Building overlay should not add `ruins=yes`

[HolgerJeromin]

How to Reproduce

• Building overlay
• select an building=yes outline
• select "ruins"

SC only adds ruins=yes, keeping building=yes.

Expected Behavior

Changed building=yes to
a) building=ruins or
b) demolished:building=yes or something like that

I am unsure what is better.

But just adding ruins=yes feels very wrong for a building which only is a bunch of stones now.

Versions affected

v62.1

[matkoniecz]

demolished:building=yes is for utterly gone building, not for one where remains are present

I would expect building=ruins and building=yes ruins=yes to have roughly the same meaning

But just adding ruins=yes feels very wrong for a building which only is a bunch of stones now.

in such case that is not a building anymore and I would expect ruined:building=yes to be used, in StreetComplete mapping you would need to create note

maybe having both "ruined building" with current image and "building gone, ruined remains of building visible" with new image would make sense?

[mnalis]

TL;DR: let's use ruins:building=yes as per lifecycle prefix instead of other problematic solutions.

---

I would expect building=ruins and building=yes ruins=yes to have roughly the same meaning

I definitely would not. For me, building=yes means there exists a building. I.e. (per wikipedia):

A building or edifice is an enclosed structure with a roof, walls and often windows, usually standing permanently in one place, such as a house or factory.

If there is only a pile of bricks in general shape of the former building (i.e. there no walls and/or no roof; i.e. they are ruins), that is IMHO definitely not building=yes, whatever extra trolltags¹ it might have in addition.

I'd retag it as ruins:building=yes instead, which is IMHO quite clear and much preferable to claiming it is still a building, and then adding trolltags like ruins=yes or operational_status=broken or has_walls=no + has_roof=no.

(I'd also avoid building=ruins due to the fact that most ruins of a building are not type of a building, but something completely different -- sham ruins in good standing excepted, maybe, but those are not what we are tagging here, but historic problematic ambiguity of that specific tag - see its wiki).

• IOW, this picture below is not a "building" of any kind by any stretch of imagination, but a ruin to be tagged with ruins:building=yes:

  

• Some ruins are in slightly better shape, but still definitely not building=yes but still ruins:building=yes:

  

• And then there are some which are in much better condition:

  

  That one would/might be acceptable to tag as building=yes + ruins=yes (although to me it is even a candidate building=yes + abandoned=yes, as it seems to me that it fits the description quite well: "man-made feature that have fallen into serious disrepair, appear to be unused, and can not be put back into operation without expensive repair, yet remain identifiable as the same feature." -- it wouldn't be cheap, but expensive repair (e.g. roof repair + reinstall windows / doors, some drying and new paint job) technically² might turn it into usable building again; so it is not quite "ruins" to me). But if someone were to tag it as ruins:building=yes, they wouldn't be too wrong either - those are ruins of a building for most practical purposes.

So, the only solution which fits all 3 images is ruins:building=yes IMHO.

in such case that is not a building anymore and I would expect ruined:building=yes to be used

ruins:building is more popular spelling of ruined:building AFAICT? But otherwise, yeah, I agree.

in StreetComplete mapping you would need to create note

which is exactly the problem and the reason for this issue, as far as I understand it? Variants of those first two pictures are IME more prevalent variant of "ruins", and current tagging does not suit them at all.

maybe having both "ruined building" with current image and "building gone, ruined remains of building visible" with new image would make sense?

It might work for SCEE (and would work for me generally), but I'm not sure about SC... Would the average user understand the difference between different "classes" of ruins? I'm not sure. Perhaps with a good text 🤷‍♂️

Footnotes

1. yeah, I'm aware I'm replying to inventor of the term 😃 -- it does not change the fact that the secondary tag like ruins=yes should not completely invert the meaning of the primary tag (like building=yes). ↩

2. it likely won't happen though; it is likely more economical to tear it down and build anew - while that would be more expensive then quick refurbishing (especially if you're not picky but a squatter), you could then sell apartments there for much higher price too. ↩

[matkoniecz]

TL;DR: let's use ruins:building=yes as per lifecycle prefix instead of other problematic solutions.

it is used more than I thought, see https://taginfo.openstreetmap.org/tags/ruins=yes#combinations vs https://taginfo.openstreetmap.org//search?q=ruins%3Abuilding%3Dyes is quite stark

which is exactly the problem and the reason for this issue, as far as I understand it? Variants of those first two pictures are IME more prevalent variant of "ruins", and current tagging does not suit them at all.

hard to say, not a native English speaker

If there is only a pile of bricks in general shape of the former building (i.e. there no walls and/or no roof; i.e. they are ruins), that is IMHO definitely not building=yes, whatever extra trolltags1 it might have in addition.

I agree and I would not use building=ruins either for it

I'd also avoid building=ruins due to the fact that most ruins of a building are not type of a building

+1, not a building

IOW, this picture below is not a "building" of any kind by any stretch of imagination, but a ruin to be tagged with ruins:building=yes:

+1, not a building

Some ruins are in slightly better shape, but still definitely not building=yes but still ruins:building=yes:

+1, not a building

That one would/might be acceptable to tag as building=yes + ruins=yes (although to me it is even a candidate building=yes + abandoned=yes

This one does not look like ruins at all to me from photo and is closer to abandoned=yes

But if someone were to tag it as ruins:building=yes, they wouldn't be too wrong either

Hard to say based on a single photo but it seems wrong tagging here

[matkoniecz]

I guess that I can link https://wiki.openstreetmap.org/wiki/User:Mateusz_Konieczny/ruined_building if someone wants even more text to read

[matkoniecz]

TL;DR: let's use ruins:building=yes as per lifecycle prefix instead of other problematic solutions.

do you propose to also remove building tag?

Note that it will be quite problematic in cases where ruin is still a building

[mnalis]

do you propose to also remove building tag?

Yes, for those first two pictures I propose removing building=* and adding ruins:building=* instead.

Note that it will be quite problematic in cases where ruin is still a building

Yeah, for that third picture that it is still a building IMHO too.

So I personally would tend not to consider that a ruin (i.e. ruins=yes or ruins: prefix), but (as noted) "merely" abandoned=yes added on in addition to existing building=*.

I.e. what the answer "Abandoned" already does:

---

TL;DR: I only propose changing that "Ruins (derelict, ruined and unidentifiable building)" answer to replace building=* with ruins:building=* (instead of keeping building=* and adding ruins=yes, which it does currently). Perhaps improve the description text if people find it confusing (for me, that 3rd picture is still quite identifiable building and not ruins)

(existing icons in SC answers look pretty good to me!)

[westnordost]

Well. Maybe. I get your point about being a troll tag and all. But, as long as its use is not advised against in the wiki or better deprecated, I see no reason why it should be changed in the app. As usual, the process should be 1. first discuss in OSM community, find consensus there, 2. document that in the wiki, 3. make changes to editors / editor presets

Let's not skip 1 and 2. Until that is not done, it makes no sense to change the behavior of StreetComplete only. See also https://taginfo.openstreetmap.org/keys/ruins#projects

[matkoniecz]

I.e. what the answer "Abandoned" already does:

And "Ruins" selection still refers to it as a building

If it is not a building anymore and it is pile of rubble then this selection is not applicable. And in practice OSM has quite wide definition of buildings. I would not put SC here on forefront of changing tagging schemas.