Residential fences/property boundaries

What is the consensus regarding mapping residential fences i.e. fences along property boundaries? The documentation doesn’t explicitly discourage this, as far as I can tell, but I haven’t seen any examples of neighbourhoods with fences thoroughly mapped.


I don’t think there is a consensus - but I doubt anyone would object as long as your source of the data is compatible with OSM licence requirements.

What you see in OSM is its own hierarchy of needs: first streets get mapped, then major POIs, then turn restrictions and area polygons, then finally when everything else is mapped you start to see people turn to less obvious things (litter bins, fences etc).

I’m interested in the same issue. It seems what is most commonly used is landuse=residential, however I don’t like that usage very much. I’ve imported a dozen property boundaries and houses in a neighborhood nearby, and have used that landuse=residential tag, which you can see here. I also applied county specific identifiers to the property boundary, which at least I can use to differentiate between individual property boundaries and large areas that happen to be tagged landuse=residential.

I’m really not sure what tag is appropriate, but if you have any ideas please share them. And if not, just ensure that you tag them with something that can later be used to automatically re-tag them if a property boundary scheme is created.

One other thing, keep in mind that oftentimes fences and property boundaries won’t always line up. However, if you’d like to map fences as well as property boundaries, and if they do line up, then use multipolygons. For example if a fence surrounds the back and sides of a property, then create a way for the fence and a way for the rest of the property boundary. Then create a type=multipolygon relation using the two ways as role=outer members.

Take a look at this area:

Impressive! - to say the least.

I will have my thoughts and questions on the subject of fences, but I wanted to ask you about something that I have struggled with and you, apparently, have answered for yourself. It has to do with duplicating information in the database. I am looking at your Cardinal Wiseman Catholic Technology College.

You tagged

  1. Grounds of the college as amenity=school + name + full info, including address.
  2. The outline of the building as amenity=school + name only (which is the same as above).
  3. Node in the center of the building tagged amenity=school (no name or other info).

Could you please run your rationale by me, and if we have to open another thread so as not to hijack this one, that’s fine, too.

My instinct would be to err on the side of having as little data in the database as possible arranged in the most efficient way possible, and rely on rendereres and query tools to properly interpret/use my data. Faced with the same assignment, I would have

  1. Tagged the grounds of the college same way as you did.
  2. Drew the outline of the building and tagged it building=yes or maybe building=school, no name.

I would be concerned with Mapnik et al putting the same name on the map twice, which it did in your case, and you seem to be okay with that. Could you please point me to a past discussion of the best practices for this sort of tagging assignment, or give me your own use cases that call for each of the three tagging options that you used.

Granted, when I tag a school (or a church, etc.) the way I described above, many QA tools will point to a building with no name, etc. Also, Mapnik is reluctant to draw a school icon unless there is a POI node, the way you did it. Is that the main reason?


P.S. I guess I am trying to think like some abstract search algorithm. Is it going to tell me that there are three schools in this area?

Back to the original question and the points made so far.

I don’t think anyone would frown upon anyone mapping fences. I also think that mapping fences around individual properties is extravagant. It’s fine for a small area (maybe the mapper’s own neighbrohood) just as a showcase of what can be done and an exercise in completeness (see monxton’s effort). But at some point the mapper’s energy and time is not infinite and this is where csdf’s comment comes in handy. Believe it or not, I was about to ask someone to break down for me the usefulness of types individual mapping activities to the OSM community, and this is more or less what he did. (Big thanks!)

That said, I don’t know what kind of self-discipline and attention to detail it will take to say to oneself: “I am not going to do any POIs until all streets in my area have been reviewed and I am not going to draw any are polygons until I have mapped out all major POIs”. Some activities are just more fun and provide a bigger bang for one’s volunteer buck as far as the finished product (if the rendered maps are the finished product) are concerned. So, personally, while keeping in mind the needs of the routing apps and checking turn restrictions and new streets every chance I get, I am actively mapping land use polygons in my area, mainly shopping plazas and residential neighborhoods. It just makes the Mapnik map so lively :slight_smile: (“lived in” as one of the oldtimers calls it). And this is where this conversation comes in.

Personally, I would very much inclined to map the fence (or sometimes a wall, or hedge) that serves as the boundary of a neighborhood. To me, it allows to kill two birds with one stone: get a nice shaded and named polygon for the neighborhood AND map out a physical feature (fence) at the same time.

Granted, of the 10 or so neighborhoods I have mapped out (as named residential polygons), I have not tagged the fences yet. Simply because Bing imagery allows to see the shape of the neighborhood easily but does clearly show fences, walls and hedges, and one needs to either physically survey each area or rely on much more detailed imagery, and who knows how that sits with the OSM license.

I would say: when the nature of the linear barrier (fence, wall or hedge) is known, map it for a neighborhood and turn it into the boundary. As far as individual properties, map the fences when there is absolutely nothing left to do in OSM (or in your area). Hope this helps.

Sorry to come to the thread a bit late, but it seems like a fine opportunity to blow my own trumpet. So I will. :slight_smile:

I’ve spent a while mapping houses, garages, driveways, fences, hedges, etc. around Cambourne:

It was mentioned at the end of this blog post over at Itoworld:

As mentioned above, there are other priorities — obviously mapping streets and footpaths is of more use, and probably POIs as well, but for door-to-door routing and for gorgeous looking maps at large scale, you can’t beat this!

In fact, when the house numbers are in wacky arrangements like on Woodpecker Way in Cambourne, OpenStreetMap becomes the tool of choice for guiding visitors and deliveries to your front door. It’s just a shame that the Bing imagery runs out two thirds of the way around. :frowning:

Man - you must have had a slow couple of days!

The archetype is Blackadder’s work in Sutton Coldfield (postcode district B72): and Chris Hill has also posted on what’s involved

I’ve had a go and it hugely adds to the effort for an existing survey. If you want something to do in your lunch break or need 20 minutes away from the computer then its fine: I think this is what Andy (Blackadder) did. It’s not enough to draw stuff in from aerial imagery, it needs to be surveyed too: is it a fence, wall, hedge; is it still there.

A quicker approach is to trace ‘property boundaries’, i.e., based on fences and other information. There has recently been a lot of discussion about the best way to tag these on talk-tagging, much of it summarised on the wiki (

Cadastral boundaries per se don’t seem to me to suitable for OSM: they are not amenable to ground verification, and its not clear what value they have in OSM anyway. Cadastral maps also have inaccuracies: I once bought a house where the Land Registry had everyone on the street owning only half of the land for their own garage, but also owning half of the land for their neighbours garage. I don’t know how many transactions had completed with this error, before our own, but we had an unusually thorough and competent lawyer.

Obviously the law is different where you live. Here the cadastral map can’t be wrong. It is what is leading legally. It would be the idiot property developer that had the garages (and houses) built in the wrong place.