Proposal talk:Garden specification

From OpenStreetMap Wiki
Jump to navigation Jump to search

I like the general idea of this, just not sure about the sub types. A rose garden, for instance, is normally part of a larger park or garden rather than a thing in itself. Perhaps that's ok though, in which case you might want a herb garden, water garden, rock garden (rockery) etc. I think the main thing is to differentiate between formal gardens and residential gardens.Seventy7 18:15, 18 May 2010 (UTC)

Yes, the main reason I started this is to differentiate formal gardens and residential gardens. But I think the rosarium is ok - if it's a part of larger garden, then it's not a place "where mostly roses are grown"; this value is meant for gardens where roses are pretty much the only thing you can see there. Xificurk 21:30, 18 May 2010 (UTC)

Pub beer gardens

I like this proposal, but does it somehow include the gardens main usage. Only example I can currently think of right now is a pub garden (I.e one attached to a public house) aimed mainly for enjoying a drink in? Martin Renvoize 09:35, 20 July 2010 (UTC)

We already have amenity=biergarten for that one. --Lyx 13:46, 13 November 2010 (UTC)
A .de biergarten is really not the same sort of thing as a .uk beer garden attached to a pub though - that's why the former tag is in German :) The right thing to do for a Brit-style pub is probably to add beer_garden=yes to the amenity=pub POI. Do that first, but if you have detailed imagery you can locate the garden itself as well: no reason not to tag it garden=beer_garden. --achadwick 15:08, 5 June 2011 (BST)


What is the difference between "castle" and "monastery" except the fact that one is near castle and the second near monastery? --Bilbo 09:50, 19 May 2010 (UTC)

There are some other differences - the monastery garden is usually maintained by the monks themselves in contrast with castle garden where it's a professional gardener or specialized company; I would say the monastery garden is more of a "quite" place, etc. It is true that most of it can be deduced from the fact that garden is near castle or monastery, but I think it's better to be more specific than another vague value describing wider range of gardens. Also according to tagwatch there already are few garden_type=castle,monastery areas. Xificurk 14:37, 19 May 2010 (UTC)


What do you want to say with garden:type? Relation to surroundings (backyard, castle, monastery, university) or type of using (scientific (botanical), as a private vegetables or flower bed). Most of your examples of garden:style are about architecture. This is fine. But what do you want to say with kitchen (type of use) and rosarium (main plant)? There exists more types of plant special gardens Erica, Rhododendron ... --vsandre 13:54, 3 June 2010 (UTC)

The idea is that usually you can describe the garden by 1) the main type of usage (which is mostly in 1:1 relation to its surroundings) - garden:type - and 2) by the architecture of a garden - garden:style. kitchen style stands out a bit, but if you think about it, it's a kind of an architecture (functional style of a garden) - it excludes other styles (architectures), and it can be combined with type of residential, monastery, or even botanical; similarly for rosarium. Anyway, the proposal suggests only the most common values, I expect people will use whatever suites the purpose best, and only the time will show which values are the most popular. There were two main reasons for this proposal - differentiate residential gardens from scientific botanical, larger castle gardens etc.; and provide a "standard" way for more detailed description of gardens (if you take a look at tagwatch, there are currently many different attempts to do so). --Xificurk 23:59, 3 June 2010 (UTC)
I like the idea behind, but your choosen tags are bit misleading. How do you want to describe a botanicle garden near a castle? Or a Japanese styled garden with roses? The problem is you combined using, architecture and surroundings in two tags. If I am totally wrong, please add some examples (with picture and detailed tagging) --vsandre 09:07, 4 June 2010 (UTC)
There will always be some grey areas, just use a common sense. botanicle garden near a castle - is it scientifically structured and labelled collections? yes => botanical; no => castle. The same case can be a residential garden near castle and so on. I pretty much doubt you will find combination like a Japanese styled garden with roses, anyway as the proposal says rosarium is not meant for gardens where among others is a bed of roses; take a look on wikipedia:Rosarium. The problem is that trying to categorize a garden is like categorizing a music - each person has a different opinion on the number and the borders of categories... - so it doesn't make much sense try to precisely define limited number of values and force people to use only those (you can't do that anyway). As I said, I expect the time will show. --Xificurk 10:18, 4 June 2010 (UTC)


For neatness and consistency with (some) other tagging schemes, I'd prefer this to be closer to the iterative refinement pattern documented at Any tags you like#Syntactic conventions for new tags. Something like:

garden=residential → refines the definition
garden:style=zen → describes the stylistic qualities

somehow feels a little nicer to me. --achadwick 17:26, 4 October 2010 (BST)

That was my main thought on this proposal. We've generally avoided using the word "type" in favour of this iterative refinement pattern
-- Harry Wood 20:59, 16 October 2010 (BST)
Agree with not using type. But what to use in its place? garden:function=botanical/arboretum/community/civic/residential/**?? Warin61 (talk) 22:36, 11 October 2018 (UTC)

Residential gardens (vs. interesting ones)

They need to be distinguished from the sort of garden that someone might go and visit otherwise cities and towns could end up as a blob of indistinguishable green in places where most people have gardens.

Does access=private encapsulate the distinction we need already?

Also, adding a qualification tag to leisure=garden with the intent that specific values won't get rendered by some renderers looks awfully like the situation with disused=*, a pretty awful tag which has gathered a fairly resounding "no" in the past from implementers (see the bug reports). Could something like


represent this in a way that's more reasonable for implementers? How are you intending to ensure that this will be implemented? --achadwick 17:38, 4 October 2010 (BST)

The problem is that in most cases garden is only part of larger landuse=residential, so the residential=* would pose as the main tag on the garden area and that's kind of weird. --Xificurk 03:04, 16 November 2010 (UTC)

Update 26th Oct: I've reported this as a bug since it's a fairly big aesthetic concern for many city districts. ticket 3302. We'll see where that goes. Leaning towards access=private myself right now. --achadwick 11:56, 26 October 2010 (BST)

Also, long German-language thread on this very topic with more examples: --achadwick 11:57, 26 October 2010 (BST)

New vote and type addition

Hi everyone.

I am looking forward to a new vote for this proposal.

One thing I was hoping to get some feedback on is the addition of a new type of garden: the "community" type. This is something that is very common in lots of countries, and it would make sense to have a different tag to differentiate private residential gardens and community veggie patches that are open to everyone (most of the time run by an organisation or the local council).

Cheers Chtfn (talk) 02:25, 21 April 2014 (UTC)

Those are usually tagged as landuse=allotments, aren't they? The leisure tag seems redundant. --Xificurk (talk) 15:25, 21 April 2014 (UTC)
Not sure if landuse=allotments is exactly the same, as those are usually not really "community" operated but strictly divided by fences and hedges. Maybe there are also other types of community gardens? E.g. those gardens found in a monastery are usually community gardens (this is also where I see a potential problem: overlap with existing values). --Dieterdreist (talk) 17:06, 5 May 2014 (UTC)
I agree with Dieterdreist, the landuse=allotments tag does not cover small community gardens that are open and not divided in separate plots. The gardens I am talking about are open small-scale edible gardens that tend to appear in developped cities lately, to build a sense of community, educate and promote local production of food. Not sure how applicable this tag would be to other cases. Another issue could be to define what the limit is between community garden and inner-city farmland – maybe the difference between open and private? Chtfn (talk) 04:49, 25 August 2014 (UTC)


Can this set to approved status now since it has 40,685 uses by over 1124 and is listed on the main garden tag page? That all sounds like approval to me.

No. A proposal can only be approved by voting. The proposed keys are in use (as already stated on the respecive feature pages), but not necessarily with the same tag values and meanings as in this proposal.
Be careful to distinguish proposals (a mechanism to change tags) and feature pages (documentation of actual tag usage). --Fkv (talk)
Fkv, thanks for the distinction. I hadn't thought of that. Is there a way to update the proposal and have it voted on or is it essentially stuck in limbo and dependent on the original creator be moved ahead? I wouldn't really care that much, except I feel like it having proposed status at this point is a little miss leading for new mappers and people finding it in searches since it is so widely used. There should also be a point where a tag should move past proposed if it is so widely accepted but just doesn't have the thumbs up by three people on its discussion page. At that point voting is just superficial. Plus, it pollutes the proposal categories and makes it seems like its a lower rung tag when its not. Especially since its being used already on the main garden page. Maybe there should be an "in use" status for proposals that are in limbo like this one. Any ideas as to what can be done about it? --Adamant1 (talk) 01:12, 26 July 2018 (UTC)
The status should be set to "rejected". No need to restart voting, as the tags are already in use and documented. When new mappers are directed to a rejected proposal instead of the feature page, the cause can only be a bug in the search function or a wrong link. --Fkv (talk) 01:44, 26 July 2018 (UTC)
OK thanks. Just to be clear though, your saying if the proposals aren't eventually voted on then their are considered rejected? But if it can't be approved when it hasn't been voted on, how can it be rejected when it hasn't been voted on? If its considered rejected after a certain amount of time has passed without a vote, how much time should it be? I only ask because I keep getting attacked for doing certain things, this being one of them, and I want to be clear about what the standard is so I can avoid confrontations in the future. Thanks. --Adamant1 (talk) 23:47, 26 July 2018 (UTC)
There is a minimum of 2 weeks between RFC and voting (see Proposal process#Voting and no maximum. I think that 2 weeks are too short to get sufficient feedback, particularly for complex proposals which introduce multiple tags and keys and deprecate or change the meaning of existing tags. It is generally a good idea to leave a proposal open for some time and see how it works out in practice. Mappers are allowed to use non-standard tags as long as there is no conflict with standard tags, so they are also allowed to use proposed tags that have not yet been voted on. One example is Proposed features/information which has been open for almost 10 years without voting, and most of the proposed tags have come into widespread use since. One advantage of leaving a proposal open is that the proponent can still change some details that turned out problematic. One example for this is Proposed features/area:highway where some issues still need to be resolved.
When the proponent himself lost interest in the proposal (or at least when it seems so), it is possible to set the status to "abandoned". There is also "obsoleted" when the proponent abandons the proposal in favour of another.
Many proponents refrain from starting the voting period because only a handful of people use to vote, and most of these do not have an idea of what the proposal is all about. E.g. if the proposal is about climbing and none of the voters is a climber, the voting will lead to a random result. That's why some proposals were rejected although they were fine, and some flawed proposals were aproved. The sole advantage of an early voting is to get the proposed tags to the standards track quickly. However, some people achieve the same result by just writing (or rewriting) the feature page, silently bypassing the whole proposal process. Actually, nobody cares about proposals anymore. Most people do not even care to read the wiki. They just use the tags suggested by editor templates (in JOSM etc.) and validators, no matter if they conform to the wiki or make sense in any way.
--Fkv (talk) 08:41, 27 July 2018 (UTC)