I’ve been looking at brick bonds recently and started nano mapping them on buildings and walls. I naturally differentiated between building:material and building:facade:material, but now it was pointed out to me that the building:material is documented as “what you see in the facade”. This is a less than ideal approach for several reasons, in my opinion:
sustainability: concrete is a far less sustainable1 building material, but most buildings are now built from concrete, because it is less labour and cost intensive than for example bricks.
related to that - historical authenticity: The time of building in brick (for example) has passed in most of Europe and Northern America due to cost efficiency. I would prefer to have historical/ architectural truth reflected in the map.
Against the argument that a building:material different from what you see is in the facade, I would say that for one, you could have watched the building go up and/ have it document it on street-level imagery. 2nd, if a facade is built in stretcher bond or stack bond, you can be fairly sure that it is not built from brick, because those are not very supportive bonds; stretcher bond is however the easiest bond to use for a facade, because the bricks will all be flush behind the facade. 3rd, if all you see is render, surely you wouldn’t expect the whole building to be made from render.
demo.f4map.com already interprets building:facade:material correctly. If anyone was worried about the rendering (no pun intended) on 3D maps - which they shouldn’t in the first place - they can be rest assured that it’s taken care of.
I have no solution on what the fall back should be for “I see a material in the facade, but I’m not sure whether it’s just the facade or the material for the whole building.” It should be building:facade:material, but it would need a lot of retagging (2,000,000+ uses of building:material, but some could be correct, of course).
Yes, that’s what the key means. We defined it such fourteen years ago in 2012, and the use case we had in mind was indeed 3D rendering.
As mappers and developers interested in 3D, we’re still heavily relying on that key and this use case appears to be more popular than ever. So if you want to start mapping some new kind of information, such as the building’s structural material which may not be visible from the outside, please don’t misuse the existing key contrary to its definition and intended purpose. Instead, make up a new one or check if the existing building:structure key with values such as brick, cement_block, engineered_reinforced_concrete and steel_frame fits your use case.
If we wanted to change the name of the building:material key, the problem wouldn’t be adding support for a new key to 3D renderers. The problem would be migrating the existing two million tags to use the new key. This would not be feasible without a large-scale automated edit. And there is usually a lot of reluctance to perform changes to tagging standards which don’t add new capabilities but merely make key names nicer.
Since an incomplete attempt to start a migration (such as calling on people to just start using the key with a new meaning without taking care of the existing data we’ve collected over the past fourteen years) would be quite harmful to the 3D rendering community, I’d very much like to avoid that outcome.
I agree with your post, we should not change the definition of building:material as it is established for a long time, although it could have been choosen better to make clear this is about the outer visible shell and not the material of the building in a structural sense.
building:structure seems generally fine, but it is kind of odd, brick and concrete blocks are distinguished rather than using a structural term like “masonry”. I looked it up, the suggested values are indeed reinforced_masonry, confined_masonry, unreinforced_masonry (i.e. there would be some room for building:structure:material with values like brick or concrete_block for masonry).
I would also suspect that “steel_frame” is used for all kind if steel structures, not just buildings whose structural connections are designed as frames, should this be detailed in the wiki? There are some very few uses of “steel” (like 21 globally)
Note that this is an easy part, trivial to do. The harder part is convincing people that it should be done, convincing users to stop using it, change software, change documentation…
In practice you can convince community to stop using some tag but it will not make it free for different use anyway, as old one will keep occasionally appearing. And tag with so significant use will linger for long.
And if people decided to use tag this way they are likely to independently do it in future. Better to invent a new value.
Though tagging internal structure runs in some verifiability issues.
I don’t think the interesting buildings are the one with stretcher bond, where it looks like it’s built from brick, but is actually concrete blocks and just the facade is brick. The interesting, historic buildings are the ones where the building:material is in fact brick, facade and structurally, so it’s not a huge problem for me personally.
And since f4maps is able to understand and render either, it’s even less of a problem.
It’s just not reflecting the ground truth using something called building:material, when it is in fact only the facade.
I don’t think the interesting buildings are the one with stretcher bond, where it looks like it’s built from brick, but is actually concrete blocks and just the facade is brick. The interesting, historic buildings are the ones where the building:material is in fact brick, facade and structurally, so it’s not a huge problem for me personally.
what about wood frames filled with brick, is there suggested tagging ?
There’s building:material=timber_framing, 5,839 uses. It’s documented on the Wiki page as
timber framing
This tag is used to describe the structure, rather than the facade.
See also building:structure=*
I don’t understand that comment. Is it meant to discourage readers from using the tag building:material=timber_framing? Is it asking mappers to use the tag but only when the building is actually timber-framed, as opposed to a modern imitation? Is it warning data consumers that the tag may have been (mis)used for buildings where the timber framing isn’t visible on the outside?
Can we agree that building:material=timber_framing is appropriate for the building:material tag of a half-timbered house, where the timber framing is externally visible? Including maybe also false half timbering that is only decorative but doesn’t serve a structural function?
There’s also building:architecture=timber_frame, 3,718 uses, which doesn’t seem to be documented on the Wiki.
I don’t understand that comment. Is it meant to discourage readers from using the tag building:material=timber_framing? Is it asking people to use the tag but only when the building is actually timber-framed, as opposed to a modern imitation?
I think this is to emphasize that it is about the structure which could be visible or hidden behind plastering, and not necessarily about the facade
Yes but that directly contradicts how the key (building:material) is defined…
The key has mostly been used in Germany (4,852 of 5,841) and the German wiki page, where most German mappers will be looking first, doesn’t have this confusing sentence. It just says “Fassade aus Fachwerk”: half-timbered facade. So I think we should just change the English wiki to say the same.
