Cobblestones vs setts vs paving stones

In trying to achieve clearer definitions for surface=*, I’ve started two discussions, one on Portuguese pavement, and another on a more specific type of pavement technically called “polyhedral” in Brazil. But I feel like I need more feedback.

It seems like setts are generally viewed as a type of cobblestone, as currently written in the wiki, but several sources cite that cobblestones are natural and smooth while setts are stones worked to a cuboid shape:

Based on this, it seems to me that the Portuguese pavement would be better described as a type of sett than a type of cobblestone (such as cobblestone:flattened). It has been suggested that, because gaps are tight and the surface is quite flat, that they actually resemble what in OSM is mapped as surface=paving_stones.

It has also been suggested that a surface for the Brazilian polyhedral pavement should be chosen according to its typical smoothness, with paving_stones being the smoothest and cobblestone the least smooth, which in my opinion would make the polyhedral pavement a type of sett. As such, I believe the wiki definitions could be more broadly defined as:

  • pebblestone: loose small rounded natural stones

  • cobblestone: jointed large rounded natural stones

  • sett: large natural stones cut to a flat top, usually jointed and forming a regular pattern

  • paving_stones: smooth tightly-fitting natural stones or artificial blocks, usually forming a regular pattern

This would make cobblestone no longer a superset of of sett and sett no longer required to be regular at all times.

I’m wondering what is the general opinion on this. Does it seem reasonable? Or does it look wrong?

Edit: perhaps the Brazilian “polyhedral” paving would be called in British English crazy paving (also on Wikipedia), but crazy paving seems to be flatter due to larger slabs and more even jointing, and the gaps are narrower.

The problem with your proposed definition is that, in my experience, most sett is tagged surface=cobblestone, and most surface=cobblestone represents sett. Or to put it differently: Many people don’t know and/or care about the distinction.

To me, that means the distinction between sett and cobblestone should be made with an optional additional tag, e.g. surface=cobblestone + cobblestone:type=sett/true_cobblestone.

I just inspected the definitions of sett and cobblestones in German (since 6 November 2013), Dutch (since 23 December 2015), Danish (since at least 26 October 2017 or arguably since 29 July 2014) and French (since 16 August 2017) and they all specify similar distinctions to those being proposed. Almost all translations are using the same images as in the original article in English too.

Perhaps the only major difference is that they all consider sett as having a regular shape.

Some of the images in the translations are automatically transcluded from the English version, these aren’t all independent data points. As for why the English version had that image, there apparently was this undiscussed edit in January which introduced it. And because another user edited the same template only a few hours later, few people will have noticed that edit, given that only the most recent change to a page is shown on the wiki watchlist by default. Since no reason was given for the change, I’ve reverted the edit for now.

A more general problem with using the wiki as a reference on this topic is that the wiki is acting in a very prescriptive manner here. Some years back, surface=sett didn’t exist at all, and there was no doubt that sett should be tagged surface=cobblestone. Then someone just went ahead and added surface=sett to the wiki. There even was some mailing list discussion that showed a lack of support for this change, but unfortunately, said edit was never reverted – even though that would have been perfectly justified and imo the right thing to do. So what you see on the wiki today is not necessarily how the tags are used, it’s how a relatively small group of passionate people want them to be used. There never was a vote, consensus or overwhelming tag usage to support this move.

Admittedly, I’m a bit annoyed because of the way this entire “sett” concept has been sneaked into the wiki over the years through undiscussed edits.

Well, the fact that both images were equal is, to say the least, an oddity. Still, the descriptions have not been the same as in English in these languages.

Additionally, some applications already consider these values distinct, such as GraphHopper (line 100 added 22 December 2013), OSRM (lines 208 and 217 added 24 March 2014) and recently defunct Mapzen (lines 661 and 664 added before 24 February 2015).

Also, if what’s written in this post is correct, then the usage of surface=sett increased from 29k 8 months ago to 55k today. So it seems like people are finding a distinction in reality quite often.

I both agree with Tordanik and ftbrien:

  • Most surface=cobblestone are setts;
  • it would be sensible to separate them properly.

I remember discussing this very point at SotM-10 with Lulu-Ann, and the one point which sticks in my mind is that cobblestone proper is a really awkward surface to walk on if one is wearing high-heels, but setts are not. This distinction would be true for many other cases: pulling a suitcase; pushing a pram or wheelchair; blind persons; elderly people (who often need to avoid uneven surfaces etc).

On the other topic: I think portuguese paving is a) not paving stones, and b) very distinctive with very different properties from other surfaces (from an old experience rather slippery when wet). I think there may be some other surfaces formed by embedding tiles/stones in a surface (or perhaps pre-manufactured), but this is again different from paving stones which are not usually mortared.

Whilst we are at it could we get rid the horrible surface=paving_stone:20; surface=concrete:lane etc. In my view these can be met by subtagging/adjectival tagging. They just make it that little bit harder for someone to add surface tags.

I’d not bothered checking on this until now. I just had a look at setts on wikipedia. The image there is of something I would not call cobbles (I wouldn’t have known what to call it until I just looked). Those setts are flat. I grew up in a town where cobbled roads still existed (they probably don’t exist any more). Setts are not cobbles. Setts are essentially flat. Cobbles are rounded and a little awkward to walk on. The images of cobbles on wikipedia show stones with much larger gaps than the cobbles I’ve seen, wide enough I’d be tempted to call them something else (like “crappy cobbles”). The cobbles I’ve seen were much closer fitting (and were possibly trimmed so they could fit closely), but the important thing is that they had slightly rounded tops.

Does this distinction matter in an OSM context? I think so. I wouldn’t have a problem riding a bicycle over sett. I’d dismount from the bike to traverse cobbles because it would be very difficult riding over them. I don’t think namespaciing cobblestone is a good idea because setts are not cobbles. Setts are closer to brick paving (with irregularly-shaped bricks). Cobbles are a hazard for somebody riding a bike (depending on degree of roundness and gap size).

When you put it like this, I also agree with both of these statements.

At the risk of getting a bit off topic here, I’m in full agreement with that point. In fact, I’ve written a proposal to define some subtags to replace the paving_stone:X values (and allow for a lot more detail on top of it). :slight_smile:

Two German mappers seem to have disagreed slightly on what surface=cobblestone means here.

I understand that you have the opinion that all of the surfaces below should be mapped as surface=cobblestone, right?

The first image was the first to be assigned to cobblestone in the wiki and stayed assigned to it for 2333 days.

The second was assigned to cobblestone in the wiki for 1461 days.

The third was assigned to cobblestone in the wiki for 731 days. During this period, the second image was also assigned to cobblestone.

Instead of advising the second image together with either the first or the third, since 18 August 2014, you have ensured that only the first image stayed assigned to cobblestone in the wiki. Why? If it was for contrast with paving stones, then the second image should have been kept to express the full range of pavements that the cobblestone value applies to.

Since that date, the usage of sett in OSM has risen from 6k to 55k (+49k or +792%), whereas the use of cobblestone has risen from 119k to 162k (+43k or +36%).

Only the middle one is cobblestones, the other two are setts.

Cheers Phil

Links to Wikipedia were added to the descriptions of cobblestone and sett in the OSM wiki on 5 November 2013 by a German user and on 27 February 2016 by a Portuguese user, but removed by Tordanik on 18 August 2014 and on 28 May 2017.

Here are some important edits in the Wikipedia articles:


  • 4 March 2009, British user: “distinct from a Cobblestone by being quarried or shaped to a regular form”

  • 8 January 2010, Mexican user: “often incorrectly called ‘cobblestone’, is a broadly rectangular quarried stone”

  • 4 April 2012, British user: “often idiomatically called ‘cobblestone’, is a broadly rectangular quarried stone”

  • 3 December 2012, British user: “often idiomatically referred to as “cobbles””


  • 10 March 2008, British user: “Cobblestones were largely replaced by quarried granite setts in the nineteenth century. Cobblestone is often wrongly used to describe such treatment.”

  • 1 May 2008, British user Harry Wood on image label: “A cobblestone-covered street” > “A stone-covered street”

  • 11 August 2012, American user removed sett images

It is also interesting to point out that Harry Wood (user page suggests it’s the same person as in Wikipedia) wrote on the OSM wiki on 1 May 2008 that mappers were already using both cobblestone and sett.

I’d say (but I’m just a Briton using the local idiom where I was brought up) that cobblestones do not have flat upper surfaces. The sides may be trimmed to a regular rectangle but the top has, at the least, curved edges. Not fully domed, but not fully flat. Awkward to walk on.

The things called cobbles in the wikipedia image, domed stones with large gaps between them, appear to be designed to prevent traffic of any sort. I start to wonder if those are what Terry Pratchett had in mind when he mentioned “cat’s head” cobbles.

I also wonder if we’ll ever reach agreement on these things. :slight_smile:

It’s also worth remembering that a single person, with a bee in his bonnet, can put a lot of stuff on various wikis using various aliases endorsing a position that only one person (himself) thinks is correct. So you can’t entirely trust anything you see on Wikipedia and other wikis.

Are they how these British guys describe them?

Cobblestone on Wikipedia says that:

  • (below the domed stones image) “Cobblestones such as these are designed for horses to get a good grip.”

  • “Setts were relatively even and roughly rectangular stones that were laid in regular patterns. They gave a smoother ride for carts than cobbles”

That sounds like the literal translation of the German word “Katzenkopfpflaster” (sometimes “Katzenkopfsteinpflaster”; literally “cat’s head stone pavement”) that has shown up in the forum and in the wiki to describe cobblestones, along with “Kopfsteinpflaster” (literally “head stone pavement”). German Wikipedia links Katzenkopfpflaster to this image (used today as an example of cobblestone:flattened in OSM) and redirects Kopfsteinpflaster to Großsteinpflaster where it is written in German “The term Kopfsteinpflaster refers to round or head-shaped stones such as Katzenkopfpflaster or Bonbonpflaster, which is rarely found in Germany.”

Surely, but questionable (check out the citations). Wikipedia surely has more eyes checking their articles than OSM does in its own wiki. The parts that I have cited have been kept in the articles for years and none of the subsequent authors have put them in question, except, perhaps, to express annoyance at possible nitpicking. Have you also looked at the other references I gathered at the beginning of this thread?

I don’t mean to defend Wikipedia, but it was used as a description for cobblestone and sett in the OSM wiki for a little more than 2 years. This is important if we want to understand the assumptions of mappers and their impact on the data we have today.

As someone born and bred in England, I would refer to both 2 and 3 as being cobbled streets. I’d never heard the term setts, before.

Sounds like image 3.

Up to this point, I see no support for image 1 as representative of cobblestones. I will add to that some statements I found elsewhere in OSM channels. British user Jonathan Bennett wrote on 20 February 2012, in agreement with the British references at the beginning of this thread:

British user Andy Mabbett wrote on 10 May 2015:

British user pmailkeey wrote on 11 May 2015:

The first URL is no longer available, alternative images of the same place can be found here and here.

So this:

Could perhaps be changed to this:

But that’s not technical and would prompt the question: how much room is there for surface=cobblestone:flattened to be curved? Or should cobblestone:flattened include only natural uncut stones? Fortunately, cobblestone:flattened usage is very low.

A bit late maybe, IMHO figure 1 are natural stones, flattened but cobblestone, fig 2 is undoubtedly cobblestone, surface=rough and fig 3 could be a rough pavement made by so called klinkers a baked clay product. Which come in many variations and names, which no one would describe by it specific type name. I personally use surface=paved for clay and concrete stones. The concrete ones come in many forms and colours for instance to make sett as well. :slight_smile:

To further enrich the debate, I’ll just add the translated opinions of some German mappers in this thread:

And here’s a summary of the definitions of these surfaces on the English OSM wiki over time, juxtaposed with tag usage history.

The bottom timeline uses codes to refer to the top images and the descriptions in the middle. So 21A means images 2 and 1 were used simultaneously, with 2 first followed by 1, together with description A. The purple horizontal line represents half of all usages of cobblestone.

Considering the 218k OSM objects with surface=cobblestone (162k) or surface=sett (56k), it seems to me that:

  • ~25% are surface=sett that in reality have a flat top (56k out of 218k)

  • ~25% are surface=cobblestone that in reality have an arched top (whenever the image for cobblestone was 6 or 9)

  • ~50% are surface=cobblestone than in reality may have an arched or a flat top

Remarkably, usage of sett has accelerated by the time that the aforementioned thread in the German forum started getting replies.

Considering both vernacular and technical usage, maybe it would be most useful in the long run to define these surfaces like this:

  • paving_stones: blocks or stones with smooth flat top, tight gaps, images 1 and 2

  • sett: hewn stones with flat top, wide filled gaps, comfortable to cycle and walk on, uncomfortable on high heels, images 3 and 4

  • cobblestone: hewn stones with slightly arched top, wide filled gaps, uncomfortable to cycle on, difficult on high heels, images 5, 7 and 8

  • cobblestone:raised: natural or hewn stones with very round/irregular top, wide empty gaps, fixed to a bedding, difficult to cycle and walk on, uncomfortable to drive on, images 6 and 9

  • cobblestone:flattened: never

  • pebblestone (for contrast with cobblestone:raised): loose natural small rounded stones

Back to my first post in this thread, this would make the Portuguese pavement and the polyhedral paving a type of sett most of the time. Portuguese pavement does resemble image 3 and is uncomfortable on high heels (as on sett), and crazy paving does resemble image 4 except for the regularity of stones when viewed from above (also uncomfortable on high heels according to multiple anecdotal reports on the Internet). Crazy paving would sometimes be sett, sometimes paving stones (like flagstones would), depending on how wide the stones are in relation to gap sizes.

2 would be called Block Paving in England, today. They are used for footway crossovers (“dropped kerbs”) and personal hard standings, and occasionally for low traffic residential roads, more likely private ones.

Paving stones is generally only used for pedestrian areas with large, rectangular, lumps of concrete. These will eventually crack if even just domestic cars drive over them, especially if the vehicle is turning, so are never used for intentional vehicular areas. When reporting ones that are broken (usually by abuse by motor vehicles, my council just calls them “slabs”.