|
I ask because council Ward names seem like an important source of local place name information in particular. Wards are arguably administrative subdivisions, in a more active sense than Parliamentary constiuencies - they are not just about elections, but councils' service delivery is often based on wards, for instance. |
|
Generally we want verifiability, ideally on the ground. So anything where you can go to the place and see it with your own eyes (and which has some degree of permanence) is ok to map. Something that is not verifiable at all should not be mapped. Something that is verifiable, but only with recourse to third party sources, may be mapped but it requires a good reason to do it. If the data cannot be surveyed by OSMers, then one of the big strengths of OSM - that we have lots of people with local knowledge who can help keep the data accurate - is out of the window. Adding non-surveyed data to OSM purely because "someone might find it useful" is not sustainable (there is simply too much geo-related data out there to add it all to OSM). In most cases, if someone finds third-party data useful, they can get it from that third party, and OSM should not be abused as a "distribution channel" for third-party data. If third-party non-surveyable data is to be kept in OSM then there must be a very good reason for this particular kind of data. Administrative boundaries are one category where we've generally said: This is so useful even for us mappers, let's include that data in OSM. I can't say if this applies to ward boundaries but the question you should probably be asking is, can they aid other mapping work? |
|
Is there any particular reason to include them? The data is already available from Ordnance Survey under a more permissive licence than OSM can provide. OSM can't improve on what is already authoritative data - we can't suddenly decide that ONS have got the Charlbury & Wychwood division wrong and that it should include Leafield, for example. OSM data editing standards would dictate that such objects would need to share nodes with coterminous objects, such as roads and rivers. That would make it harder for new users to edit those roads and rivers (e.g. to split at a bridge) without accidentally breaking the ward relations. If you need to analyse OSM data in the context of the boundaries, just import OS Boundary-Line into your PostGIS database and run a query across the two of them. Not everything needs to be in OSM. I can see the maintenance issue for sure. My reason was really driven by what people use maps for, what they might typically want to find. Council ward names are often names of localities, for instance. Some of these are missing, because they're names used rather than always written in areas. Another example is the local area (quarter) names in central London. These have been added in some cases, but not all,, eg "Blackfriars" and "Aldgate". There are however around 25 in total, now used in City fo London Ward names, such as 'the Cheap" or "Barnard Castle". Some of these are also reflected in street names, rather than tube stations,, but that's part of the story, not the whole of it, as these are areas, not streets. It seems a pity to leave off these known names, shown on local signage etc. I am sure they could be added manually though.
(25 Jan '21, 23:47)
Jim Killock
I think wards in the City of London are completely distinct from political ward boundaries ) (https://en.wikipedia.org/wiki/Wards_of_the_City_of_Londonelsewhere in that they are fairly permanent, but names such as the "Cheap" and "Castle Baynard" (not Dominic Cummings favourite place) are names of wards not of localities in OSM. If these can be ascertained by on-the-ground mapping then they are fair game (boundary=political). Wards regularly revised by the Boundary Commission really are better dealt with as Richard suggests
(26 Jan '21, 12:34)
SK53 ♦
Thanks, the names of London wards can certainly be found "on the ground" (they have noticeboards for instance) but the boundaries, I doubt.
(26 Jan '21, 23:57)
Jim Killock
|
|
I think any accurate data has a use. We may not see what that use could be. But someone may well find a way to use it in a very clever way. One concern I have is there a storage limit to what OSM or to be precise the third parties will store for us. |
Council Ward names are an absolutely lousy source of place names. First of all ward names may persist through multiple boundary changes. Second, wards change regularly so the connection with a ward name & the associated area is loose at best. Third, wards in rural areas are so large they will encompass multiple place names.
See also the recent assignment of names to MSOAs by HoC library for the difficulties of choosing a suitable name for an arbitrary area chosen for a specific purpose (census stats in this case, but same considerations apply to wards). They
The one I know best in this respect is Aspley Ward in Nottingham which has a very loose relationship with the two places called Aspley (a suburb & a large 2000+ 1930's council estate). Much of Aspley is not in Aspley Ward: the wikipedia article is a hopeless mess on this front as it chooses current Aspley Ward boundaries which a) wont last much longer and b) are completely different from the last 2 incarnations of that ward.
In general OSM in the UK is pretty well provided with place names of the neighbourhood, suburb variety, so if you find something missing check first with the local community.