what's your source for the building type changes?

highway=track is for "Roads for mostly agricultural or forestry uses." (highway=*) ie. for roads which you can drive a motor vehicle on. This is a single path track constructed for mountain bike use so highway=cycleway is best in my opinion, but I understand that cycleway is contentious, as an alternative we can use highway=path + path=mtb. I've added the path=mtb and restored to cycleway for now.

The bicycle access and highway classification are two independent settings. https://www.mapillary.com/app/?pKey=543018103351921&focus=photo makes it abundantly clear that bicycles are only forbidden from that point onwards, so bicycle=yes can still remain up until that point. The issue that bicycles have no where to go/exit at that point is one routers solve and not something we should address in bicycle=* tagging.

I agree that the distinction between motorway_link and trunk here is debatable, but due to the length and service road joining Batman Avenue I would lean in favour of trunk until the tollway entrance sign.

Hi I realise this is an old change but what source/justification did you have for creating a bridge at way/596670708 ? Footways crossing waterways could be a bridge, or the waterway could be a culvert, or it could be a ford (creek crossing). Without a survey or very good imagery it's impossible to know. In this instance it's very unlikely to be a bridge and likely just a ford (creek crossing).

Since it looks like you've done this as part of a coordinated effort at Kaart (?), how widespread is this, were there other similar changes?

I also restored the deleted inner way from https://osmcha.org/changesets/120119176 which represented that this ridge area is excluded from the tree cover area from the Bing imagery that looks roughly okay. Unless you know that this area is covered with trees and you really did intend to delete it?

Hi welcome to OSM. In your edit here it looks like you've accidentally tagged the trees landuse multipolygon as a lake instead of the around the lake area that you improved the geometry for, you can see your changes at https://osmcha.org/changesets/120119031?filters=%7B%22users%22%3A%5B%7B%22label%22%3A%22markw11%22%2C%22value%22%3A%22markw11%22%7D%5D%2C%22date__gte%22%3A%5B%7B%22label%22%3A%22%22%2C%22value%22%3A%22%22%7D%5D%7D. I've fixed it by placing the lake tags on the proper area and restored the trees landcover tags.

I added the turn:lanes tagging and tweaked a bit per Maxar imagery which appears to show more recent changes to lanes changeset/120374112

It looks like you've added a driveway for the full admin boundary here at note/3157838#map=19/-38.01944/145.39789 did you mean to just map the driveway next to it?

For the ways like way/929237617 and way/929237615 I couldn't see any route signage if these have no formed path and no route signage then per osm.wiki/Australia/Walking_Tracks these should likely not be mapped. Are they part of any formal route?

I've updated this in changeset/119327254

It's generally considered better practice to use turn:lanes=* on the existing way rather than adding a new way where there is no physical separation, which does not seem to be the case here from what I can tell.

I'll update the tagging to use turn:lanes

hi I think it's better to use the hazmat=* tag, so hazmat=no for this. Is there a reason for using access:conditional?

This just looks like you're taking the piss. I believe Dian was just making the point that turn lanes shouldn't be split from the main way 100m up when it's just white paint that separates them, and instead you should use turn:lanes=* and similar tags. It doesn't mean you need to push the slip/link road so far that it's a 90 degree angle, for example https://osmcha.org/changeset/118114029 is a better middle ground in my opinion as it mostly keeps the way to the centerline of the road but avoids extreme 90 degree joins.

per note/2941525

reverted in changeset/117869922

Hey Graeme, the note was for the missing steps from the path here up to the lookout. I left a note until we can get a gps trace to map it. So I've reverted this changeset in changeset/117870564 and will re-open the note. Mapillary also shows no steps on the cyclepath.

The only argument I can see for having it in the harbour is it may be a better labelling position cartographically (less clutter and easier to read), but that shouldn't be our focus, we should map it accurately as data first and foremost.

I agree with repositioning the place=city Sydney node from the harbour to the CBD. The CDB is the centre so that's where the node should be.

I've deleted node/9506803261/history as that's not an appropriate use of place=plot per the wiki and we don't need to keep a node to show where the Sydney city node used to be.

I didn't get around to asking on talk-au, but I noticed someone else restored place=city, which I think is more correct. I wouldn't call it double mapping because the node is part of the relation so data consumers can identify it as the same feature. So one feature, but two geometries, one as the area, one as the central point.