I've been working on it, deleting some areas, and retagging some as `natural=wetland`, `wetland=tidalflat` when the satellite imagery makes that look likely.

Thanks for updating this. It was closed for some time during construction of the new road to the north, but I passed this way again a few days ago, and it is open again.

Thanks for updating this, I'd not heard that news.

OK, sounds like it doesn't mean anything here on OSM, so I'll see what I can do in terms of deleting the relations and ways.

I'm not very experienced with imports, but the one I did I documented what tags I used in the imported dataset and how I translated them into OSM tags on my import userpage (@keithonearth_imports). While the one import I did is far less complex than yours, and I could have done a better job of phrasing it, I think it's a good thing to do. I'm not sure about the rules about imports, but I recommend the approach.

I was going to say it's only been a couple of years since the only import I've done, and I've already forgotten the details. But it turns out that it's been 7 years. I couldn't remember how I translated the CoV data tags a few months later, or which ones I used and which ones I discarded, but I don't have to because I wrote it down.

Hi James, I've seen a number of sections of sea on the coast of Haida Gwaii, tagged as `natural=water` and `intermittent=yes`, below the `natural=coastline`, like this one: way/299759857/

This is a strange tagging convention, and I'm not sure how to interpret it. Areas of sea that are exposed at low tide? Reefs? Something else? Do you know what the features were intended to represent in the Canvec data? How could we better tag them? As the features are currently tagged they do not mean anything, and are messy.

This changeset added the same address to multiple building traces. This should not be done. If there are really more than one building with the same address the address should be added once, either on a central node, or on a landuse area. In this case the one building was incorrectly mapped as many, and I've fixed that, and deleted the unnecessary address tags.

I'm going through fixing some address errors in this changeset too. The usual stuff: the same address added to multiple buildings, or buildings with multiple addresses only being mapped with one. Only buildings with one address seem to usually be correct.

I'm going through fixing some wrong addresses for this changeset too.

oops, I forgot to link to the example relation: relation/3989104/

This import seems to include coastal regions of sea tagged as both `natural=water` and intermittent=yes`, I've just found this as relations, but there may be some mapped as a single circular way. (example relation:

Some of them seem to be areas of shallow water, probably to the low tide line. Others include deeper water. To my knowledge OSM does not map low tide lines, nor does it make sense to me to map some areas of sea as `natural=water`, as opposed to `natural=bay`, or some other named feature.

I'm tempted to just delete the relations as I come across them. Do you have any input?

Here's a link to my edit making the changes:

changeset/153187014

Here's a link to the changeset in question: changeset/123327250

Thank you for reminding me about this voteforpedro, I'd meant to give alex circuit a chance to explain themselves, and then remove the `bicycle=no` if they failed to provide a reasonable rational for their change. Or at least the opportunity to write a full sentence. But I forgot to deal with it.

They have not made any attempt to explain themselves, and I am opposed to their change that banned bicycles from this route. I've gone ahead and changed the `bicycle=no` to `bicycle=yes`. I think I got them all.

Thanks for your answer Joel. I think it's good not to over complicate things. I think it's good to represent areas with simple outlines as a single trace. If we used multiple ways in relations to represent everything that can be represented as a single way it would become much harder to edit the map, because it'd be harder to see how it's structured.

Additionally, by changing it to a relation, it made the edit history harder to find, it only part of one of the ways making up the relation.

It's not a big deal, but I personally consider unnecessary relations to be not as good an approach. In all honesty I don't know what the consensus is.

Thanks for recombining the ways, deleting the relation, and making sure the history remained intact. It looks good.

What's the advantage of making the Burnaby Mountain Conservation Area a relation with three members? It had been mapped as a single way marking the outer boundary for 7 years.

Thanks SomeoneElse!

Sorry not to see your message until now, Joel.

I also understand that the mean highest high tide line is the best one to use for the coastline.

That said, I do think there is generally some beach between this line and where the terrestrial plants start, and if there are too many overlapping features they obscure each other in many renderings. I don't think we can say with total accuracy from the imagery where the mean spring high-water line is, so spreading the coastline and the highest point of beach may or may not be technically accurate, but allows users to see the info better. Many renderings make the fact that there is a beach impossible to see, if it is entirely below the coast line.

I think it is more useful to take my approach, to keep things clear, and I hope my approach is ok.

I should also point out that it was changeset/148518163 that I deleted the railway line with.

Hi, and thank you for your work on the map.

I've noticed what looks like a mistake to me, a short section of railway that is disconnected from any other railway line by many km of mountainous terrain.

I think it's actually a section of canal, mostly obscured by the trees. These narrow canals are used in this part of India commonly for irrigation and small hydroelectric.

But I don't think we can be sure from the satellite imagery alone. I've just deleted it, but if you have been to the location, and know more than me I'm happy to add it back. I would recommend that you add "Survey" or "Local knowledge" to your changeset summary if that's the case.

Thanks for your work on the map, and I hope you keep it up!

Wow! That's a pretty funny mistake!