OpenStreetMap
A Suggestion to Fix Poor LSN in the UK
Posted by alexkemp on 20 August 2016 in English. Last updated on 8 February 2019.This is a research document; it is going to attempt to explain:—
- The fundamental basis on which Location, Search & Naming (LSN) facilities in OSM work
- Why those facilities fail for a substantial part (40%) of the UK
- How to fix it
You need to know that the writer has been mapping for only 5 months, and therefore only part-understands what he is talking about. One (possible) advantage is that his is a fresh eye, plus he has the ability to think for himself. As the writer enjoys stories, much of this will be presented in that form.
- Local government in England:(counties, districts and unitary authorities) (pdf)
- Local government in England:(Parish and town councils) (pdf)
- Counties, Non-metropolitan Districts and Unitary Authorities + GSS (ONS archive)
On Thursday 9 June 2016 I began to map outside of my home patch in Nottingham NG3 and met the Boundary marker which, since 1877, has marked the Boundary Line between the City of Nottingham and Gedling, and also between NG3 & NG4. I was now heading for Carlton, Gedling.
One feature that had been common throughout my NG3 mapping was that LSN had consistently failed with OSM. When I was mapping close to St Anns OSM said that I was in Thorneywood, and so on. By the time that I reached Carlton I’d gotten the basic map methods under my belt & could pay more attention to the condundrum of the fact that when I was working in Carlton (a Suburb) OSM said that I was in Bakersfield (at the time a Neighbourhood, but now a suburb), or even Thorneywood (another Neighbourhood).
Practical examples can focus the mind, and this post is typical. I placed the Diary arrow in Highfield Drive, Carlton, but the result was: “Location: Thorneywood, Sneinton, Nottingham, East Midlands, England, United Kingdom”, which isn’t even the correct District. That was just embarassing.
Eventually it became clear that OSM was giving precedence to Neighbourhoods over Suburbs, which seems perverse, and I was deep into conversations with Will & Jerry (my nearest active, senior mappers) on how to handle it all. One of the results of those conversations was Bakersfield being changed from a Neighbourhood to a Suburb, which helped a bit, but that was only a minor part of the truth. A clue was given when a commenter (somewhere) said wrt LSN that “areas are important, not nodes”.
On 16 July 2016 I met, documented & mapped Nottingham’s Unparished Areas.
How Location, Search & Naming (LSN) facilities in OSM work
It starts with admin_level=10 BoundaryLine areas. If they do not exist, then OSM will do the best that it can but, as documented above, often fails miserably. If they do exist then, as longe as yur speling is gud, then OSM will be able to find your search item and/or locate where you are and/or name that locality accurately.
Why Location, Search & Naming (LSN) facilities in OSM do NOT work
Whilst there are ~10,000 Civil Parishes (CPs) in England (which is what an admin_level=10 BoundaryLine area is documenting), a very substantial area of the country is unparished (referred to as an “Unparished Area”) (see all the “unnamed areas” in the Civil Parishes page). I’ve spoken to a very helpful & knowledgeable lady at the Local Government Boundary Commission (0330 500 1525) (hello Joe) who, to my dismay, confirmed that Unparished Areas are the Black Hole of the Boundary world. It was ‘dismay’ because many mappers have a cast of mind which insists that, if the authorities do not recognise it, then neither will they. That means that they do NOT want an admin_level=10 BoundaryLine area to exist for the Black Holes, and that means that LSN will never work for those areas. Oh dear.
Locally to me, the cities of Nottingham, Derby, Stoke-on-Trent and Leicester are Unparished (plus my home town of Hull), as also are Arnold and Carlton (one non-parish) plus Beeston. Notice that the last two are each named, even though they (supposedly) do not exist. St Anns and Thorneywood are both in the City of Nottingham whilst Carlton is within Arnold and Carlton. Thus, none of those 4 neighbourhoods/suburbs was at that time within an admin_level=10 BoundaryLine area, and my assertion is that is the reason that LSN features of OSM were inoperative for them.
The previous paragraphs are mixtures of anecdotal & written evidence, as I did not realise at the time why it was going wrong, but much of it did get documented due to my diary entries. The following is fully evidential:—
Nottinghamshire Civil Parishes - names for unnamed areas: the location arrow for that diary entry is placed in Slab Square, Nottingham. I moved the arrow a tad on 2 occasions when the BoundaryLine setup changed to discover the effect. These are the three location results + the BoundaryLine setup that applied at the time:
- (Original)
Location: Nottingham (Unparished), City of Nottingham, Nottinghamshire, East Midlands, England, United Kingdom
‣ admin_level=10 name=Nottingham (Unparished)
‣ admin_level=6 name=City of Nottingham
‣ admin_level=6 name=Nottinghamshire - admin_level=10 relation removed
Location: Lace Market, St Ann’s, Nottingham, East Midlands, England, United Kingdom - “City of Nottingham” relation changed to admin_level=10;8;6
Location: Lace Market, Nottingham, Nottinghamshire, East Midlands, England, United Kingdom
Searching for ‘Nottingham’ in OSM gives:
CityNottingham, Nottinghamshire, East Midlands, England, United Kingdom
Notes: The admin_level=10 relation was removed without consultation (as I understand it, this is considered by OSM to constitute abuse). Nevertheless, I took the opportunity to test out the effect of those changes, as shown above.
The Nottingham UA relation was changed by me to “admin_level=10;8;6” at 5am in the morning and returned back exactly as before shortly after completing the tests within the hour. That seemed a perfectly innocent action to me, but apparently not.
How to Fix It
All the evidence suggests that no admin_level=10 BoundaryLine area means no accurate LSN in OSM. The simplest fix to my mind is to enter a BoundaryLine for each Unparished area, but that is proving controversial. Even more controversial is my suggestion that, if folks want to discriminate between a CP and an Unparished area, then use designation=civil_parish for CPs & designation=non-civil_parish for Unparished areas. designation is an acceptable key for these folks, but both values are not, even though one non-standard value (and many others) has been promoted on a wiki for many years.
The main problem comes with Unitary Authorities (UAs). It is possible that, with a bit of tweaking, that the UA admin_level could be set to let it work. But maybe not.
Contra-Indications: Completing a recent Diary entry I set the pointer between Valley Road and Prospect Road. That location has an admin_level=10 area set. Saving the post, OSM said it was:— > Location: Thorneywood, Sneinton, Nottingham, East Midlands, England, United Kingdom
With Nominatim it seems, at times, that no-one actually knows how it all works. I’ve presented here my discoveries as I got my legs under OSM’s table across the last 5 months, plus a simple method to fix it. My pessimistic belief is that no-one will be able to agree on a solution, and my ignorant, blundering efforts so far seem to have only provoked personal abuse and sleepless nights.
Parish Councils + Town Councils
The Ordnance Survey refer to a “Civil Parish (CP)” in their .shape files, but the Parliament Research Paper refers to:
- Parish Council
- Town Council
- Neighbourhood Council
- Community Council
- Village Council
- City Council
The above are the different styles that such bodies can freely adopt.
1894: Parish Councils Act : Parish and Town Councils first establishment.
1972: Local Government Act : source of modern legal foundation.
Powers: In theory Parish & Town Council powers are identical to those of a District. However, they mostly do not have the resources to do anything more than Leisure Services.
County Council: (admin_level=6) : Education, Highways
District Council (admin_level=8) : Bin collections, Cemetaries, Parks & Boundaries
Parish Council (admin_level=10) : (everything else)
Unparished Area (admin_level=??) : parish powers handled at the district level
Coda: The Scale of the Problem
Clearly, we need to know how much of the country is affected by this issue. I’ll publish the full figures in a separate Diary entry, but here is the bottom line:
Parished: 11,329 Parishes; Average area/Parish=1,253 hectares; Total area=14,199,250 hectares (61%)
Unparished: 3,069 Parishes; Average area/Area=3,007 hectares; Total area=9,229,902 hectares (39%)
Discussion
Comment from TomH on 21 August 2016 at 07:58
The fundamental problem is that unparished areas do not have well defined boundaries other than where that abut a parished area, and no two people will agree on exactly where one suburb ends and another one begins - indeed in many places in London there are multiple overlapping names for example.
On top of that any suburb boundary is definitely not an administrative boundary because it does not define an area of administrative control.
So we probably shouldn’t add these areas where they are not well defined and verifiable (which I think is pretty much the default) and even if we do that shouldn’t be administrative boundaries, they should be some new thing like a suburb_boundary or something. Yes search would then have to be taught to process that but good data always comes first in OSM and we fix the consumers to understand new data rather than abusing the data to suit the consumers.
Comment from alexkemp on 21 August 2016 at 12:41
Hi TomH
Suburbs: as I understand it, suburbs appear under OSM only as nodes and, therefore, by definition cannot have a BoundaryLine.
Unparished Areas: oddly, these are all as carefully defined by the Ordnance Survey as are the civil parishes, and therefore are all perfectly verifiable. They are also all named by their respective District councils, and by those that live within them (I mean, can you imagine: “Where do you live?” “Oh, I live in Unnamed parish 9346”).
The issue, Tom, is that because the OS treat them as black holes some folks insist that OSM must do the same, which then shafts LSN. That seems nonsense to me.
Comment from Colin Smale on 21 August 2016 at 12:43
This statement is untrue: Powers: In theory Parish & Town Council powers are identical to those of a District. However, they mostly do not have the resources to do anything more than Leisure Services.
Parish and Town councils have very limited powers and are in no way equivalent to a District. This also applies to large town/city councils such as Salisbury - which is simply a Parish Council with a fancy name.
A Civil Parish council can pretty much unilaterally choose to style itself as a Town, Village, Community or Neighbourhood Council. The differences are only symbolic, there is no change to the powers of the authority. City status can only be bestowed by the Crown however, you cannot just call yourself a City. Similarly with Borough status - this is simply a binary flag which a given LA either has, or has not. Neither City nor Borough status changes the fundamental powers of the council.
You didn’t speak to the Boundary Commission by the way, but to the Local Government Boundary Commission for England. The Boundary Commission deals with parliamentary constituencies and their constituent wards, and the LGBCE deals with local government district boundaries and electoral wards. Parish boundaries and their warding arrangements, together with their electoral situation (parish meeting or parish council, grouping etc) are however an internal affair for the parent District/Unitary concerned.
Parishes, and Unparished Areas, are of little use for many use cases as they are not involved in postal addressing and are frequently divergent from local perception. However there are occasions when you need to know the responsible authority for a given location, including not only local government but also things like emergency services and health care provision.
Comment from Colin Smale on 21 August 2016 at 12:59
Alex, the OS is not the owner of these boundaries - they are just the messenger. In their data they do not refer to “unnamed parishes” - there is no such thing. The polygons we are talking about here are “filler areas,” which do not have an official name (and indeed no name is supplied by OS).
It appears that your most fundamental worry is that LSN is not working (properly). I am not 100% sure but I believe that Nominatim is not officially a core part of OSM, in the same way that Mapnik (which renders the map tiles) and its governing style sheets are not. Maybe they can be considered examples of good practice, or possibly even reference implementations. The normal attitude if you consider it broken is to (offer, help to) fix it. I agree with you that there is room for improvement here, but the Right Way (and in this case probably the Hard Way) to address it is to improve Nominatim, not to work around these shortcomings by manipulating the underlying data to produce the results you expect.
Comment from alexkemp on 21 August 2016 at 15:34
@The Maarssen Mapper: You state “This statement is untrue: Powers: In theory Parish & Town Council powers are identical…”.
That statement is drawn directly from the Parliament Briefing paper (“section 3. Parish council powers”). Think again.
Comment from alexkemp on 21 August 2016 at 16:21
@The Maarssen Mapper: You state: “the OS is not the owner of these boundaries” and “In their data they do not refer to “unnamed parishes”.
I’ve re-read what’s written just to be certain, and I never ever refer to the Ordnance Survey as the “owner of these boundaries” (nor think of them that way), so I’m uncertain what you are talking about there.
I never say that the OS refer to “unnamed parishes”. I say “see all the ‘unnamed areas’ in the Civil Parishes page”. I’ve got to call them something so that folks know what I’m talking about. You have named each one “Unnamed shape ….” so that is the name that I’ve used.
Hmm. I see now that you have now removed each file referred to in each page. Having problems?
Comment from alexkemp on 21 August 2016 at 17:30
@The Maarssen Mapper: thanks for the correction re: the LGBCE; I’ve altered the relevant phrase.
Comment from Colin Smale on 21 August 2016 at 19:23
What I meant by OS not being the owner of this information, is that they do not define the boundaries, nor the names of parishes etc. They are formally informed of any changes through changes in secondary legislation and formal notifications from competent bodies.
Re: “unnamed parishes” - I was responding to your reply to TomH about Unparished Areas, where you suggest that OS defines these boundaries, that they are verifiable (how?), that they are named by their District councils, and that one may/would refer to them as an “Unnamed parish”: > “oddly, these are all as carefully defined by the Ordnance Survey as are the civil parishes, and therefore are all perfectly verifiable. They are also all named by their respective District councils, and by those that live within them (I mean, can you imagine: “Where do you live?” “Oh, I live in Unnamed parish 9346”).”
The Localism Act 2011 only confers meta-powers to parishes in that a parish council does not automatically get these rights; it first has to demonstrate/resolve that it is “eligible” and must remain so. It can only use these powers to do things that “anyone” can do, so it can’t do many things that are reserved to higher powers. It seems these powers have proven very useful in certain specific cases, but I am not sure if there is a general power transfer going on. To start with, parishes only have one source of tax income (the precept) and are answerable to the electorate for that.
I haven’t removed any files or links as far as I know… If you can give a specific example I will look into it!
Comment from alexkemp on 21 August 2016 at 21:14
@The Maarssen Mapper: Ah, I see that you have restored access to all files in the BoundaryLines pages. That is good; it is a tremendous resource:
So, you are someone unable to admit openly/publicly to your mistakes. Very foolish. It means that you doom yourself to repeat them endlessly (you are still lecturing me on local government, yet I no longer trust any of your statements following your earlier gaffe & refusal to accept that you made one).
Comment from Colin Smale on 21 August 2016 at 22:01
Alex, you don’t do yourself any favours with these baseless comments. I have not removed any files, nor “restored any access” nor taken any other action which may explain your observation, which by the way you have still not explained. Your conspiracy theories are incredible. I would like to remind you that you brought this current animosity on yourself by stomping on the work of others without taking the trouble to investigate the background - and then you have the gall to accuse me of that same offence when I correct the situation. Whether you trust my statements or not is of course up to you, but I make them in good faith. Your accusations are bordering on slanderous. I don’t really care about your private beliefs. Please keep your ad-hom comments to yourself. If you can’t behave like a decent OSM citizen, you might consider whether it’s an appropriate environment for you.
Comment from alexkemp on 21 August 2016 at 23:03
@The Maarssen Mapper: What led to my recent comments came in 2 stages:
These two items (appear to) confirm each other: you declare that statements from a Parliamentary paper are incorrect but refuse to acknowledge the correction; you (appear to) silently reinstate access to files previously removed.
Now, I understand network problems totally. I am happy to accept that I made a mistake and that you never removed any files. Just in case the latter was down to me rather than you I restarted my whole system, and met this further attack when the system came back up. Will you get off my back, please?
Now…
This entire farrago started because I made some additions to your private wiki (the phrase “private wiki” is an oxymoron; by definition, all wikis are public). As soon as you complained to me about those additions I departed & would have never referred to it or you again. You, however, from that moment have made it your business to descend on all I do and nit-pick and make as much complaint & disturbance as you possibly can, seeking to get me ostracised and (if possible) rejected from OSM.
Look at the length of your comments in this Diary entry. You just cannot stop yourself, can you?
From my point of view, your behaviour towards me has amounted to abuse. I’ve tried everything that I know to defuse the situation, but you will not stop. Please, just go away & leave me be.
Comment from SomeoneElse on 22 August 2016 at 00:08
@alexkemp On https://www.openstreetmap.org/changeset/41428333 and elsewhere I asked if you could “please tone the comments down a bit”. This doesn’t seem to have happened.
I really don’t understand what you are trying to achieve here; are you trying to work together with other people towards a solution or are you just trying to have an argument? Comments such as “So, you are someone unable to admit openly/publicly to your mistakes. Very foolish” unfortunately suggest the latter.
If you really are trying to work towards a consensus then I’d suggest that an apology for some of the intemperate language used here and elsewhere (e.g. your comments on https://www.openstreetmap.org/user/alexkemp/diary/39062) is in order.
Comment from alexkemp on 22 August 2016 at 02:48
Hi @SomeoneElse
I’ve tried, I really have.
My language is well tempered; there is zero bad language. I am direct & to the point, yes. There are zero “baseless comments” - every comment is referenced & explained with reasons. My language is reasoned throughout; because of this I find it difficult to understand how you call me “intemperate”?
Did you read the storyline in 23:03? I have not sought this chap out; he has sought me out. He hits on me again & again solely, as best as I can see, to find opportunity to cause me harm.
How can I work to a consensus with this fellow? His desire is to cause me harm, which is why you end up being called into play. Have you noticed that I have not attempted a similar ploy?
Oh dear; this is not what I wanted before going to bed.
Comment from smsm1 on 22 August 2016 at 12:05
I presume town council is the same level as borough council. Example being Ipswich Borough Council.
Suburbs in Edinburgh have been getting areas, as the points were causing the geocoding to produce exceptionally weird results, which was causing people to stop using OSM based services such as CycleStreets.
Comment from alexkemp on 22 August 2016 at 14:33
Hi @smsm1
Yes, your Edinburgh experience certainly seems to mirror my own in Carlton & Thorneywood.
In general, in the UK a Borough Council is
admin_level=8(same as District) whilst a Town Council isadmin_level=10(same as Parish Council). They are normally different things. That changes with Ipswich, as it is both a District + an Unparished area (in my booksadmin_level=10).Suffolk Council has a list of Borough, District & Parish Councils within Suffolk (see also find-a-council at gov.uk). Ipswich Borough Council is upon that page & you can find contact details there. There is also a link to Google maps, which takes you to Grafton House, but not to an Area.
Ipswich does NOT appear as a CP within the Parish CP .gpx files. However, there are two Unparished Areas in Suffolk: Unnamed_shape_6717 and Unnamed_shape_7194. As best as I can tell:
Unnamed_shape_6717 == Lowestoft (there is a current proposal to create a Lowestoft Town Council). Lowestoft falls within Waveney District Council (admin_level=8) + Suffolk Coastal District Council, part of Suffolk county (admin_level=6). Corton (admin_level=10) exists to the north, but not Lowestoft to the south. The current OSM boundaries seem well-fitted to the OS GPX, as best I checked, apart from the missing Lowestoft.
Unnamed_shape_7194 == Ipswich (bingo!). Checking the boundaries to the north, there is Mid-Suffolk (admin_level=8), Ipswich (admin_level=8) + a tiny Whitton and an Akenham (both admin_level=10) but no Ipswich parish nor area (admin_level=10). Once again, the current OSM boundaries seem well-fitted to the OS GPX.
Note: in yet another bizarro episode, within all os_boundary .gpx file index pages the GPX file download links have had some forward-slash (‘/’) directory separators swapped for back-slashes (‘\’ == ‘%5C’ in a browser). The bottom line is a 404 on attempted download. I’ve rewritten the links above so that they work.
Comment from SomeoneElse on 22 August 2016 at 21:37
Re your comment above , by “intemperate” language I didn’t mean “bad” language; I just meant what you probably describe as being “direct & to the point”. As I’ve said elsewhere if you’re trying to persuade people to a particular point of view then you need to try and persuade them; calling them “fools” is unlikely to win any friends or influence people.
The truth is, geocoding and reverse geocoding are complicated. In $dayjob I’ve used a few commercially, and it’s normal to have to be very careful to structure the request going in or to pay careful attention to the data coming out. A statement such as “All the evidence suggests that no admin_level=10 BoundaryLine area means no accurate LSN in OSM” sounds very simple and straightforward, whereas in reality life’s a little more complicated than that.
The first problem is it assumes that there’s only one geocoder that works with OSM data. Nominatim does an excellent job (one customer that I was working for chose Mapquest’s instance of it over the one native to the OS they were using because it produced better results and was easier to use), but it doesn’t get everything right - as you say above it works better with areas rather than nodes. Other Geocoders do exist for OSM data (Photon’s the first one that springs to mind, but I’m sure there are others).
Secondly it’s not clear to me that adding “unparished” areas at admin level 10 actually helps in all cases. Just to take a couple of examples, here’s a location in Forest Town, Mansfield, and here’s Hyson Green Asda.
Nominatim gets the latter “correct” (after deletion of the admin_level=10 unparished area) and the former “wrong” (the admin_level=10 unparished area that you added still remains for Mansfield).
Thirdly however it suggests that we should only add data to OSM that can be understood by Nominatim (or change data already there so that it can). There’s an old saw that’s been around almost as long as OSM itself “don’t tag for the renderer”. For “renderer” read “router”, “geocoder” or any other data consumer. If the standard rendering style on osm.org doesn’t understand a particular form of tagging, or Nominatim doesn’t understand an address, it’s not an excuse to map things differently - there’s always the expectation that better renderers, routers and geocoders will come along, and it’s good to have the “correct” data there for when they do. You could argue of course that it’s “correct” to have an admin_level=10 for Mansfield at the same level as the admin_level=8; I’m somewhat skeptical personally, but would certainly listen to arguments behind something saying “here’s an area for which there’s no admin_level=10”.
In the case of the Nottingham and Mansfield suburbs, it’s always going to be complicated. As TomH said in the first comment above, “no two people will agree on exactly where one suburb ends and another one begins”, and as you said Nominatim really works better with areas rather than nodes. in the Forest Town example that’s a problem - Forest Town is clearly a suburb of Mansfield, but I’m not sure I could clearly draw a line between “Forest Town” and “Not Forest Town” (even though I’m probably the most active OSMer to the area).
Comment from alexkemp on 22 August 2016 at 22:41
Hello @SomeoneElse
Finally! You are talking to me about the content of the Diary Entry! Hooray….
I was very, very serious when I said upfront “I only half understand what I’m talking about”. In fact, that was possibly exaggerating the extent of my understanding. But, a person needs to start somewhere.
The value of the “admin_level=10” parishes is that no-one, including Nominatum, needs argue about where a Parish (Town, etc.) Council begins or ends; it is known exactly. I therefore consider my opening statement to be non-controversial (“LSN starts with admin_level=10 BoundaryLine areas”). Any controversy begins (1) if you accept that statement, and (2) when you discover that a substantial portion of the Country does not have an “admin_level=10” parish, because of the Unparished areas. The one problem with this admirably simple statement is that it demonstrably is not universally true (your discoveries + see my “Contra-Indication” in the main body above, discovered by me just as I had almost finished my wonderful treatise; I was spitting tin-tacks afterwards).
The paragraph above is the closest to the truth that I’ve seen anyone say, including those that speak for Nominatum. It’s a good enough working hypothesis until something more accurate is found.
PS
1) I obviously am not ‘tagging for the renderer’. You CAN accuse me of tagging for the LSN, but that is different.
2) I suggest adding admin_level=10 for Unparished areas because Nominatim already understands that tag + it affects LSN. Any other tag added would be for humans, not the renderer.
3) I would want to know the exact area of the country that is Unparished as opposed to CPs. It should be possible to find that out, and would help focus the mind as to the exact dimension of the problem.
Comment from alexkemp on 23 August 2016 at 14:19
Extent of country unparished: 9,229,902 hectares (39%)
Extent of country parished: 14,198,481 hectares (61%)
I’ll publish the entire stock of figures in a separate entry after d/checking (already found/fixed a small error).
Comment from SomeoneElse on 24 August 2016 at 12:50
No, it’s the same sort of thing as mapping golf courses as “beach” so that it it renders nicely on OSM’s standard map, or mapping farm tracks in Romania as “unclassified roads”, so that a particular router will use them (both of these these things have happened, BTW!).
An “admin_level=10” tag in England means that there’s a parish council. If there isn’t, while it may be useful to say “there is no parish council here” on a description of a relation describing the “hole”, it doesn’t make sense to have an “admin_level=10” tag.
A quick investigation of nominatim’s results will show that “LSN starts with admin_level=10 BoundaryLine areas” isn’t true either (assuming by “LSN” you mean “Nominatim”; and as I said above, that isn’t the only geocoder for OSM either).
It’d be great to fully understand why e.g. https://github.com/twain47/Nominatim/issues/231 happens, but “making up parish councils” doesn’t actually help fix the problem; in fact, as I mentioned in the Forest Town / Hyson Green example above, Forest Town is still geocoded incorrectly, just differently incorrectly.
Comment from alexkemp on 24 August 2016 at 13:34
Hi @SomeoneElse
God, I hate it when someone is correct in an opinion that differs from me. However, I do still adhere to the opening premise of this Diary entry, which I would paraphrase as follows:
“LSN” is used deliberately in the article to both avoid unfairly fixing the blame on Nominatim and also provide a smokescreen for my ignorance, as I’ve seen clear indications in the last 5 months that a great many different services are used to provide LSN facilities, and I have zero idea as to how they all interact together. I also have great confusion in that I cannot seem to drill down to the Nominatim code.
I’ve found sufficient evidence to indicate that the primary factor in LSN results is the presence of an admin_level=10 area (BoundaryLine). The fact of contra-indications means that other, as yet unknown, factors are also at play.
Wouldn’t it be wonderful to have some proper, documented help? (and here comes yet more controversy…). My experience of too many software writers is that they believe
codeto be forreal men, whilst documentation is for wimps.Comment from alexkemp on 24 August 2016 at 14:07
For information:
Comment from alexkemp on 24 August 2016 at 16:46
(I’ve just discovered some info)…
On 12:05 22 Aug @smsm1 said: > (in Edinburgh) the geocoding (is producing) exceptionally weird results, which was causing people to stop using OSM based services such as CycleStreets
If you have a look through the data you will see that the CITY_OF_EDINBURGH is a Unitary Authority and is entirely Unparished. It therefore has zero
admin_level=10areas, and is a prime candidate for all the problems set out in this post.