[Tagging] Updating definition and description of place=square

Martin Koppenhoefer dieterdreist at gmail.com
Tue Mar 24 12:11:49 UTC 2020

Am Di., 24. März 2020 um 03:33 Uhr schrieb Joseph Eisenberg <
joseph.eisenberg at gmail.com>:

> > Place=square was defined until 3 days ago as “a
> > named square” and “a town or village square which is an open space
> common in
> > urban centres, typically crossed by streets but can also be a pedestrian
> > area or more rarely green areas.”
> >
> > I am perfectly fine with this documented definition
> But the first part wasn't a definition. "A named square" is not a
> defintion at all, since the word "square" is undefined.

is it? The English wikipedia has an article "Town square"
https://en.wikipedia.org/wiki/Town_square and tries to give a definition
(although the article is in bad shape and also has a warning that the
affirmations are not backed by citations, and it links to articles in other
languages which have different definitions, and which may be partial, due
to these other languages having several similar concepts). Also the article
states that "Other names for *town square* are *civic center
<https://en.wikipedia.org/wiki/Civic_center>*, *city square*, *urban square*,
*market square <https://en.wikipedia.org/wiki/Market_square>*, *public
square*, *piazza <https://en.wikipedia.org/wiki/Piazza>*, *plaza
<https://en.wikipedia.org/wiki/Plaza>*, and *town green*." but these are
not all 1:1 synonymous.

Oxfordlearnersdictionary also has a definition: " [countable] an open area
in a town, usually with four sides, surrounded by buildings"

We did not so far define the words "street" or "road". It is taken as
granted in the highway tag definitions that you know what it is.

> If this means
> "a feature that includes the word "square" in the name" as the page
> suggested back in 2015-2016 this is even worse, since it is completely
> culturally determined. I would be justified to tag all "alun-alun"
> feature as squares, even those that are 100% soccer pitch now, and
> those function as a walled palace garden.

whether and how many of all objects with "square" (and similar) in the name
are actually the kind of object we are tagging with place=square will
likely depend on the culture and language. If this works for Italy with all
(or almost) piazza, piazzale, piazzetta, largo, campo, it does not imply it
works in Indonesia as well.

> The first second definition was a little better: " an open space
> common in urban centres..."

yes, but it implies a certain kind of urbanism, in particular how centers
and peripherical areas look like, and I believe it is not universally
applicable (if it is intended to exclude squares in the outskirts).

> Though this could be used for a leisure=pitch or leisure=park or
> leisure=garden or an amenity=parking, or a fenced-off roundabout
> etc...

no, the place=square can hardly be the same area as a leisure=pitch or
leisure=park. You won't have the park cover the whole area and end at the
buildings, there'll be a way or street along the buildings.

> But then the second half of the definition offers several more
> possibilitiies:
> "typically crossed by streets" - That one is unclear, does it mean a
> street intersection/ road junction? Most mapped place=squares are NOT
> crossed by streets, it turns out.

it may depend how you define "street" and "crossed". I guess it includes
pedestrian streets, and crossed may also be seen as "at the borders". If
you read it like this, there won't be many examples that do not fit (can
you post one?).

> "But can also be a pedestria area or more rarely green areas.”
> A highway=pedestrian area is certainly a type of open public space, so
> that is fine, and the most classic squares fit that definition.

although not all pedestrian areas are squares. It could also be a parking

> But what does "more rarely green areas" mean? Is a green area just a
> flat, mowed lawn, or can it be an elaborate garden with trees, knolls,
> ponds? Can it be a leisure=pitch? Can it be a park with trees, picnic
> areas?

I believe it refers to the surface. It the square must not necessarily be a

> I don't think that is what was intended: generally a "square" seems to
> be designed to be used for events or for people to congregate, at
> least historically, so if it is green, it is just grass, not trees,
> flowers, shrubs, gardens, water features, etc, else it's a park or
> garden.

It may have changed the purpose, but still functionally, architectonically
be a square. I agree with the notion that squares are made to stay, while
streets are made to walk/drive. There mustn't be "events", and many of them
are made to make a building stand out more (e.g. a square in front of a
castle or church).

> That's whey the prior definition is inadequate: it is non-orthagonal,
> it can include many types of features, and is impossible to translate
> into different cultures.

agreed, it wasn't a good definition before (it did not explain to someone
unfamiliar with the concept, when to apply the tag), but at least it didn't
exclude some kinds of squares by setting criterions that seem not
universally shared.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20200324/e1d7936b/attachment.htm>

More information about the Tagging mailing list