[OSM-talk] Lot's of locality names in an otherwise empty area

Andy Townsend ajt1047 at gmail.com
Mon Nov 21 12:49:39 UTC 2016

On 21/11/2016 11:42, Richard Fairhurst wrote:
> Sebastian Arcus wrote:
>> Well, looking at the map, it looks like each and every parcel of
>> land and section of field has a locality tag associated with it.
> It's very common in the UK, too, for uninhabited sections of woodland and
> hillside to have placenames.

... and fields, of course.  Where I was brought up the names in use were 
mostly just descriptive ("The Twenty Acre Field", "Piggy Thompson's 
Fields", etc.), but they were in OSM terms at least "loc_names".  Very 
few were verifable beyond "find a local old person and ask them" though.

However "names on a map" doesn't always mean "names of places". Ordnance 
Survey data in the UK is riddled with them, and some are little more 
than historic names.  Anything that's taken OS data on board without 
local vetting will share that problem.  As an example, 
https://en.wikipedia.org/wiki/Common_End,_Derbyshire was originally a 
"village" in wikipedia; it got changed to the curious "a place noted on 
a map" at 
when various people (including me, who has lived down the road for 30 
years) said "it's not actually a village!".

Obviously names change over time.  In the Common End case I suspect it 
was never much more than a farm, like Owlcotes to the north (another 
"place" according to OS maps).  Another example of that is here:


There there's a modern village ("Rossnowlagh") but two townlands 
("Rossnowlagh Upper" http://www.openstreetmap.org/relation/5625290 and 
"Rossnowlagh Lower" http://www.openstreetmap.org/relation/5625293).  
Those two were also imported as 
http://www.openstreetmap.org/node/5224127 and 
http://www.openstreetmap.org/node/52242180.  The "Upper" and "Lower" 
versions aren't signed on the ground and aren't villages any more 
(though likely once had significant populations); the modern village 
http://www.openstreetmap.org/node/2349484921/history I added based on 
survey, after checking with #osm-ie what best to do.

>> it still seems a bit odd - and begs the question if those tags
>> really need to be there.
> Why not? Be conservative in what you change/delete in OSM, be liberal in
> what you add.

Indeed - but there's no harm in asking the question, and as Colin Smale 
said yesterday, the logical people to ask, if you can't find a local 
80-year-old, are the people that added it.



More information about the talk mailing list