name - the boundary itself is not important!), which would be especially important for a topographic and landscape map. A related question is the name of the valleys. I saw several proposed tags in the wiki: http://wiki.openstreetmap.org/wiki/Relations/Proposed/Region http://wiki.openstreetmap.org/wiki/Proposed_features/Valley but not really an official tagging scheme. Valley names are very important features for a topographic map. Similarly, we have areas for entire mountain groups, which are fundamental for a topographic map in the alps. Again, the boundaries of such areas are not so important, but it's mostly used as an indication for the name placement. Thanks! ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Tagging of topographic areas with a name
On 08/06/2013 04:14 PM, Janko Mihelić wrote: 2013/8/6 Yuri D'Elia wav...@users.sourceforge.net Similarly, we have areas for entire mountain groups, which are fundamental for a topographic map in the alps. Again, the boundaries of such areas are not so important, but it's mostly used as an indication for the name placement. I don't know about the others, but I've been thinking about this one, and there's a simple solution. Drawing a big polygon around the whole mountain is not very effective. There are no clear boundaries for a mountain. But what we can do is put a tag like mountain=* on all natural=peak nodes. Maybe even on alpine_huts and other features. That way some software could find arbitrary boundaries using that data and SRTM data. Maybe valleys can be solved in the same way. Might still be problematic. A forest, sometime lakes, rivers for sure and many other big polygons will cross the boundary of the mountain group. It's kind of unfortunate, because a mountain group will span across italian regions and include parts of several valleys. Of course, likewise, valleys have the same problem. It's not a hierarchical information either. It's really a topographical information, and I feel like tagging objects within or using relations might be really problematic. Just imagine what kind of spotty tagging would you have for big mountain groups. Huts and peaks would definitely not be enough for a decent boundary. But also drawing big areas is kind of ugly :(. Fortunately, the boundaries of the area are not important in themselves. Nobody renders valley or mountain group borders. But we *do* use such boundaries for name placement. I'm thorn. ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
Re: [Tagging] Tagging of topographic areas with a name
On 08/06/2013 04:27 PM, Yuri D'Elia wrote: Might still be problematic. A forest, sometime lakes, rivers for sure and many other big polygons will cross the boundary of the mountain group. It's kind of unfortunate, because a mountain group will span across italian regions and include parts of several valleys. Of course, likewise, valleys have the same problem. It's not a hierarchical information either. It's really a topographical information, and I feel like tagging objects within or using relations might be really problematic. Just imagine what kind of spotty tagging would you have for big mountain groups. Huts and peaks would definitely not be enough for a decent boundary. But also drawing big areas is kind of ugly :(. Fortunately, the boundaries of the area are not important in themselves. Nobody renders valley or mountain group borders. But we *do* use such boundaries for name placement. I'm thorn. I'm attaching a crude osm file I edited quickly to demonstrate the problem. Valleys usually end exactly at the mountain ridges. Valleys also end at the border of a mountain region or at the border of another valley. Between valleys, the border is purely arbitrary (it's mostly determined by geographic properties). In the alps I would expect a mosaic which is essentially totally filled with valleys. A relation would be great to re-use existing geometry, but some new boundary type will also be needed to mark the end where's no additional geometry can be reused. I also created two (inexact) mountain groups. Mountain groups actually form a complimentary mosaic, as you see in the file. A mountain group would start at the middle of a valley (which I didn't do in the example, but you get the point) and end at another one. The only exception might be where you have very large valleys, like the Val D'Adige, where the group doesn't start in the middle exactly (but doing so wouldn't exactly be wrong either). For mountain groups I do not see any existing geometry that could be reused, except occasionally for the nodes where the valleys cross. A new boundary type is definitely needed, and the edges could be shared with a mountain group relation. ?xml version='1.0' encoding='UTF-8'? osm version='0.6' upload='true' generator='JOSM' node id='-385' action='modify' visible='true' lat='46.6430679777813' lon='11.052495557349316' / node id='-366' action='modify' visible='true' lat='46.49188168894685' lon='11.043871730621769' / node id='-347' action='modify' visible='true' lat='46.459302571583' lon='10.36979709656485' / node id='-345' action='modify' visible='true' lat='46.348707668264105' lon='10.364026723600112' / node id='-343' action='modify' visible='true' lat='46.31049752324693' lon='10.494821428967915' / node id='-341' action='modify' visible='true' lat='46.356644089129794' lon='10.727664750611696' / node id='-339' action='modify' visible='true' lat='46.42752467234013' lon='10.91235837302667' / node id='-337' action='modify' visible='true' lat='46.59815137198975' lon='11.092740082077869' / node id='-335' action='modify' visible='true' lat='46.601607818458966' lon='10.867083282707044' / node id='-333' action='modify' visible='true' lat='46.61197583489281' lon='10.763597361976476' / node id='-332' action='modify' visible='true' lat='46.60802634834044' lon='10.573873173970435' / node id='-321' action='modify' visible='true' lat='46.83957020991874' lon='10.644301092245405' / node id='-319' action='modify' visible='true' lat='46.805147933633975' lon='10.724790141702512' / node id='-317' action='modify' visible='true' lat='46.82088656606407' lon='10.826838757978491' / node id='-315' action='modify' visible='true' lat='46.930928039250446' lon='11.023030816030195' / node id='-313' action='modify' visible='true' lat='46.8955821746472' lon='11.125798084533468' / node id='-311' action='modify' visible='true' lat='46.81252599107442' lon='11.23718917976429' / node id='-308' action='modify' visible='true' lat='46.697313033657714' lon='11.088428168714096' / node id='-306' action='modify' visible='true' lat='46.67315605116175' lon='11.047464991758245' / node id='-304' action='modify' visible='true' lat='46.62678384457391' lon='10.862771369343271' / node id='-303' action='modify' visible='true' lat='46.641587804517854' lon='10.611243089789806' / node id='-223' action='modify' visible='true' lat='46.39911725379921' lon='10.983395899954589' / node id='-221' action='modify' visible='true' lat='46.4202488648184' lon='11.004914609498558' / node id='-219' action='modify' visible='true' lat='46.436595031792024' lon='10.971137954815664' / node id='-213' action='modify' visible='true' lat='46.644455683599396' lon='11.210808472618753' / node id='-211' action='modify' visible='true' lat='46.52888351182877' lon='11.30495191439448' / node id='-209' action='modify' visible='true' lat='46.476944108594104' lon='11.358850831441648' / node id='-207' action='modify' visible
Re: [Tagging] Tagging of topographic areas with a name
On 08/06/2013 07:04 PM, Martin Koppenhoefer wrote: For other areas other data types might be more adequate: Some years ago on the German ML there was this interesting idea to define (fuzzy) areas (e.g. lower scale topographic regions like the European Alps). You put existing objects (like nodes, ways or relations) into a relation with the roles inside or outside and some algorithm would calculate an area that includes all inside and excludes all outside objects. You won't have to be very precise with this, as this kind of rough information is only required on lower scales where some kilometers more or less won't change anything, just a few nodes should suffice to define something as huge as the Alps, and you could reuse (preferably simple and stable like peak-nodes) existing geometry. The message from fly, about about boundary=topologic/geographic though would solve nicely valleys, mountain groups _and_ other topographic features under a single umbrella, and it's quite easy to achieve. to fly: Is this some form of official proposal? Calculating a concave hull from points, especially where you have nested geometry is very messy process (I used to do it as a gis developer in the past). I wouldn't really expect decent results even for name placement. +1, usually you will have a river or stream there, as it is the locally lowest point (i.e. the needed geometry is already there). An argument against reusing rivers to define mountain groups is that they often add a lot of complexity and you'd usually not need the borders of a mountain group with the precision this allows for (adding relations augments complexity and raises the barrier for other mappers to edit). Ridges can also be quite complex. Also, many times they end way before the end of the end of the hill or do not exist at all (flat top mountains). Just to say that the geometry might not always be there. Also, is there a tagging scheme for the lowest point/depression of a valley? (I was looking for it recently). ___ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging