Ah, I see. I've reverted my changes to the roundabout now, including the central grass circle.

Have a look at osm.org/way/962036167. I've retagged this central grass island as 'was:landuse=grass' cos it already was overlapping some of the roadway prior to my changeset + so it will need redrawing too. Maybe you'd be able to have a go at this.

I'd interpret the wiki as referring to whether to map physically separate cycleways (e.g. osm.org/way/156111836 on this stretch of road) as a simple 'highway=* + cycleway=track' or to map as a separate 'highway=cycleway'. I'd always aim to map these as separate ways.

From a (non-routing) mapping perspective, it would still misrepresent the infrastructure to map this as a separate way.

This isn't a crossing, it's a signalised junction. Vallance Rd <-> New Rd is a single carriageway with no protected cycleway. Crossings usually are perpendicular to the flow of traffic.

This junction just has advisory with-flow lanes for cyclists, with broken white lines meaning that motorists, including ambulances to the adjacent hospital, can still drive into these lanes. It's not much different to all other cycle lanes.

At the junction there's just an early release of 2-3 seconds for cyclists. Although there are markings for a 'turn right in 2 phases' here... if willing to chance it when the lights change, you can veer out of the painted 'advisory' lane at any point and turn right (if e.g. no cyclists are coming from the opposite direction.) I've done this several times on early Saturday mornings.

Even if sticking to the painted lane to go straight across, it's very obvious you're still on an unprotected roadway (especially as Tower Hamlets doesn't have the best standard of driving!)

If there was e.g. an island then it would be mapped as a separate way, such as at https://goo.gl/maps/HpGo7p1pf4UyGkuq8.

The issue here is that (apart from the physically separated bit at the north of the street) the cycle track is connected to the road along the length of the road. Between each pair of wands there are links between the cycle track + the road. No house on the west side of the road is inaccessible from the cycle track.

As you had changed it, to cycle from Stamford Hill station to any of the houses here (all are on the west side), you would have to cycle the full length of West Bank and do a u-turn and go back north via the roadway.

This track (basically a protected cycle lane) is too inherently part of the road to be mapped as a separated way.

Despite the elephant track marks, this is just a junction rather than an actual crossing, basically with the blue anti-skid asphalt marked out to guide cyclists. AFAIK it is not an offence for motorists to enter them and they can definitely 'left hook' into 'crossing' cyclists turning into Whitechapel Rd from Vallance Rd or New Rd.

The UK forum 'consent' is that mapping should be of actual physical separation and that painted lines/markings are not a physical barrier/separation.

Tagging the road with cycleway=lane is the best way to represent the layout here.

I know you have a keen eye for the coloured asphalt - you could tag here with "cycleway:lane:colour=blue".

Hi. I removed this tagging earlier today, once works were announced as finished.

I'm not sure if two-way has started along osm.org/way/252614807 to Fleet St. though. I saw conflicting maps this week, but I'd be surprised if it's not two-way.

When I surveyed round there on Thu night, I thought there would need to be some additional changes at osm.org/node/258663529 where eastbound traffic would merge. Layout might have changed over the weekend though.

There could also be new traffic access restrictions at osm.org/way/4299212 now the oneway has been reversed along it also.

Yes, I'd agree "destination:" would be the appropriate prefix to use here. I will use "destination:rcn_ref" to reflect that the route being signed is a regional cycle route.

Thanks for the advice.

Hi. The guidepost in question here is indeed for the C6 route and is a member of the relevant route relation with 'role=guidepost'.

I just used "guidepost:rcn_ref" as a metatag for adding detail for the route number +'route class' signed on the guidepost. I deliberately didn't want it to be rendered on any map.

Using rcn_ref on nodes is intended for cycle node networks (mostly in Holland + Belgium) where a node (or 2-3 nearby nodes) have unique values of rcn_ref. I know the OpenCycleMap developer frowns on its use on nodes in London, as it is not a node network here. If all signposts were mapped + tagged with rcn_ref, lcn_ref and ncn_ref, then OpenCycleMap would be littered by pink + blue + red bubbles.

In the second last update to OpenCycleMap this 'rcn_ref' tagged node this appeared similar to osm.org/node/418847948#map=18/51.98970/4.50095&layers=C.

I expected it to vanish with the latest map update, but it now appears as a pink dot, suggesting to me that the OpenCycleMap developer had spotted it and reacted in some way to render it differently but I can't be sure yet if this is for the current version of this node or for the previous one.

Hi. It looks like you might have applied the wrong set of tags to osm.org/way/457358275 plus also to the 2 ways connected at the north end.

The full 'T' shape is tagged as a zebra crossing.

Hi. I've done some slight adjustments on both sides of the roundabout to slightly 'favour' a straighter route for forward directions.

The mapping should reflect actual physical separation though. I can't remember how the layout used to be mapped here, but I suspect the physical separation of carriageways was exaggerated and either didn't permit access to osm.org/way/850929142 or didn't reflect that the single carriageway ends to the east of osm.org/node/8905731637

The angle didn't look like it should provoke a turn left instruction in some apps. It seems like this is more a routing app issue than a mapping one though.

Hi. The purpose of a stop_area_group is to allow routing via public transport where there is an interchange.

In this instance, it creates a symbolic link between the bus stops for Watford Junction and the railway stops in the train station, avoiding the need to have these fully connected by footways in order to permit a continuous route by routing engines.

Hi. I did some further edits last night after another survey of the area, looking for some of the changes you had made. Most of my prior edits that you'd changed were accurate though. Among your changes that were inaccurate:

- I'd noticed during a survey on Thu that the separated cycle track (osm.org/way/888020105 ) had recently moved further west, but you had moved it further east to where a simple painted cycle lane begins

- the westbound cycle track ends on the east side of osm.org/node/6154905994 rather than extending through the crossing

- there is only a painted cycle lane on the north side of osm.org/way/693294644, not a one-point entry separated cycleway

- the temporary cycle crossing at osm.org/way/948893550 only crosses 1 motor carriageway, not 3 carriageways

- osm.org/way/971266072 is not a dual carriageway

- there is still a works access road at osm.org/way/886590035

- there is no separate northbound left-turn carriageway yet at osm.org/way/886590027

My edit wast night will have restored some of the changes that you'd made at the NE arm, though I will have been more strict regarding what is physically fully separated (e.g. the northbound wanded cycleway=track that can be entered at any point)

The mapping of cycling infrastructure is best view with the CyclOSM map, but it hasn't at present rendered to the current data.

One thing I am unsure about is the speed limits here, especially with the east-west route through the roundabout. I wonder if that was something you'd be able to shed any light on from your surveying here.

Regards,
Mac

Hi. Yes, it was intentional.

The relevant footways that are mapped separate to these cycleway crossings are footway=crossing, i.e. they're not sidewalks.

The documented (global) tagging for separately mapped 'bicycle only' cycleways is with "foot=no"... but in the UK (to my knowledge) pedestrians can legally still use them, so there is no actual access restriction for pedestrians on these. I know some OSM users are very protective about inaccurate use of "foot=no".

Instead I've added "footway=separate" just as a potential aid for routing engines to direct pedestrians via the intended pedestrian crossing rather than via the cycleway crossing.

Hi. The proposed changes I mapped here will change the layout from that seen on that video. To my knowledge the scheme (Mansell St cycle link route between CS2 + CS3) hasn't yet started at this junction.

Hi,

There is shared use signage 'round the corner', viewable at https://goo.gl/maps/PLC9oQDNR9Lte7Sz9 - from memory I believe there are now 2 of these signs at this location.

There is nothing (either signage or tactile paving) to indicate that this shared use ends between there and the toucan crossing near the Cycle Hub (https://goo.gl/maps/RKpKxfD2mnVLJZqJA).

I'd agree it could do with some repeater signage along it.

Also, it's not clear to me how the infrastructure would allow eastbound cyclists to get to St. Mary's Road from the end of Quietway 2.

Hi,

Can you check yje royal_cypher value highlighted on https://www.keepright.at/report_map.php?schema=87&error=124152741, which suggests that royal_cypher="EVIR" maybe should be "GVIR"?

This road is definitely not oneway.

Please stop doing all these inaccurate "oneway updated" edits in London. If you don't know the areas, you shouldn't be guessing.

Hi. Are you sure about the oneway=yes tagging on Gillespie Road and Drayton Park? To my knowledge this was supposed to be kept as a two-way road.

Thanks. I've done some further tweaking to the cycle tracks along here now.