Wrong changeset description. Should read:
Removed duplicate place node for Sehzade Camii Prince's Mosque

Do you verify the location of the airport metro station of Goldline? I am not sure it is actually at this location judging from the construction visible on satellite imagery. Would be best to be verified on the ground

Calgary of course, not Edmonton *facepalm*

Restored metro station in changeset/73183128

Can you explain why you re-tagged LRT line 2 to light rail? The name is light rail, because it started like that, but line 2 has meanwhile changed to heavy metro rail vehicles, the infrastructure has characteristics of a rapid transit (metro) system, such as high passenger throughput, exclusive right-of-way and use of full metro rolling stock. I would suggest you revert the tagging of infrastructure and routes back to subway therefor.

In regard to maintainability: As you already pointed out upon changes to a line (like opening of finished segments) we always need to update both individual nodes and ways as well as the accompanying relation. So it's just the question which is the cleaner way to capture the elements of the members of a route. As I said earlier I think it is not correct for the route "HarbourFront <-> Punggol" to include elements that go beyond this section, even if they are tagged as inactive. Also for data consumers it's easier to tell that a "construction:route" / "planned:route" / "proposed:route" relation is actually a whole route section that is not active (and most likely discard it). This can of course also be realized by using the lifecycle prefix on member roles. So in the end it's just a question of taste and preference.

If it's easier for you to maintain the routes with tagging members we can re-tag Singapore accordingly.

I think it makes sense to tag network as the modality and operator as Transport for NSW.
I have updated the source document for the Subway Validator to cover light rail and trains as well. Should be included in tomorrows run and then we can work through the list of routes and stations to update :)

I agree with you wholeheartedly. This combination of monorail and APM has been practiced in OSM data for years though. So if we come up with a new tagging variant it should be a distinct improvement covering more variations like the air cushion funicular at https://en.wikipedia.org/wiki/Dorfbahn_Serfaus or the cable propelled https://en.wikipedia.org/wiki/MiniMetro
One could even argue that these should not even belong within the railway=* tagging group and instead form something like transit=* - For simplicity of the tagging I would suggest to keep them inside railway.

The simplest would be to just summarize all these different movers as railway=people_mover and then specify further via people_mover:type=automatic or people_mover:type=funicular

No. I've deleted it meanwhile

I see. It would help if there was consistent network=* tagging applied on the Sydney Trains. I can see one of the two:
- TNSW - Sydney Transport
- Sydney Transport
Most routes are consistently tagges with operator=Sydney Trains
Judging by the articles https://en.wikipedia.org/wiki/Transport_for_NSW#Public_transport_services and https://en.wikipedia.org/wiki/Sydney_Trains
I wonder if the network name shouldn't be "Sydney Trains" consistently. I guess it's a question of whether you look at the network from a modal level ("Sydney Metro", "Sydney Trains", "Sydney Light Rail") or as the intermodal overarching public transport network ("Transport for NSW" or just "Transport").
Once you have picked a consistent network tag I can set them up in the Validator as well

It can include other mass transit routes as well. Are you thinking of Sydney light rail?Any other networks in the Sydney area that could be included? I will update the document then.

We discussed this in Telegram chat a while ago and Line 2B extension has its own relation with proposed:route tag

See these changes applied here:
changeset/71131794

The North East Line MRT currently operates from the Souther Terminus HarbourFront to Punggol. That's what the active route relation should cover. It would be quite odd to include the extension section to Punggol Coast which is planned to open no earlier than 2023 in a route relation that has Punggol as endpoint.

The public transport route model in OpenStreetMap says that multiple line/service variations are linked via a route_master relation. I suggest to add a route:construction relation for the segment from Punggol to Punggol Coast and add this route to the route_master.

Anytime.
I assume you are aware of the OSM subway validator which helps to keep public transport tagging correct and complete: http://osm-subway.maps.me/index.html

Seems I fell for a trap of some "funny" mapping there. There used to be a node on that location which I spotted when reviewing some religion tagging mistakes: node/6094463057/history
I moved that information to the way. At least this change helped surface the wrong data and you could fix it. Thanks for correcting this.

You are welcome. Looking after each other is what makes OSM data quality constantly improve.

Yes, they are "a millenarian restorationist Christian denomination with nontrinitarian beliefs distinct from mainstream Christianity"
https://en.wikipedia.org/wiki/Jehovah%27s_Witnesses

See also the listing within christian denomination values in OSM: religion=christian

Thanks for fixing the operator tag. I was following information I could follow on the home page back then.
Will also reach out to better understand why you are not tagging as light_rail

Sorry for the late reply. Are you relying on these meta-relations that act as collections of objects?
What's your use case? I am happy to help finding a better alternative. Relations in OSM are not intended for this usage and hard to maintain.
Quote: "[Relations] are not designed to hold loosely associated but widely spread items. It would be inappropriate, for instance, to use a relation to group 'All footpaths in East Anglia'." from osm.wiki/Relation