Hi,

This section is definitely two-way. I just should have spotted + removed this access tag when I split Ashe St into separate one-way + two-way sections.

The original access tag had been added in osm.org/changeset/17492047 before the road was split.

Hi. Are you sure that "cycleway:both=track" is correct for Balliol Road + Kelfield Gardens?

It looks suspiciously like "cycleway:both=no" might have been intended here.

Hi. This changeset changed an accurately mapped short section of single carriageway into a 'physically separated dual carriageway' as a supposedly "necessary" change. In addition this messed up several bus route relations.

Hi. I've adjusted the tagging here to "maxweight:imperial=5 ton" which might reduce confusion between ton/tonne.

Hi,

Adding role=route to any members of a PTv2 bus route (i.e. route tagged with public_transport:version=2) is not permitted. All ways in the route relation should have an empty role. See osm.wiki/Buses#Adding_streets_to_the_relation

Hi. Names seem to have been added in osm.org/changeset/34080984. My edit was purely to make the relations more valid as PTv2 route. I wouldn't have any knowledge of the bus stop names.

Hi Andy,
My edit was just to join 2 disconnected segments of road that had 2 overlapping end nodes instead of a single shared node. It had merely come to my attention via OSM Inspector's 'Routing' layer.

I can't comment on the primary road status here, but on zooming out I can see that the surrounding area looks very suspect with regard to highway=primary.

I could easily revert my changeset if that would be appropriate.

Regards,
Mac

It seems that OSM Inspector's various different tests are carried out at different times, e.g. PTv2 tests at about 11pm, Routing tests (http://tools.geofabrik.de/osmi/?view=routing&lon=-0.66788&lat=51.94329&zoom=8&overlays=unconnected_open_ends_1) at about 6pm.

It seems to me that the various tests might get updated on the page at different times of the day too. Where OSM Inspector mentions the time of the update that is the time in Germany.

Until the past month the PTv2 results were updated at midnight almost on the dot, but have been anywhere between 12am-1am recently (including before the clocks changed.) The update for the Routing test seems to update at a later time during the night despite being tested hours earlier.

For stops, tag is "public_transport=stop_position" with role=stop. Stops/platforms should be the top of the relation before the ways/roads. If both stop+platform are used for an individual stop, the order is stop1-platform1-stop2-platform2... (i.e.stop before platform).

role=station isn't valid in PTv2. A station should have an empty role, but ideally a stop and/or platform should be used instead.

Also, light_rail routes (e.g. at Stansted Airport) are flagged as invalid just because light_rail wasn't mentioned in the PTv2 proposal!

Hi. My changeset actually corrected the bus routes.

It was the most recent changeset for osm.org/way/145395450 that removed this (or possibly another nearby) way from all the bus routes that were affected.

I merely fixed the gap after seeing the issue on OSM Inspector, but the PTv2 validator only refreshes once daily at around midnight-1am in UK. Tonight's results show these routes as valid; I have had a go at correcting the invalid London-Norwich trains which I expect will now be valid after tonight's update to the validator.

BTW I don't use any plug-in for reordering routes.

Hi, there's something wrong with this conditional turn restriction. From Peter St to City Sq would be a LEFT turn, but the restriction type is set as an 'only_right_turn'.

Hi, osm.org/way/778812607 is clearly incorrectly drawn as a service road by the AI and needs reviewing.

That node is a stop_position (railway=stop). It is not a station node (railway=station) and there should also be a station node.

Hi, I checked around this area again yesterday and did some more rough edits based on photos I'd taken, including a photo I took over the fence at the recent road closure by holding my phone high in the air.

I'm fairly confident that the asphalt sidewalk to the west of the paved sidewalk (osm.org/way/829265241) is just temporary as the road from Monier Bridge would curve in the opposite direction to the current bend.

This bend never made sense to me prior to it being opened for the current alignment.

The building(s) north of osm.org/way/826630433 now seems finished. I was able to walk around most of it's perimeter. Will update the map here when I find the time.

Regards,
Mac

Hi Andy,
It definitely does seem like the these bus stops should be removed as https://bustimes.org/services/140-penicuik-musselburgh#map matches the OSM route.

My changes were simply to put the roads in the correct order as should be the case with PTv2 relations.

Feel free to remove the bus stops if you're confident that they aren't being used e.g. by some route variant.

Regards,
Mac

Hi,

I checked out most of the Warner Place Liveable Neighbourhood today... and unfortunately none of the roads I checked have gone oneway yet + no modal filter either. I've retagged these upcoming changes with "proposed:" prefixes for now.

I didn't check the west end of Gosset Street, so I left that as it is tagged, although I doubt the two-way has been removed yet there either.

Shouldn't be too long till the oneway does get enacted though.

Mac

Hi. I tend to do most of my editing in Potlatch but then if the changes might involve bus routes I open the same changeset in JOSM and check/repair if the roads in e.g. PTv2 bus routes are still in correct order in the relation. In checking the bus routes I think I might added a missing role=forward to one or two of the bus routes which probably explains Milton Road/Elizaebth Way roundabout.

Unfortunately with Potlatch when you split a way in 2, although it maintains both fragments of the way in bus route relations etc,, it seems to assign the 1st fragment (in the direction the way is drawn ) as the pre-existing way (i.e. the one with the history) and I'm not aware of a way in Potlatch to override this.

You're right about the absence of crossing island on the east side. I'll have a look again at this and amend it.

It could equally be said that during recreational cycling a family with kids might put off using an (apparent) cycle route!

This could be the best off-road cycleway in the world, but that still doesn't make "what currently makes for a good loop round the park" an actual "network=*cn" cycle route, just a 'fantasy route'... which is why it should be excluded.

Any signed/official cycle route that is rubbish is an actual route though (even if it might include sections where cyclist have to dismount.)

The Lambeth audit just showed the circuit in their map as this was considered "suitable" for novices doing Bikeability cycling training courses. See https://bikeability.org.uk/bikeability-training/bikeability-level-1/: "The Bikeability Level 1 course aims to develop mastery in cycle handling in an off-road environment". There are 4 providers of this training in Lambeth listed at https://bikeability.org.uk/find-a-course/?la=lambeth

The Lambeth Bikeability map you're using as a source for cycle routes is just not a appropriate source for the 'local cycle network'.

BTW in Teddington there was a plan for Q21 to pass into Bushy Park but that apparently has been blocked by the park's authority so, even when it is the intention to do so, it's not always feasible to deliver routes inside parks.

Also, if you look at the Lambeth LIP3's Healthy Routes map (essentially future Cycleways) the only proposed route around here will be along the road outside the east edge of Brockwell Park - see https://moderngov.lambeth.gov.uk/documents/s98795/Appendix%202%20Draft%20Local%20Implementation%20Plan%20IC.pdf#page=26

It's a path that is indeed "suitable for cyclists of all levels" as suggested by the bikeability level. Every shared use path in the country would have an identical bikeability level. That does make them cycle routes.

Similarly all yellow residential road on their map are "suitable for cyclists at Bikeability Level 2", just as official Quietway residential would be but they aren't all routes.

As the website states, "Lambeth's roads have been audited for 'Bikeability' which is split into three levels and taught through cycle training which is available for free to all students, residents and workers. Our cycle map shows all roads and their Bikeability level."

"Suitable for" does not make any of these paths/roads actual cycle routes, not from a OSM perspective.

Level 1 equate to a shared use footway/cycleway on OSM and Level 2 to a highway=residential, so the bikeablity of these are already represented in the dataset.

The 2004 LCN map clearly shows that a proposed route had instead been planned on the east side via the roads OUTside the park and I think road improvements were made there too just a few years back.

It's just a traffic-free path though, not a circular route. The Lambeth map is a 'cycle map' (rather than an actual 'cycle route map') with most roads in the borough coloured by bikeability.

The bits on the 2004 map were already mapped as the only members of this relation in changeset 63853583, but the extra ones you added in just don't belong in a route.