Recent diary entries
Yesterday the UK government published a consultation on options for a new Thames Crossing close to where I live. The web page is here: https://www.gov.uk/government/consultations/options-for-a-new-lower-thames-crossing
Of course, I made straight for the map of the possible routes - all very broad-brush at the moment. What surprised me most was that, despite owning its own mapping agency the Ordnance Survey, the government has used an OSM Mapnik rendering for its base map! Maybe this is a regular occurrence, but it's a first for me.
New alignment (and name - Peninsula Way) of A228 near Hoo missing in Google Maps, but bang up to date in OSM: see http://tools.geofabrik.de/mc/?mt0=mapnik&mt1=googlemap&lat=51.4263&lon=0.54859&zoom=15
Bing maps have the new road, but call it Ratcliffe Highway (which is actually the name of the remaining portions of the old road, now a minor road): http://www.bing.com/maps/?v=2&cp=51.43160359990295~0.5541890554995277&lvl=13.999999999999998&dir=0&sty=c&form=LMLTCC
Today's ITO analysis shows that Dartford has now reached 100% (matches against OS Locator, or rather discrepancies resolved), which boosts it up to 16th in the table.
A month ago, I emailed Kent County Council's Highways department as follows:
I'm confused about the name(s) of the A225 road passing through Sutton at Hone. The Kent County Council Highways Gazetteer (March 2011) lists MAIN ROAD with a total length of 57.36 metres and DARTFORD ROAD with a length of 2156.95 metres. Yet there seem to be postal addresses ranging from 1 to 300 Main Road, Sutton at Hone, suggesting that the bulk of the road running through that parish is called Main Road, not Dartford Road. I wonder if you could check the accuracy of the listings in the gazetteer. The reason I am asking is that I am updating the map of the area for OpenStreetMap.org.
Today I received this reply from Dartford Borough Council:
It would appear that there is an error within the NSG as you claim. The numbering along the road would suggest that the change takes place on a 500m stretch of roadway either side of the railway bridge, towards the southern boundary of the Borough.
A site visit revealed no signage or road markings to provide a practicable start/end point that are recognisable on the ground, providing a simple route to correcting to the error. Several available datasets have differing opinions on the location of the change, from the railway bridge itself to 30m south to 300m south of that location.
I have been unable to get definitive documentation, for example a road adoption plan or section 38 agreement, to identify where the end of Dartford Road and the start of Main Road is at this time. This could be due to the road pre-dating a lot of this type of documentation and having been within the realm of several border changes.
I am seeking advice from Geoplace as well as the NSG custodians and others on the implications of making a change to the NSG without proper documented proof.
For your current need, I would suggest the station road junction 30m south of the railway bridge at OSGB36 555806, 169287 or WGS84 0°14'24.431"E 51°24'2.49"N would be a logical start end point.
I will update you when there have been further development.
Anyone else think this changeset (http://www.openstreetmap.org/browse/changeset/8170139) looks dodgy? Has anyone reading this contacted the user about it?
I've started using the website http://gebweb.net/optimap/ to plan my mapping forays. I click on all the streets I want to visit (usually based on the ITO Analysis site) together with my start point and end point and it works out a route for me. I then enter the points as Favourites in my TomTom iPhone app, numbered according to the sequence the optimap site calculated, then navigate from one to another. Sadly, it's based on Google Maps! If anyone knows a better one, let me know. It's a shame the TomTom app doesn't have a file upload facility though: it's a bit tedious entering all the locations manually. I daresay there are other routing apps that do have that capability though.
Tonight's target area is Greenhithe and Stone.
Having done all I can on Medway in terms of street-names for the moment, I've turned my attention to the area around where I'm currently staying. Knocked off a few street-names in Swanley, St Mary Cray, and St Paul's Cray (Sevenoaks and Bromley as far as the ITO analysis is concerned).
As of this morning, Medway is 39th in the league table of completeness (street-names versus OS Locator), with 99.57% of the 2,809 Locator entries now matched (or otherwise reconciled) against names in OSM. A month ago it was 121st and 88.71%. I think the 12 discrepancies still listed are all either in Chatham Dockyard or in Gillingham docks, so we may just end up copying those from Locator/Streetview.
Spent a frustrating couple of hours today trying to gather missing street-names and confirming what I already suspected - that most of them were in inaccessible areas, namely the military base in Gillingham and Chatham Historic Dockyard. Spent two hours driving around and came back with three street-names. Anyway,I decided in the end to copy the ones on the military base from OS OpenData - StreetView in the first instance, Locator if necessary. Chatham Dockyard - I may pay the £15 to go in during normal visiting hours or I may try to get them to let me in for free by explaining what I'm trying to do.
Still thinking about the Contributor Terms. Have declined them for the time being while I make up my mind. I imagine I will accept them in the end.
Found the new Kingdom Hall of Jehovah's Witnesses in Gravesend today after one of them called and left a leaflet with the address on. I knew the old one had been demolished and flats built on the site, and I knew the new one was somewhere near Lord Street car park, but didn't know exactly where. The address is 2 South Street, Gravesend, DA12 1DA.
I've added "alt_name = Eastgate" to the stretch of Rochester High Street between Star Hill and the old city wall. I'm not sure what the status of this name is. It appears on all OS maps since the 19th century. It's still there on OS Locator (which is what drew my attention to it). However, the numbering of the street simply continues the numbering of the High Street within the city walls. The Royal Mail's address file doesn't recognize it - properties in Eastgate have High Street addresses. Yet I'm pretty sure there used to be a street-name sign at the Star Hill end when I was a lad (we used to wait for buses outside the Seeboard show-room). The Medway Map Service (http://medgis.medway.gov.uk/) doesn't recognize Eastgate as a street-name either - which presumably means it's not in the official List of Streets. (If you click on the street in Identify Feature mode, it names it as High Street.) Maybe it should be old_name rather than alt_name, but somehow the fact that the OS still uses it persuaded me to go for alt_name. Google throws up some references to numbered addresses in Eastgate - e.g. "W. Naylar and Son, 191 eastgate, Rochester, funeral directors". These are generally in historical documents, though.
I was a bit surprised, knowing that Gravesham had reached 100% and 13th place in the ITOWorld table, to see it move up another place to 12th yesterday. As places on 100% appear to be ranked by the total number of roads, the only way Gravesham could have moved up further is if some other place dropped out of the 100% rankings. Sure enough, Wrexham seems to have done just that. It actually looks like a way was deleted at the weekend, leading to the appearance of a single mismatch against the OS Locator dataset.
Meanwhile, Medway has moved from 121st to 95th in the last couple of days, as there seem to be several of us working on it. Now 91.67% complete (by this measure).
I've been working in a novel way today - novel for me at least. I'd been looking at the Keepright web site (keepright.ipax.at) quite a bit lately, and manually fixing the odd error here and there.
But today I noticed that you can extract the errors in the area you're looking at to a GPX file. So I did that for my local area, although I unticked some of the error-types first as they were things I wasn't particularly interested in (like 'floating islands').
Then I had a look at an app I've had on the iPhone for a while called MotionX-GPS. I knew you could create waypoints in the app, so I wondered if there was a way to import waypoints from a GPX file. Sure enough, there is. You have to use a bit of a roundabout method: on the PC, you email the GPX file to GPSimport@motionx.com, and it then sends an email back to you. The idea is that you open this email using the Mail app on the iPhone: the email contains a link which, when you click on it, takes you into the MotionX-GPS app and into a dialogue that allows you to import the waypoints from the GPX file. According to the help in the app, you can import a maximum of 12 waypoints at a time, but actually I just sent one with 15 waypoints and it imported them successfully.
So then I sorted my waypoint list by proximity, and selected the closest one - this shows the location of the keepright error on a map and a description of the error, so I then drove to the location (with the help of my TomTom - sorry!) and checked out the missing details (mostly they were places of worship and schools without names): I edited the waypoint in MotionX-GPS to add the missing details and then checked out the next nearest waypoint.
Once all 12 were done, I returned to base and made the alterations in Potlatch2. This bit was entirely manual: I don't know of a way to get the information from MotionX-GPS into Potlatch or OSM automatically - would be great if you could. Each time I resolve an error in Potlatch I also edit the error in Keepright (I use the 'ignore temporarily' option to show that I think I've fixed it).
So, 24 POIs fixed today, and another 15 sitting on the iPhone waiting for another mapping foray.
I won't know for a couple of days, but I *think* I *may* have done enough today to bring Gravesham to the magical 100% mark. Fingers crossed!
ITO World have a useful analysis tool comparing street-names recorded in OSM against those recorded in the OS Locator data set. http://www.itoworld.com/product/data/osm_analysis/main , but you may have to be registered and logged in to see it. They've done the analysis by local government area, giving a handy 'league table' of the completeness of each area (in terms of street-naming).
10 days ago, Gravesham was 142nd in the table, with 181 of its 1,133 street-names apparently missing. It was therefore rated 83.32% complete. As of today, it has risen to 76th position, with only 47 streets missing, and is now 95.23% complete.
It's been an interesting exercise resolving the discrepancies so far. Usually, the street-name is indeed missing from OSM, and in some cases this is because there is no street-sign on the ground (especially in rural areas - I've yet to find signs for Buckland Road, Wrangling Lane, Lockyers Hill, etc.). Discrepancies regarding apostrophes are extremely common (the ITO analysis separates these out as 'minor' errors), and for these I usually flag the other version as an alternative. Quite often, two street-signs for the same street will have different spellings, again often involving apostrophes, but also disagreeing about whether words are separate or joined together (e.g. Deneholm Road vs. Dene Holm Road).
Occasionally, the OS has clearly got it wrong, as in the three avenues in Northfleet which have been named '------ AVEUNE' in the Locator database, or the DRIVE which has become a DRLVE.
I think I got all the missing roads I was looking for in Walderslade on Sunday, plus about two-thirds of the missing street-names. Need to go back for the rest.
Did a spell of about an hour around Burgess Hill last night (before rehearsal), and have just finished putting it all into OSM.
Looking to map some more areas of Walderslade/Tunbury Wood, and tackle an area of Lordswood with lots of missing street-names.