LMCU:NOM_ARRET et LMCU:OBJECTID ont été ajoutés à node/1670199377, une partie de changeset/14247809

Yes, correct. I've now reversed the oneway direction.

True, the changeset comment is inaccurate. However the change to the junction tagging is valid, as there is a give_way on the 'circular' at node/1294552639 that you had tagged.

Hi,
The X1 relation has nothing to justify its former "public_transport:version=2" tagging.

Its ways were not ordered (although I have improved this) and it uses invalid (in PTv2) "forward/backward" roles. (Of lesser relevance it also has no stops.)

Crucially though, this relation combines the 2 directions of travel into 1 relation, so it should never have been tagged as a PTv2 relation.

Ideally, and with a little bit of work, it could be converted into 2 x PTv2 route relations for the 2 directions of the route.

Without creating a 2nd relation, the only way to make this particular relation valid (for one direction of travel) would be to remove ways that are only used in the opposite direction of travel.... but this would then lose valuable information.

Hi. There is a give_way at node/2730892177. Therefore it is not a junction=roundabout

Hi. This way is definitely correctly tagged.

There are some details on the introduction of the shared use here in a 2014 article at https://www.hackneycitizen.co.uk/2014/01/20/hackney-council-pavement-cycling-trial-victoria-park. Although this was introduced as a trial (as far as Wetherell Rd), it was never removed (or improved upon!)

A few years ago, a shared use sign beside some tactile paving (https://goo.gl/maps/LC6F5Hn1sJdR3GSJ9) was replaced at the NE end of this way by a large Controlled Zone sign (this is actually just inside the TfL 'red route'), but this pavement is still a shared use path.

To the SW, there is signage at the next junction (https://goo.gl/maps/u6Aehcgp9PSMi6QNA) for cyclists from Homer Rd who have a dropped kerb to access the pavement. There are at least 8 further shared use signs along the pavement between this point and Gascoyne Rd, plus more signs down to Wetherell Rd.

In practice, this all serves as a night-time detour for cyclists when Victoria Park is closed. As it is shared use, it is not oneway although SW-bound cyclists still have the option to use the road which has some of its pre-trial lanes. (However as mentioned in the article, "Victoria Park Road is a designated ‘A’ road used by heavy goods vehicles and buses.")

Despite being low quality infrastructure, I suspect it never had too negative an impact as it is seems to be mainly used at night + by a small number of cyclists. It's also not a busy pedestrian route at any time of day.

Re standard landuse practice, I'd direct you towards the wiki at osm.wiki/Land_use#1._Mapping_the_reality_on_the_ground:

"For example to map the land use of a school draw the edge of the shape following the boundary of the land that is used by/available to the school; to map a residential area include land that actually belongs to the houses and their gardens/private driveways.

Land that is dedicated to the public highway (both the carriageway and the pavement/verge) is not land that is available for other uses i.e. you can't build a house on it, and so it shouldn't be mapped as such."

Thanks. This is now corrected.

Hi,
None of the cycle lanes added in this large changeset are correct.

In each case, either the cycling infrastructure is already mapped as a separate cycleway or as cycleway:left

As an example, way/12109461#map=19/51.52868/-0.01649&layers=Y has a stepped cycle track that is mapped as separate cycleway, but you've further added non-existent cycle lanes on both sides of the road.

Traffic on the approach roads do have right-of-way when they have a green signal at this junction and traffic on the "roundabout" have to give way to them (at their red signal). Therefore this is not a junction=roundabout.

See junction=circular:

"The tag junction=circular is used only on road intersections where traffic on the junction does NOT have COMPLETE right of way."

"Where traffic on a circular junction itself does have complete right of way, this is a roundabout. The tag junction=roundabout should be used on roundabouts."

At junction=roundabout:

"A roundabout is a... road junction where the traffic on the roundabout has ALWAYS right of way."

It mention, "There CAN be traffic light, as long as the circular road has right-of-way". In other words, if there were only 'non-junction' traffic signals on a roundabout (e.g. for a pelican crossing), then it would still be a junction=roundabout.

The station entrance is opened but the rest of the building is still fenced off (including the Oxford St side) as construction seems unfinished above the station entrance.

Hi. I've removed what I'm sure would have been errant extra tagging with 'node=2133959589' from 4 areas of grass (in changeset/120430257).

Hi. The errant tagging has now been removed.

Thanks for pointing this out, Daniel.

I've now fixed the error gap here.

Regards,
Mac

Hi. I've changed Vallentin Road to oneway=no.

The road has new a 'no entry' restriction at its west end (buses + cycles are exempted). This is already set as a pair of turn restrictions.

General motorists can still enter at the east end of the road, travel westwards almost to the end of the road and then do a u-turn and exit the same way they entered.

Hi. I've corrected the turn restriction relation/13933331 into 2 separate 'no left' + 'no straight ahead' turn restrictions (except cycles).

I can see that you've now resolved this issue.

Thanks. These are all now removed.

Hi,

"crossing=toucan" is not an approved tag for a crossing with traffic signals. See crossing=*?uselang=en-GB#Approved_tags

"toucan" is only valid as a value for "crossing_ref". See crossing_ref=*

Since changeset/115290764 the extent of the relation for "EV1 in Spain" also included the Ireland section of the EV1 route.

This was due to a circular reference within the relation/2763798 parent relation, with this relation wrongly being included as a submember of one of its child subrelation (Spain's EV1 superrelation).