Talk:Key:building:condition

Verifiability

The value of this tag seems to be too detailed, leading to subjective conclusions. --Jgpacker (talk) 13:50, 23 October 2014 (UTC)

Can you give an example? I think that the definitions are pretty clear. Of course there is no clear line between e.g. partly_ruinous and mainly_ruinous, because degradation comes continuously over decades and centuries. It's just the same ambiguity as for tracktypes, meadow vs. heath vs. scrub vs. wood, etc. We have to classify things somehow. Cartography is an abstraction of an analog world. --Fkv (talk) 15:42, 23 October 2014 (UTC)

If I were to group this tag's values in similar-meaning groups, it would be: "preserved, renovated, reconstructed", "partly_ruinous, mainly_ruinous" and "completely_ruinous, recognizable_remains, ruinous". The first group could be said to be verifiable (not ambiguous) given the history of repair/reconstruction of a building.

Tags like tracktype=* and smoothness=* went through many discussions with people trying to get as close as possible to some verifiable criteria, but even so some people don't feel satisfied. Once in a while a discussion appears because of things like meadow vs heath vs scrub vs wood, and generally people prefer to have a more objective schema.

One thing I found odd is that more than 90% of this tag's current use contains values that don't match the documentation. Not sure if they already were like this when the page was created.

By the way, it was nice to see you were watching the page. Unfortunately most people don't (or simply don't know how to turn on notifications).

--Jgpacker (talk) 16:43, 23 October 2014 (UTC)

We can add some photos to support the descriptions. The problem with smoothness=* is that the photos contradict the tag descriptions instead of supporting them.

I was not aware that there are so many undocumented values, with "average", "poor" and "good" making up the vast majority of uses. These are indeed subjective. What is "poor"? I don't know! I also don't know what building:condition="baik" (forth most frequent value) or "Baik" (fifth most frequent, in spite of the uppercase letter) means. "baik" does not seem to be an English word. Maybe Indonesian?

The wiki should normally document the most frequent values, but in this particular case it would be an ugly task. I'm going to contact Walter Schlögl (Walterschloegl), who is the original author of this documentation (originally located on castle_type:de=*). I hope he'll find some time to participate in this discussion.

--Fkv (talk) 20:33, 23 October 2014 (UTC)

Thanks to fkv for taking up this documentation. I can tell you the history about my usage of this tag. When I started to edit castles and ruins I surged for a tag to document the condition of a ruin. At that time building:condition was used undocumented with the values good / average / poor which might fit for a building which is not a ruin. So I added my own values which fit more for a castle and/or ruin. I don't know what's the difference between average and poor but I would classify a building if it's still possible to life in it or not, although this would depend also on the country it is in. Here my definition of the ruinous values

• recognizable_remains ... you can find some foundations of a wall but no walls any more
• completely_ruinous ... there are some walls left but no rooms, no roof
• mainly_ruinous ... most part of the castle only consist of some walls but there is at least one part with a roof left.
• partly_ruinous ... most part of the castle consist of buildings where the roof is still there but there are also parts with walls only.

In case it is partly_ruinous many mappers would map both parts separately and tag one part as completely_ruinous and the other part according to the condition, at least if the preserved part is still in use. --WalterSchloegl (talk) 16:01, 24 October 2014 (UTC)

What if windows are broken, or there are small holes in the roof? I have been using partly_ruinous for these cases. When the roof is completely down in some part of the building, I would consider it mainly_ruinous, because the building as an entity is no more habitable, and it is certainly dangerous even in those parts where the roof is still in place. --Fkv (talk) 16:25, 24 October 2014 (UTC)

This definition sounds not so bad, I could also aggree to that.--WalterSchloegl (talk) 19:43, 24 October 2014 (UTC)

I'm not sure if that would completely remove possible ambiguities, but if a description like this is put in the page for the values, I would be ok with removing the claim that this is not verifiable (i.e. the {{Verifiability}} template). --Jgpacker (talk) 21:30, 24 October 2014 (UTC)

Big Picture

The English Wikipedia has a less detailed classification that would make a lot of sense in OSM in relation to castles (at least on a higher abstraction level) but we have to deal with a lot of (evolved) fragmented tags here. The building:condition=* proposal somewhat overlaps with ruins=yes and the abandoned:=* namespace. And there is the case of Burgstalls or completely lost castls where (almost) nothing remains. It seems a bit counterintuitve to map these as historic=castle since there is also historic=ruins which seems to be better fitting but even that seems to be rather optimistic in many cases. With the current documentation I feel completely lost what to do for such places. --Stefanct (talk) 05:49, 15 August 2020 (UTC)

The other tags are not as specific or have different meanings. abandoned=yes means that the object has been disused for a long time and left to decay but it is not necessarily ruinous yet. abandoned:foo=bar is a tagging scheme to keep old tags visible to mappers while hiding them from applications. (It's equivalent to note=*.) historic=ruins is essentially deprecated, as it doesn't tell what the ruins originally were (a castle or a cowshed or something that wasn't even a building), and it was never rendered in the standard layer. It's also unclear whether you should tag every individual building with it, or the whole site altogether. ruins=yes I do use for benches, boards, hunting stands etc., and it can also be used on buildings as well as castles (which may consist of multiple buildings), but it's obviously less fine-grained than building:condition=partly_ruinous/.../recognizable_remains.

Let's take an example: You have a castle (historic=castle + castle_type=defensive + name=* + maybe ruins=yes) with strong walls (barrier=wall) and 3 buildings inside: a keep (Bergfried) (building=tower + building:condition=preserved + tourism=viewpoint), former apartments of the lord, with no more roof (building=residential + building:condition = completely ruinous) and remains of the walls of a chapel (building=chapel + building:condition=recognizable_remains).

--Fkv (talk) 08:50, 15 August 2020 (UTC)

That's a very subjective view of yours and I strongly disagree with most of the first paragraph (also I think you are confusing abandoned:=* with disused:=* - the former indicates some form of degradation (which could be specified more precisely with an additional tag like it is proposed here)). The german wiki of historic=ruins completely contradicts the english one because it is super outdated - I'd rather argue to not use ruins=yes since we have mostly overcome such broken tags. Tagging historic=ruins, ruins=castle is way more OSMionic than slapping a ruins=yes on anything.

I am confusing nothing. historic=ruins + ruins=something won't work, because you cannot set ruins=building=shed or ruins=tourism=information=board. You can only set ruins=shed or ruins=board, but then you lose all the structure. --Fkv (talk) 21:57, 15 August 2020 (UTC)

Most importantly though, you seem to ignore the whole issue I am trying to convey: the whole mess of various tags for lifetime specification needs to be resolved before something like this proposal makes any sense - it would actually make things worse.

--Stefanct (talk) 09:30, 15 August 2020 (UTC)

This is not a proposal. It is a feature page, documenting current usage of this tag. I wrote it because the tag was in use but not documented. It is not the right place to lament over the mess that other life cycle concepts are. Here we should only discuss the mess that the building:condition values are. This mess is the consequence of some people introducing this key via mass edits without writing a proposal first. Now the damage is done and we must try to make the best out of it. Replacing building:condition with historic=ruins etc. is out of the question in view of the information loss. --Fkv (talk) 21:57, 15 August 2020 (UTC)