Yes, that's how sentences work, you can say one thing and then qualify it with another. "It is sunny today but there will be rain around 2pm."

Facetiousness aside, a typical case is https://cycle.travel/map?from=&to=&fromLL=51.500175,-0.241000&toLL=51.498920,-0.242396 where the sane route is pushing along a short stretch of highway=footway to avoid a 400m detour. c.t uses a pedestrian symbol in this case to indicate that the user should push (and in the app will give distinct turn-by-turn directions etc.).

Contrast with https://cycle.travel/map?from=&to=&fromLL=51.748672,-1.256575&toLL=51.751087,-1.248749, where you can't even drag the route onto the path through Christ Church Meadows, because pushing a bike is forbidden there.

This is why it is important to differentiate between "you can push a bicycle here" (bicycle=dismount) and "you cannot push a bicycle here" (bicycle=no).

Yep, hence "absent permission of the landowner". A public footpath doesn't confer a right of way for cycling or pushing a bike. But a landowner is entirely within their rights to grant permissive usage, whether explicitly or simply through tolerance. We have a section of NCN near us which is officially a public footpath, for example, but the landowner has given permission.

Hello from a cycle router author.

No cycle router will send cyclists along highway=footway by choice, with or without access tag. Some may, however, potentially incorporate a highway=footway into a route, with a clear direction to get off and push, if it saves a long or dangerous diversion otherwise.

bicycle=no is, however, not likely to be the right tag on (for example) way/206286744. If you want to add an access tag, then the least ambiguous one would be bicycle=dismount. bicycle=no is best used for the situation where bicycles are forbidden _whether pushed or ridden_. This is the default situation on public footpaths (absent permission of the landowner, because a bicycle is not a "usual accompaniment"), as well as many permissive footpaths such as through the grounds of Oxbridge colleges.

Hi - could you say why you removed the note tags from way/319586047 in this changeset?

Hi - the long-standing convention in OSM, dating back to 2011, is that the National Byway is network=rcn. This is due to its Regional Route-like character; the fact it's not truly a national network but just covers some regions of the UK; and for differentiation from the (Sustrans/WWCT) National Cycle Network. I've retagged the superroute relation accordingly.

(writing as someone who has a spare National Byway sign above their desk ;) )

Indeed - "my" route (NCN 442, which I designed and largely implemented) has two missing sections which will be fixed one day...

NCN 43 is intended to go via roughly Tirabad and Trecastle to join the two sections, I think. But I'd be pleasantly surprised if that happens any time in the next 10 years!

Great - thank you for the survey! I think mapping it as a separate link route is fine for now.

I'm 99% sure that it's a "future route" rather than a "meant to be removed route" though. Sustrans' direction of travel is very much towards traffic-free paths rather than on-road routes, and this is a (really quite good) traffic-free path whereas the on-road route is pretty mediocre in places - e.g. the A4047 section. Improving NCN 46 seems to have been a theme throughout the Heads of the Valleys road scheme - e.g west of Cefn Coed where there had been a gap in the route for ages.

If you take a look at the Sustrans planned works map (https://www.sustrans.org.uk/national-cycle-network/our-plans-to-improve-the-national-cycle-network/) then the northerly route is shown as "New Route Section - Traffic free" and has advanced to "Stage 3 - Detailed Planning" (coloured in purple).

The garage exit is weird though and nor is it clear how the route will link to the Clydach Valley path - I wonder if there might end up being a signalised crossing across King Street by the Bridgend Inn.

Excellent, will be interested to see what you find out. Legalities aside I wouldn't trust OS maps on the NCN - OSM is famously more accurate than Sustrans' own data as to where the NCN is actually signposted ;)

That's weird - yes, certainly there was signage at that roundabout and continuing east, but also further along: I took a couple of photos of the signage east of Garnlydan. https://imgur.com/a/czocdcI

I do remember it being missing on the path through the residential area at Rassau and on the entrance to the garage near the Brynmawr bridges/roundabout.

I don't think it's a leftover sign - quite the opposite: this is the (expensive!) new route built as part of the A465 works which hasn't been fully signed yet!

Hi - in this changeset you appear to have removed the new route of NCN 46 north of Beaufort and Brynmawr, close to the Heads of the Valleys Road. I cycled this the other year and it was signposted as NCN 46 throughout. Could you reinstate it?

Hi - you've changed the towpath between Bascote and the Long Itchington greenway bridge from highway=footway,bicycle=yes to highway=cycleway. My recollection from boating along here the other year is that significant parts of it are unimproved (i.e. just a narrow unsurfaced path) and that describing it is a "cycleway" is a little misleading. Has this changed?

It has already been explained to you repeatedly that rural unimproved towpaths are not highway=cycleway. STOP IT.

It's on the OCC public highways map as a metalled road "Maintainable at Public Expense", and there's no conflicting PROW status, so I don't see any compelling reason to have it as anything other than highway=unclassified. (It would be lovely if it were a cycleway though!) OS's decision to show it as an ORPA seems strange.

Hi Mikey - this is one of my local roads! I think this was better as it was previously, to be honest. Motor vehicles are not forbidden from using the road (and in fact they do use it - it's popular for people driving to do a walk around Ditchley) so motor_vehicle=no is inappropriate. There is a legacy "Unsuitable for motor vehicles" sign, which dates from before the road was resurfaced, but this isn't a legal prohibition.

There is some usage in OSM of motor_vehicle=discouraged and I could see an argument for this, but at present anyone seeing an OSM-based map would expect this to be a traffic-free road, and it certainly isn't.

You can revert a changeset using https://revert.monicz.dev . You will need to know the ID of the changeset you want to revert - you can see all your changesets at @NCN%20updates/history

No it isn't. Whatever is shown on OS Maps, which has numerous visualisation flaws, is beside the point.

We map what's on the ground, and what's on the ground is blue signs for a regional cycle route. Therefore it is tagged as network=rcn. Please revert this back to the correct state before I have to take this to the Data Working Group.

(Whoops - I do of course mean that you've changed it from network=rcn to network=ncn, and that it should have stayed as rcn!)

Hi - in this changeset you have changed relation/8495916 from network=ncn to network=rcn. This is wrong. This is a regional route (numbers on a blue background) and as such is tagged with RCN, not a national route (numbers on a red background) tagged with NCN. NCN 51 runs from Oxford to Cambridge and into East Anglia, not in Devon.

Could you fix this change and any other similar ones you've made? Thank you.

> apologies if one has slipped through and I'll double-check this particular changeset to put it back

Fixed: way/659232442/history

Hi! Many apologies for any asphalt to paved change - I try to avoid that, but apologies if one has slipped through and I'll double-check this particular changeset to put it back.

highway=unclassified is generally used in OSM for minor roads of a principally rural nature, whereas =residential is intended for residential roads in built-up areas.

There is (famously) a long-standing issue in the US in that the TIGER import of 2007/08 brought in all A41 class roads as =residential, largely as a placeholder for future use. The TIGER import was famously problematic and fixing its many issues is a long-term project, to put it mildly. (As documented on the OSM wiki: "The large scale TIGER data input of United States street data overused the highway=residential tag. A significant proportion of these roads could be tagged as highway=unclassified or highway=tertiary under the guideline mentioned above.")

Something like the example you give, or way/17166426 for another one, is much more akin to highway=unclassified in standard OSM usage than =residential.