New Quest: Can a water bottle be filled here?

mcliquid commented 7 months ago

General

Affected tag(s) to be modified/added: bottle Question asked: Can a water bottle be filled here?

See also rejected SC quest: https://github.com/streetcomplete/StreetComplete/issues/4110

Checklist

Checklist for quest suggestions (see guidelines):

[ ] 🚧 To be added tag is established and has a useful purpose
[x] 🤔 Any answer the user can give should have an equivalent tagging (SCEE quests may not be answerable in all cases, but please keep this rare)
[x] 🐿️ A survey is necessary (may require knowledge of the subject, though easy answering for everyone is preferred)
[x] 💤 Not an overwhelming percentage of quests have the same answer (No spam)
[x] 🕓 Applies to a reasonable number of map data (Worth the effort)

Ideas for implementation

Element selection:

nodes, ways with
          (amenit~drinking_water|water_point or man_made~water_tap|drinking_fountain)
          and access !~ no|private
          and !bottle

Proposed UI:

Recycle the AddWheelchairAccess quest & UI
Simple "Yes, No and Limited" answers

Another option would be an Image Selection. I could prepare three images which shows a 1 liter bottle (classic bicycle bottle) in the three situations.

Helium314 commented 7 months ago

Yes / no / limited would be fine, if there is some description what limited means.

mnalis commented 7 months ago

While I love microtagging water sources, I'm not particularly fond of bottle tag as it is very poorly and unclearly defined (see the wiki and SC discussion), and so different people have conflicting ideas what yes and no actually mean...

But if it is going to be written anyway, I'd advise to at least also add Designated for bottles only answer (to retag as clearly defined fountain=bottle_refill - which is probably what huge number of those bottle=yes actually was intended to mean - as I have rarely if ever found water source that cannot be used to fill bottle if one is persistent enough)

mcliquid commented 7 months ago

Do you think it would make sense to follow up the definition with a proposal or do you see this as rather hopeless?

mnalis commented 7 months ago

Do you think it would make sense to follow up the definition with a proposal or do you see this as rather hopeless?

I (personally) generally do not think that trying to redefine meaning of already popular and widely used tags (like bottle=yes / bottle=no) is likely going to be successful. The additional problem with bottle=yes is that not only it was not clearly defined, but also that some contexts it has well defined, but completely different meaning (e.g. for fountain=bottle_refill, it indicates instead whether "empty bottles are available for purchase" instead of "water bottle can easily be filled"), and that it is highly subjective (what does "easy" mean in that definition?)

If one cares greatly and is willing to invest a lot of time, it might have better result, though. It would need research, contacting many users of the tag, starting a discussion about that (linking to all related issues, tickets, discussions, wiki talk pages) etc. and presenting all that data to even get the discussion started.

E.g. for https://wiki.openstreetmap.org/wiki/Key:camp I've manually contacted few dozen top users to get their feedback on what they meant actually there. (and I still haven't finished that compilation of answers almost a year later, and it has about 20x less uses than bottle=yes; although that is mostly attributable to my chronic lack of time).

What would IMHO be quite more likely to succeed (and would be lot less work, although still not insignificant by any means) is starting proposal process (and linking it to all the places mentioned above) to suggesting additional clearer values (like e.g. bottle=impossible, bottle=hard, bottle=easy, bottle=only) - and providing picture examples with explanations why some picture fits and doesn't fit into some category. As additional benefit it would provide even more details. So if you're willing to invest some time into that tag, that is the route I'd recommend.

While that still has some issues with subjectivity / verifyability, those would be greatly reduced with picture examples and more detailed descriptions what fits which value (see e.g. smoothness=*, track_type=* for examples of tags which managed to have useful function even if they are somewhat subjective) and via extended discussion on e.g. forum tagging category.

mcliquid commented 7 months ago

Thank you very much for the detailed answers. At this point I am closing this issue as I realise that it is not worth the energy.

Helium314 / SCEE