Hi all, I’ve lately been looking into Walk Wheel Cycle Trust’s (WWCTs) publically available data and amending a few NCN relations in Scotland that have gaps, sorting the members into a proper order etc. etc.
Now, I’m a stickler for detail so I noted that many of the NCNs in Scotland are–within OSM currently–fully contiguous, but by referring to WWCTs ArcGIS map as well as their OS based map, using NCN 76 as an example, NCN 76 terminates at the west side of Edinburgh. From the ArcGIS map, one would then likely follow NCN 1 into Edinburgh along the way that has a “Link No: 76” tagged, but this Link tag abruptly vanishes part way, before randomly appearing again and vanishing once again once it meets up with NCN 75.
From the east side, NCN 76 stops and NCN 1 abruptly starts around the edge of Edinburgh, ArcGIS and OS. I don’t see any “Link No” tags going into Edinburgh from here.
Basically I’m looking for a community opinion on what to do with the NCN continuity, because I would be inclined to break up the continuities to match what WWCTs data appears to show (ignoring the “Link No” thing entirely as it seems to be inconsistent). The only things stopping me from doing so right now:
Would this majorly break any services that rely on a continous route relation to work?
Is anyone aware of any conflicting ground truth signage on this point? As this should probably take precedence over any data held by WWCT.
I think would be one of the most large-scale, obvious-on-the-map changes I’ve made so far, so I just want to be sure this sounds sane to other people before I tear the map up
Just to add one data point from England, I created a gap where it was really clear on the ground where the “numbered” route clearly stopped and restarted, but cyclists were still signed along the “unnumbered” bit.
I’m open to suggestions about how best to represent the gap…
I’m not entirely sure what you’re proposing here - is this principally about route overlaps (which the roadgeeks call “multiplexes”), where a given way might currently belong to both NCN 76 and NCN 1 relations, or is it about routes that genuinely do have on-the-ground gaps?
Referencing NCN 76 again, every primary source I can find basically implies that NCN 76 does not exist in Edinburgh, and if you wanted to traverse the entire NCN 76 then you would need to travel along part of NCN 1 which bridges the (from primary sources) gap between the west and east sides of Edinburgh.
The current state of the NCN 76 relation includes that chunk of NCN 1 that bridges the gap in NCN 76. I believe that this is technically incorrect i.e. you would not tell someone “Follow the NCN 76 from North Quensferry through Edinburgh to Prestonpans”. Rather, you would say “Follow the NCN 76 to the west edge of Edinburgh, then use NCN 1 until you get to Musselburgh before following NCN 76 again to Prestonpans”.
I would be proposing cutting out the section in green from NCN 76. I believe there’s at least a few other NCN routes that do the same, as it seems to me (from WWCTs own data) that NCN routes should never overlap, and in this case, NCN 76 and NCN 1 overlap.
Edit: Forgot to add, if there is local signage to suggest that NCN 76 does actually overlap with NCN 1 to at least some extent, then this should overrule WWCTs data. I have seen a single Google Streetview photo that shows this to be a possibility.
Right, so you’re talking about overlapping routes.
WWCT/Sustrans policy has been that routes don’t overlap for several years now (since the introduction of three-digit routes in the 2010s, I think). Nonetheless some authorities consistently like to sign routes as overlapping - Worcestershire do, for example - and in this case, OSM should follow what’s on the ground so that navigation apps and other consumers can present an accurate picture. Where it’s a bit more erratic… use your own judgment but defer to ground survey?
@ExoMal - I covered the section from Roseburn to Whitehouse road and I can confirm that. there is no indication of the NCN76 along here. So it will be correct to remove this from the NCN76 and match up with the routes as signed.
Thanks for that, I’ll amend NCN 76 as I’d planned. I’ll leave the other NCN routes alone for now until local surveys are carried out.
Edit: Whoops, jumped the gun a little, didn’t notice that it’s not the entire section in question that has been surveyed. I’ll hold off on amending for now, especially now I’ve noticed another apparent gap in NCN 76 through Inverkeithing, although I’m fairly confident I’ve seen local signage for NCN 76 there. Will probably survey it myself soon.
Two small points in favour of keeping the centre of Edinburgh bit - it’s called “Round the Forth” and if you’re a cyclist who has downloaded the gpx file for routing purposes then you’d be left with a big gap in the middle which would be a bit annoying…
Also the section through Inverkething is shared, but I can see a sign under the viaduct at Ferrytoll with both 1 and 76 on it.
and if you’re a cyclist who has downloaded the gpx file for routing purposes then you’d be left with a big gap in the middle which would be a bit annoying…