Re: [Talk-GB] UPRN Locations Map
On Thu, Jul 2, 2020, at 5:38 PM, Robert Whittaker (OSM lists) wrote: > I'm not completely sure if/how we can best make use of the new OS > OpenData (UPRNs, USRNs and related links) in OpenStreetMap, but as a > first step I've set up a quick slippy map with the UPRN locations > shown... > Thanks for doing this Robert but unless these can be used to pull down / link addresses or other attributes in some kind of usable, license friendly way then I don't see a lot of point in adding the references themselves to OSM. Postcodes are pretty well entrenched for navigation in the UK and I don't see the average person replacing their fairly easy to remember postcode with a hard to remember 11/12 digit number. Creation of the references also seems a bit inconsistent to me. On the seafront near me there are some open shelters and they have just one reference but on the next beach down indentical shelters have 4 references, I assume one for each face/row of seats. There are a lot of B's in my area and they all seem to have two references despite only have a single address, I assume one for the business and one for the property but businesses on the local pier do not have a separate node although they are separately addressable. Kevin ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Ficticious embankments? Vandalism.
On Tue, Mar 17, 2020, at 2:08 AM, Warin wrote: > On 17/3/20 8:02 am, ael wrote: >> In cases like this I would use the source tag on the way so that others have >> a very good chance of seeing it and respecting the previous work rather than >> simply changing it to what they think it should be. It is too easy to over >> look hard work that may have gone into establishing data. A single GPS trace >> is fine if that is all there is, better to average many GPS traces, in some >> locations I have 50+. What would you put in the source tag in this case and does it make a difference? The inability to mark an object's location as "authorititive" has always seemed like a massive shortcoming of the project to me. Stopping people re-aligning things based on a bad phone GPS or badly aligned aerial imagery is impossible and even realising that things have been incorrectly moved is random at best. Kevin ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Saltash/Plymouth
I might have some traces for the straight through Devon to Cornwall route so probably the same as what you already have. It would be better if someone more local took a look. It's a shame that the state of aerial photography for the SW that is available to mappers is now so poor. Kevin On Sun, Feb 16, 2020, at 9:48 PM, ael wrote: > Are there any local mappers in Plymouth or Saltash? ... > ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] The Range - looking for tagging consensus
Added a new Range store a couple of weeks ago as shop=houseware although shop=homeware would be fine as well. I think department store has some Grace Brothers style connotations regarding layout that The Range doesn't have at all. Kevin On Fri, Nov 8, 2019, at 1:00 PM, Silent Spike wrote: > ... > I'm actually leaning towards department store, but feel like OSM needs a new > tag to specify which departments a department store has so that we can > distinguish the huge ones from smaller ones like The Range or Laura Ashley > (another brand I've struggled with categorising). ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Automated Code-Point Open postcode editing (simple cases only)
On Fri, Jul 19, 2019, at 3:21 PM, Andrzej wrote: > > Thank you for your opinion, Robert. I will suspend adding postcodes from > Code-point Open. > > Do others agree with it or would you rather have more postcodes in database > first and work on accuracy and completeness afterwards? > > Indeed, Code-Point Open is less than ideal, the issues are almost always > caused by lack of differentiation between residential and "large user" > postcodes. On the other hand, it is the only legal source of postcodes we > have, other than local knowledge, but the latter is realistically limited to > a dozen or so postcodes per mapper. Businesses website could also be OK but > they are usually copyrighted. Derived databases, like FHRS, are generally not > OK, a unless also permitted by Royal Mail. > > It's not that I don't care about complete addresses either. But my spare time > is limited, and I feel I can contribute more by adding missing postcodes in a > town vs adding complete addresses in a few streets. Others may have different > priorities. > > I disagree that having data from Code-Point Open outside OSM is sufficient. > Excluding surveyed information, everything in OSM is already publicly > available (or should be). Yet, we all keep using and working on OSM. Besides, > how to extend or combine information without adding it first? > Hi, I would love to see a comparison done between the accuracy of manually added postcodes vs. those added from the OS or ONS datasets. Someone manually added a bunch of postcodes near me and I am pretty certain quite a few of them are wrong but without going around knocking on people's doors they are probably going to stay wrong forever. I usually use the ONS Postcode dataset (https://geoportal.statistics.gov.uk/datasets/ons-postcode-directory-may-2019) rather than CodePoint Open. The datasets do have differences and which is more accurate I have no idea. The ONS centroids are mostly "snapped" to the nearest building within the postcode so are pretty easy to match up. I know that Robert is sincere in his views but the classic "don't add data to OSM because it will spoil someone else's enjoyment" always makes me chuckle. In most parts of the country the idea that the current cohort of mappers can add accurate address data by hand is pie in the sky. There are certainly issues with adding these postcodes to buildings in dense town centres but in those areas you can often find postcodes by other means anyway. I think adding postcodes to residential or rural areas from these datasets is fine but I personally wouldn't add them unless I had some on-the-ground knowledge of the area. Kevin ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Automated Code-Point Open postcode editing (simple cases only)
On Wed, Jul 17, 2019, at 12:56 PM, Mateusz Konieczny wrote: > In that case maybe it would be a good idea to merge existing address-only > nodes > with building outlines as the first step? Some buildings (usually post office locations but also some others such as council offices) can have more than one postcode so I usually put those additional postcodes on separate nodes. Kevin___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] UK coastline data
On Sat, Jul 13, 2019, at 12:47 PM, Borbus wrote: > ...That data is included with the OS tidal waters data. It's not much more > effort to use it and it's very useful data for many people. People use > maps for many different things. Rendering is not a problem. Carto > handles it just fine already. But it does expect the intertidal zone to > be between a "coastline" and the edge of a tidalflat. Unfortunately, this is one of those fairly pointless discussions that characterises OSM. I know it isn't always possible but I prefer the meaning of words in OSM tags to have the same meaning as in everyday conversation. If you want to add a way on every river in the country describing the mean low water level then fine but that doesn't mean you need to call it coastline. If tidal beaches, mudflats, marsh areas, etc. are tagged as tidal=yes (which they should be) then all you are doing by adding a way for MLW is describing the part of the river that on average is non-tidal which doesn't add any extra information that isn't already there. I notice that several people have messed around with the tagging on the Dart over the years so it probably isn't perfect anyway but changing it to coastline certainly isn't the solution. Kevin ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] UK coastline data
On Fri, Jul 12, 2019, at 8:05 PM, Borbus wrote:. > Does that matter, though? The way many things in OSM are tagged is quite > arbitrary. What if "coastline" just means "mean high water level"? A tag > for MHWL seems much more useful than "you would probably consider this > the coast rather than a river bank". > > > I don't really care either way but what would be the benefit of changing > > it to coastline (and slavishly copying the OS is not a benefit) ? > > The benefit is we don't have to arbitrarily draw the line somewhere. The > tidal limit is well-defined so it's easy to be consistent. Just because the coastline follows MLW as it goes around the coast doesn't mean it needs to follow every tidal waterway inland. That doesn't follow at all. To achieve what you want you would need to add yet another way inside of the riverbank and intertidal areas which seems like a fair bit of effort to do for every river for no real benefit to map users whatsoever. Then you need to get it all to render right where you have tidal mud banks, etc. in the centre of the river. Kevin ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] UK coastline data
On Thu, Jul 11, 2019, at 10:41 PM, Borbus wrote: > > The Dart cuts the coastline off right at the mouth, which doesn't seem > right... > I think the main reason I did that back in the day is that mapping coastline all the way up to Totnes seems extremely non-intuitive. Someone standing on Totnes quay (10 miles inland) is not standing on the coast in any meaningful way. I don't really care either way but what would be the benefit of changing it to coastline (and slavishly copying the OS is not a benefit) ? Kevin___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] New Bridge Gunnislake
Saw this on the local news last night. A 30+ mile diversion is going to cause a few problems over Easter if they can't get it sorted. I am not that close but have mapped this area a bit in the past. No real view either way on adding the restriction to OSM. Good for users who keep their data up-to-date but potentially not so good for people who update infrequently. Kevin On Wed, Apr 3, 2019, at 5:51 PM, ael via Talk-GB wrote: > With some hesitation, I have just added access = no to New Bridge > in Gunnislake. > https://www.openstreetmap.org/way/608610775 > > This is one of only two major routes across the lower Tamar, so it is > of major importance for routing applications. > > The bridge has been closed after an accident damaged the structure, and > it has been suggested that may be shut until 1st May. > > I know the area fairly well, and I believe that I had edited the bridge > several times in the past, although the history now only seems to go back > for around a year with edits from non local (armchair?) mappers. > > The relevance of that comment is that I am not aware of any very local > mappers. > > Anyway, have I overlooked a way to tag a temporary restriction like > this? I have put a note in my dairy to remind me to check the tagging in > a week or two, but I could still forget ... which would be unfortunate > :-) > > ael > > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk-gb > ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Measuring building height
This stuff is hilarious. Standing outside people's homes taking photos of their cars and house isn't going to end very well. Is it maybe possible to get building heights from lidar data? Kevin On Tue, Mar 19, 2019, at 11:17 PM, Warin wrote: > On 20/03/19 07:51, Neil Matthews wrote: > > > > So, I just tried this and I think it has a reasonable chance of giving > > a reasonable result. > > > > Take a photo of a car outside the building. Measure number of pixels > > for the car and number of pixels for the building and the height can > > be approximated by: > > > > building_pixels / car_pixels * car_height_in_m > > > > I reckon an average of 1.5m might be reasonable for the car height -- > > otherwise use something more detailed: > > https://www.automobiledimension.com/ford-car-dimensions.html > > > > > > Obviously, the further the car is from the building the less accurate > > the measurement will be. > > > > The further the camera/photo is from the building the better too. Less > camera distortion. > > > > > > Cheers, > > > > Neil > > > > > > On 19/03/2019 16:23, Tony Shield wrote: > >> > >> Hi > >> > >> Been figuring out how to do this for a while - my solution- > >> > >> rule - I used 30cm (aka 1 foot), calculator, known length of arm - in > >> my case .6m, OSM map to measure distance from target. > >> > >> With hand holding rule vertically measure the target height against > >> the rule for rule height, this is the key measurement, note the > >> measurement point. From the map measure the distance from the > >> measuring point to the target > >> > >> With this information and using proportions (which is what a tangent > >> is) - > >> > >> target height = (rule height in metres * distance from measuring > >> point to target) / length of arm in metres. > >> > >> Using this technique I have this morning measured known height of of > >> a local landmark, and the unknown height of a building. The known > >> height of 50m measured 8cm at a range of 375m. The unknown height of > >> the building with 5 floors was calculated to be 20.7 metres which > >> would on the face of it be realistic (from 3cm and 414m). (Botany Bay > >> mill in Chorley). > >> > >> TonyS999 > >> > >> > > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk-gb > ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Bridleway or track?
Yes and yes. Just tag it as a footway then with the appropriate designation. Kevin On Thu, Mar 14, 2019, at 4:14 PM, Martin Wynne wrote: > On 14/03/2019 15:29, Devonshire wrote: > > If there are parallel tracks visible (even if maybe historic) I would tag > > it as a track, if a single track is visible then tag it as a footway. > > Thanks Kevin. Do you mean wheel ruts, footprints? > > In the case of my photo, there was no evidence of any use at all, by > foot or wheel: > > http://85a.uk/track_query_960x648.jpg > > Perhaps it is not a highway at all, despite being waymarked as a public > footpath? > > cheers, > > Martin. > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk-gb > ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Bridleway or track?
If there are parallel tracks visible (even if maybe historic) I would tag it as a track, if a single track is visible then tag it as a footway. Kevin On Thu, Mar 14, 2019, at 3:19 PM, Martin Wynne wrote: > > The path tag is actively being discouraged. > > Which makes it more necessary to have some direction on what constitutes > a "track". Is sufficient width for vehicles enough by itself to convert > a footpath to a track, or does there need to be some evidence of actual > vehicular use? > > It's a common question in these parts. Here is another track? or path? > -- which is it? On the right is an old track which is legally a public > footpath. It's wide enough for vehicles, but all vehicular traffic has > long since transferred itself to crossing the field on the left instead: > > http://85a.uk/track_query_960x648.jpg > > (not the same track as my previous photos) > > cheers, > > Martin. > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk-gb > ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Bridleway or track?
Both footway and path infer a way of a certain width suitable for people to use, neither infers any legal right of use as far as I am concerned. If starting over path is a better word as people outside of OSM have clue what it is but otherwise I see them as interchangeable. Kevin On Tue, Mar 12, 2019, at 7:37 AM, Adam Snape wrote: > > > On Mon, 11 Mar 2019, 12:54 Devonshire, wrote: >> __ >> I have personally deprecated highway=bridleway|byway etc. as the combination >> of highway=footway|track|service and designation=public_footpath etc. >> contains more useful information both for map rendering and for active map >> users. Whether you wan't to do the same is up to you. >> >> Kevin > > Byway is universally depreciated these days. > > It seems somewhat odd to reject bridleway whilst using footway which shares > the same arguable 'flaw' of tagging both physical appearance and implied > access in one tag. For those unhappy with these tags, as Dave mentions, the > highway=path tag was designed to physically describe a physical path and be > used in combination with access tags. > > Now, I can understand using either the 'classic' (highway=footway or > highway=bridleway) or 'alternate' (highway=path + access tags) tagging > schemes but I'd think that a hybrid resulting in combinations like > highway=footway horse=designated is best avoided . > > Kind regards > > Adam > > >> ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
Re: [Talk-GB] Bridleway or track?
I have personally deprecated highway=bridleway|byway etc. as the combination of highway=footway|track|service and designation=public_footpath etc. contains more useful information both for map rendering and for active map users. Whether you wan't to do the same is up to you. Kevin On Mon, Mar 11, 2019, at 11:14 AM, Martin Wynne wrote: > On 11/03/2019 08:49, p...@trigpoint.me.uk wrote: > > > Access tags for a bridleway in the UK or in my experience England and Wales > > should be horse=designated, foot=designated and bicycle=designated. As Andy > > mentioned the important tag is designation=public_bridleway. > > Thanks Phil. I'm doing all that, but it's not the question. > > The question is: > > highway=bridleway or highway=track ? > > I'm not much wiser on how to decide between them. > > More generally, does highway= indicate the physical appearance as > surveyed -- or the intended use, as designated? > > If it's the latter, as for example highway=primary, how is the actual > appearance to be mapped and tagged? > > cheers, > > Martin. > > ___ > Talk-GB mailing list > Talk-GB@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk-gb > ___ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
