[Tagging] [OSM-talk] Should we map things that do not exist?

Mateusz Konieczny matkoniecz at tutanota.com
Mon Jun 8 10:28:47 UTC 2020

Jun 8, 2020, 11:39 by dieterdreist at gmail.com:

> Am Mo., 8. Juni 2020 um 11:20 Uhr schrieb Mateusz Konieczny via Tagging <> tagging at openstreetmap.org> >:
>> On 6. Jun 2020, at 00:04, Volker Schmidt <>> voschix at gmail.com>> > wrote:
>>>> I do object strongly to the invitation to remove the razed/dismantled-railway tag in the case of railway tracks have been replaced by roads with the same geometry.
>>> +1
>> Add I have no problem with removal of them.
> this is fine, we do not have to share opinions on everything. But we should be cautious to not misrepresent community consensus in the wiki. It doesn't appear to be an universally shared conviction that you can remove these objects of which the traces are less evident than of other things.
Can you edit wiki or link problematic page and quote text that should be changed?
I am not sure is it referring to Wiki in general or one of my edits (and yes, some of my edits
on Wiki/OSM make them worse - it is not really possible to completely avoid mistakes).

>> I see the point in cases of ones where there are no traces
>> but road geometry makes 100% clear that railway was there
>> and probably would not delete them.
>> I see point in cases where track of former railway is marked/
>> memorialized/etc and that info is mentioned on OSM object.
>> But for cases where new road is matching geometry but shape 
>> is not recognizable at all as track of former railroad, and
>> historic maps are needed to recognize it?
> there aren't only written/drawn sources by the way. Oral tradition can also be relevant. Your grandpa told your dad and your dad told you, why not? 
I am not against oral sources for a current data! Mapping housenumbers, street names,
peak names based on an a trustworthy oral source is perfectly fine.

But remembering about object is not changing whatever it is mappable.

Lets say that there was a castle and was replaced by a sport pitch, and place looks like
this nowadays (a theoretical example):

Castle is remembered. Is such castle mappable? In my opinion would not be as there are
no identifiable traces (possibility of archeological excavations are not really changing this).

>> Deletion should happen, OSM is not for objects that are gone.
> we were discussing objects which left traces. "are gone" is not very precise, do you intend to include things which aren't operational but traces are there? Or does is mean "left not traces whatsoever"?
Obviously, we can map railway that are not operational and left clear traces.

We are generally OK with mapping things where some traces remained.
It is accepted that thing totally and completely gone are not mappable.

"left not traces whatsoever" seems to step - if you look sufficiently carefully
many things left some sort of traces (see historical ocean cases where I thing
everyone agrees to be not mappable in OSM - and there are clear traces of them!).

are clearly mappable 

But mapping railway in
would be too far for me and I would delete it as a historic mapping if mapped (example
from my city so I know that former railway there is not identifiable without access to old maps).

>> https://en.wikipedia.org/wiki/Western_Interior_Seaway
>> https://en.wikipedia.org/wiki/Ural_Ocean
>> https://en.wikipedia.org/wiki/Sundance_Sea
>> and other >> https://en.wikipedia.org/wiki/Category:Historical_oceans
>> also left traces and are not mappable in OSM
> I agree that these are not very compatible with OSM, similar to current geological features, and even current oceans aren't very well represented in OSM. They aren't good examples for discussing dismantled railways though. They are former natural features, of millions and billions years ago (i.e. all the coastlines were completely different), while railways are man made features.
I intended them as example of something that is not mappable
despite leaving indentifiable traces. 

I agree that it is a bit mixed with poor mappability as a large scale geological feature.

I think that mapping former moat/canal in my city would be a better example

- it is 100% gone (replaced by a road)
- some traces remained (road follows course of a former moat)
- it is not recognizable as a former canal based on a current state
- though archeological/geological works would allow to prove that it was there
- well documented as a former canal in historic sources

Is this former, gone, not recognizable canal that left some traces still mappable in OSM?

In my opinion - clearly not, so I would go that with "recognizable traces" threshold. 

For reference: this bridge was constructed as a  bridge over a canal. Now it is a
bridge over a road.

location in OSM: 

Wikipedia article in Polish about that waterway:
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20200608/5cd884db/attachment-0001.htm>

More information about the Tagging mailing list