An interesting discussion, and not sure how best to deal with it. But initially, someone (?) needs to decide what to do about the 'gaps' that now occur in major routes, e.g. NCN1 which is also part of the European network, has gaps that weren't there before. How utterly confusing, especially for european tourists who come to the UK.

I shall take that as an invitation to post here, rather than register for yet another group.

It is going to be difficult to define an OSM policy relating to on the ground signing of routes that are under someone else's control, or no ones control, and where the current controller has said that they will remove the signs over the next two years. If the routes are useful the solution is above OSM level in the political sphere: someone else, presumably local authorities has to take responsibility. OSM could then continue to map what exists.

There may be another solution, if you consider that if the routes had been created today, in the era of GPS they might not necessarily have been signed. This suggests that the solution suggested in the wiki, of marking them as 'NCN abandoned', or similar, perhaps 'NCN assassinated' could be used in cases where OSM considers that they are worth marking. The latter decision is effectively what OSM does for roads and tracks when grading them for traffic levels of bicycle use.

**************************

Examples:

My comments are based on the Avon Cycleway, which seems to be one of the routes degraded on the new map: https://osmaps.ordnancesurvey.co.uk/51.41362,-2.75577,13.

It existed and was signed before Sustrans, with brown 10 signs, and the signing, now blue 410 sometimes with a brown 10 inset, will presumably continue to exist for some while, unless budget is to be diverted to an energetic sign removal programme. The signs are useful, particularly on the link routes, when picking your way round the edge of parts of the town that you do not know well, or just to avoid stopping to read the map at every junction.

The cycleway has always had the problem of some busy sections, which were marked on the original maps, for example the B3130 Barrow Road road to Winford is in the Clifton Link, still signed and still used by cyclists as the direct continuation of Hobbs Lane. One of my 1:50k OS maps shows this replaced by a short stretch of the A38 to reach the Felton Road, presumably in the pious hope that the pavement would be designated as shared. I have long ago given up using the B3130 in favour of walking the 600 m up the A38. So you could argue that, like the Sustrans network, the county cycleways never existed in complete form, except as a route suggestion for experienced cyclists, or in quadrants, using the links for less experienced cyclists who read the warnings on the map.

It might be that had the county cycleways been created today, in the age of online planners and GPS, no one would have expected signs on the ground, or possibly would have concentrated on signs warning of the presence of cyclists in the busier sections. Perhaps that is the real utility of the blue route signs.

The Sustrans changes are rather odd. The quiet narrow back road down the Gordano Valley has been reduced to 'not on the National Cycle Network', whereas the drag strips of Manmoor Lane and Kenn Road, which have some very fast traffic, remain as 'On road route on the National Cycle Network', presumably because they line up with the Strawberry Line. Anyone using the new classification as a guide to traffic conditions could be very surprised.

Keeping NCN41 north through Oldbury is perfectly sensible, reducing NCN410 from Thornbury over to Horton and then back round north of Yate seems odd as the roads are much the same quality, and it actually provides a useful guide to some of the more obscure bits in the urban fringe. Presumably because the principle applied has been to tell the work experience lad to downgrade NCN410 apart from the off-road bits.

The route North through Henbury to Cribbs Causeway is shown as a traffic free part of the network, which it is, if your definition of traffic free is teetering along a narrow shared pavement. Then there are the odd doglegs. Who thinks it is necessary to dogleg NCN4 along Reedley Road, Stoke Lane and Stoke Grove rather than heading directly to Coombe Lane? Strictly, I think this is also a local link route, not NCN4, but the OS map does not make the distinction.

I think this points to the conclusion that the NCN is a mess that has never formed a coherent network, and Sustrans moving to define its job as traffic free routes only may be sensible, even if this leads to an NCN comprising a series of fragments. This may be the object; if the gaps in the NCN become obvious it may be easier to ask for funding to fill them with off road routes. This does leave us with some perfectly useable routes, which still have signing, and are worth maintaining as ideas for outings or utility use. I suggest that where signing exists the route should continue to be marked on OSmap with the number on the signs, at least for sections that are considered useable. Possibly we need a general tag for 'Duff bit of offcial cycle route' for the parts that are not, rather than leaving users to work this out from the details.

We do seem to be heading towards a National Cycle Network comprising all the parts of the current network that I now avoid or hurry through, because they are bumpy or busy, or convoluted.

On reflection, the reason for removing the Gordano back road from the NCN could be that Sustrans wish to promote a new path along the line of the former Weston, Clevedon & Portishead light railway and having an NCN already marked along the far side of the valley, might detract from promoting that.