Has this section actually been removed from the NCN? If it’s temporarily closed due to a landslip then it would probably be more accurate to keep it as part of the NCN relation and instead map the landslip as a node with access=no hazard=landslide and check_date=(something appropriate) on the cycleway, to cause routers to temporarily route around the landslide.

My mistake, thanks for catching that.

OK fair enough, that’s a good reason for having two separate relations. Thanks for taking the time to survey it.

I’m still not convinced that the relations as they are mapped right now are correct though.

As the Bay Cycleway is the continuous route, then the relation representing it (relation/21131746) should be continuous — but currently it just contains the start and end points, and that one short section along the A5087. I think it should also contain all the ways which are in the NCN700 relation (relation/7024538).

Hiya, I don’t think the Bob Graham round can or should be mapped on OSM. It’s not a fixed route, it’s a series of summits to visit (actually in any order) and the person doing it can choose any route between them — so it cannot be represented as a relation of ways in OSM.

From a verifiability point of view, there are no Bob Graham signs or waymarkers on the route, so nobody can verify it on the ground.

While you might argue that there is a route which people commonly take for doing the summits in a clockwise order, I don’t think that should be on OSM either. The route takes a series of fell running trods which are now seriously suffering from wear and path erosion (to the extent that the BG committee have put limits on the number of support runners for a BG attempt). Anything which steers people towards all using exactly the same route for a BG is going to amplify this path erosion.

I realise you probably mean well, but I think these edits should be reverted.

Ah, good to know that’s standardised, thanks for the references!

Can you point me (and others) towards the sources which define the Bay Cycleway and NCN700 as separate things, and which show how they differ?

Is there a definitive list of the categories of things which UPRNs can be assigned to somewhere? Something which could be used to rule out the possibility of this UPRN being assigned to any kind of thing other than the street.

Are you sure the two routes are different? All the references I can find to it from either Walk Wheel Cycle Trust (WWCT) or the Morecambe Bay Partnership (MBP) refer to it equally as NCN700, the Bay Cycle Way, and a national cycle route.

The only official map I can find of it comes from WWCT and shows it as NCN700 with a dashed section on the A5087. The MBP point people towards WWCT for it.

See e.g. https://morecambebay.org.uk/discover-morecambe-bay/visit-the-bay/cycle-morecambe-bay/bay-cycle-way/
https://lancasterandmorecambebay.com/places/bay-cycle-way
https://morecambebay.org.uk/discover-morecambe-bay/visit-the-bay/cycle-morecambe-bay/
https://www.walkwheelcycletrust.org.uk/find-other-routes/bay-cycle-way/

There is some mixed messaging over the official direction of the route too. Most of the references say it goes from Walney to Glasson Dock, but the MBP suggested day splits have it starting at Glasson Dock.

So overall I’m not convinced that it is two separate routes or needs separate relations. Perhaps it was that way originally when MBP created it, but my impression is that they’ve effectively handed it over to WWCT, who have made subsequent dedesignation changes.

I removed the UPRN from the street here in changeset/186454199, as I don’t think it’s clear what that UPRN refers to when looking at the UPRN and USRN databases. If it refers to the street, why is it at one end rather than in the middle, and offset from the road centreline? As a UPRN why does it refer to a street at all? Could it be referring to a lamp post or street cabinet?

I also removed the UPRN added to a node on the street in changeset/186365281 as from the UPRN and USRN databases it’s not clear what that UPRN is meant to refer to. It might refer to the street (if so why is it at one end on a corner, rather than in the middle?), or perhaps to a drain cover, lamp post or street cabinet. Who knows? Until we know for sure it shouldn’t be in OSM.

My point was that the relation (relation/7024538) doesn’t make sense at the moment — it (correctly) has ways from Glasson Dock to Walney, minus the discontinuous bit on the A5087 — but the distance is now tagged as 7km rather than closer to 120km, and the to= tag says Rampside rather than Walney. That doesn’t seem correct to me and I’m struggling to understand the intent/reasoning behind your changes.

relation/21131746 now exists too, and I don’t understand why that one is necessary. As far as I understand it, Sustrans has renamed to Walk Wheel Cycle Trust, and they have removed some bits of road from the NCN, but the Bay Cycle Way is still NCN700 so there should just be one relation not two?

Thanks, yes the warning about “should be an area” is spurious, since the natural=fell tagging is fairly unique to the UK and ID doesn’t really understand it.

If you’re reverting something, please use a proper revert tool rather than undoing your changes manually :) If you use a revert tool, the history of un-deleted nodes/areas/relations is restored correctly. I’ve re-done the revert as changeset/186420694.

JOSM has a revert tool, but there’s also https://revert.osmz.ru/ and https://revert.monicz.dev/

Thanks.

Yes, it does seem a bit odd that there’s a UPRN for the street, but having looked at the data here I do agree with you that it seems to relate to the street.

See discussion of this on changeset/186204250.

Heya, thanks for your attention to the updates to the Bay Cycleway/NCN700.

I’m not sure the removal of the A5087 section near Rampside from the route means that its distance changes to 7km or that its end point becomes Rampside.

Instead, I would interpret that to mean that the route relation should become discontinuous.

Obviously it would be better if funding was put in to bringing the A5087 up to safe standards for cyclists, but OSM has to represent the world we’re in, and to me it seems more helpful to people to present a discontinuous route from Glasson Dock to Walney, with that section removed, than to say the route just ends at Roosebeck or Rampside.

OK, that makes sense, but is not at all obvious from the changeset description or tagging.

Is there a way you could make the tagging clearer for these UPRNs on streets? Given that UPRN stands for ‘*property* reference number’ it’s a bit weird to see one on a street; especially since streets have USRNs.

Similarly, if you could add these UPRNs to streets in a separate changeset from all the rest of the UPRNs and geometry changes, with a changeset description which describes (or links to a wiki page describing) your methodology, that would go a long way to making the derivation of them transparent and verifiable, since from what you say it doesn’t sound straightforward.

Hiya, are you sure you can define the boundaries of a fell that precisely? As I understand it, they typically have fuzzy boundaries, not tied to a particular land parcel or ownership, so have to be drawn as a node in OSM rather than a precise area.

Why is there an otherwise untagged UPRN node in the middle of the street? node/14049711630

I don’t think it’s useful to add UPRNs to the map if we don’t know explicitly what they refer to in the UPRN DB. Doing so means we lose the list of known-UPRNs-which-are-not-yet-in-OSM. If people have a need to know the lat/lon of the full set of UPRNs they can always overlay the URPN DB on OSM for full coverage.

This reverts changesets: 185263362, 185222840, 185210380, 185207665, 185193900, 185163946, 185163845, 185163411, 185163629, 185162888, 185152973, 185152629, 185152477, 185152204, 185152135, 185151736, 185151527, 185151340, 185150976, 185147920, 185147213

Heya, thanks for all the edits you’ve been making over the recent months in Lancaster and Morecambe :)

When you’re significantly reworking geometry, please make sure to align the aerial imagery to a source of ground truth before beginning. The aerial imagery changes over time, and can be several metres offset across time and between different places in town, compared to (for example) an accurate GPS reading.

The easiest way to do this is to enable the ‘OSMUK Cadastral Parcels’ layer (in Overlays in ‘Background Settings’ on the right in the editor), and then align the imagery to that before editing. The Cadastral Parcels come from the UK Land Registry and are guaranteed to be aligned to ground truth.

For this area of Morecambe, I think the current offsets of Bing imagery are -0.68, -1.99 metres, which is why the bits I’ve tweaked since you‘ve edited them look offset when you first open the editor.

I’ll go ahead and realign these houses and the ones on George Street now, but in future if you could align the imagery before editing that would be fantastic :)

I hope that makes sense, and sorry for the long message. Unfortunately it’s not a simple topic! Let me know if any of it doesn’t make sense. Ta