For example way/444889956 is also viable, it says you can walk from the bottom of the steps to the edge of the platform (which you can) and allows the routing network to connect the platform to the road/footway network. Would that work as a workaround?

I disagree with the JOSM warnings for source:geometry on a node. I'm using it to document the source for geometry of the point, in the same ways as is done for the geometry of a way. I think using different tags like source:geometry source:location just based on node vs way adds unnecessary complexity.

This is an issue in the routing engine, not in the OSM data, but I understand we sometimes need to use workarounds until the software support catches up. It's good practice to have one feature, one OSM object osm.wiki/Good_practice#One_feature.2C_one_OSM_element this means if there is one platform, then we only have one OSM object.

If I want to count how many platform there are in a given station, I should be able to just count how many OSM objects there are.

Ideally every platform would have a footway or steps leading either inside it or snapped to the edge of it. Could you just detect any routing requests from inside a platform snap to the nearest footway on within the platform area?

If the railway platform is already mapped as an area, is there any reason to duplicate this with a linear way down the middle?

I think we need to decide if it's worth continuing from this and cleaning it up, or revert it all and start again.

Hi Sean,

Check out osm.wiki/Indoor for details on how to tag indoor features, as a lot of these features are missing the required tags to give them meaning.

the level tag level=* is a very useful combination so that the result is viable in things like OpenLevelUp, eg https://openlevelup.net/?l=-1#19/-33.79724/151.18117

It would be better to go through the JOSM validator warnings which popup before uploading, it can fix things like duplicated nodes, warn you about untagged ways and other errors.

Are you absolutely sure all the features you deleted are gone now?

Also it's better to keep existing objects like https://osmlab.github.io/osm-deep-history/#/way/73149021 why was that deleted?

This is a pretty big change, and I can see you've tried to ensure the existing mapped shops like Big W, Woolworths. Maybe if you're going to do more, post the osm file first so we can review the import before it happens, which is best practice per osm.wiki/Import/Guidelines

The other issue is you've now mapped 3 parks, based on what looks like the cadastre/legal parcels.

It's good practice to map what's on the ground osm.wiki/Good_practice#Map_what.27s_on_the_ground , and on the ground there is just one park, can we change this back?

Hey Warin, In this changeset you deleted way/608927871 as "delete cycleway - not there". How do you know it's not there? It was only added one month before you deleted it, so it's very recent. It shows up on both the recent NSW LPI Imagery and the more recent photo I took https://www.flickr.com/photos/136319147@N08/45324895234/.

I've reinstated that deleted way.

looks good to me

https://osmlab.github.io/osm-deep-history/#/way/183817729

There is a plaque at the diversion between the lookout and boardwalk with a different name, so I've changed the name you added to the boardwalk based on that.

It looks like you accidentally moved gate 55 in this changeset. https://osmlab.github.io/osm-deep-history/#/node/1099313596 I've moved it back, but next time take care when dragging the map in the iD editor as its very easy to move nodes when you're trying to pan the map.

I think the original tagging of highway=service makes sense for the road to Hillsong as it matches highway=service and isn't a residential area.

I've changed this, adding the LPI Base map one as an alt_name and leaving the original in tact pending further evidence from the ground.

What's signposted could very well be different to what's on the LPI Basemap. You should only add that as an alt_name unless you've surveyed the sinage.

I you added the wrong source tag, this is NSW ;)

I'm not sure about this one, I have a suspicion that it's signposted as the fire trail from the highway. I've left a comment on the original entry for more information at changeset/22666062. The Mapillary imagery here doesn't help.

Hey, in this change you added Mirang Road Fire Trail from the highway way/285602681, an armchair mapper since changed this to Eckersley Road, can you confirm if Mirang Road Fire Trail does infact start from the highway or not?

way/639199871 is a duplicate of way/636736592 so I've deleted the duplicate.

It looks like the lookout is already mapped at way/208009693 with the formal name and includes "Mount Ainslie Lookout" as the alt_name which geocoding engines should pick up, so I've deleted the one you added as a duplicate.