Thanks! Just came across some bbqs I had added on the ground in the last 6mo that now said they needed checking.

What is the URL/full name of the actmapi list, and when was it last updated?

Sure, i think i understand the theory. In this case, if i remember correctly, we're not talking about a clean parking lane here. Its embedded into the kerb with nice trees in between and the sidewalk is not directly beside but with an additional gap. No renderer can currently guess that i consider this is a nice street to walk down, so i rely on the map. The renderer needs to not only know widths and distances between parking lane and sidewalk, but also how the parking lane is split up into sections, leaving space for the trees. Moreover, every editor needs to do the same to provide an intuitive visualisation for me to edit, and we end up with having both real and pseudo objects on the map that can be edited. On the other hand,a pedestrian router could instead treat area between a road and sidewalk as a routable area (like for pedestrian areas). In my opinion for messy cases micro mapping is easier to work with than pseudo-geographical elements, but i would use parking lanes for true parking lanes without this complexity. I will go read the wiki again, but my feeling is my preference is still for coexisting solutions.

Thanks for sharing your view. I come from a pedestrian way finding perspective, so I'm afraid my objectives are different. When looking at a map, i want to know that there is a sidewalk that is physically separated from the road and it is useful to know the separation is by parking. Someday navigation software will catch up with complex pedestrian needs. In the mean time, I'll keep mapping my needs and hopefully try to be useful to car users too along the way.

The newly traced parts ended up being more precise, so I merged the two versions
osm.org/changeset/70382949

That changeset deleted quite a few other ways too, so we'd have to check what else is there. I was thinking of doing it this weekend, but if you want to have a go first, you're welcome to.

I'm a bit late to comment, but changing the level on the shops breaks the relationship with the ways, e.g. Retro should not be accessible from any street level as is shown now. You would need to update the level on the ways too.
Previously level=0 is the lowest ground level, i.e. the supermarket level, which also matches the shopping centre's labelling. What did you use?

J'ai aussi récupéré le fleuve Senegal: changeset #64189142

Hi Jan,
Thanks for checking. Yes, they're every few metres because they're used for the construction of the underlying tunnel.
However, the proposal for this feature is only now being commented on the tagging mailing list.
osm.wiki/Proposed_features/Qanat
Feel free to give your feedback
Cheers,
Joseph

Around Tabora (didn't proofread message)

I tried something like that in urheilupuisto (quickest exit towards tapiola is through a oneway door), but exit=yes didn't work at least
osm.org/node/5501156915

Looks like oneway has no effect for pedestrians (and isn't quite the right term in this case anyway)
https://github.com/opentripplanner/OpenTripPlanner/blob/master/src/main/java/org/opentripplanner/graph_builder/module/osm/OSMFilter.java#L155

In my case, I followed someone else who took the lift from P3 to 2 and realised the layout was wrong (and that route is quicker).
Distances still seem wrong, but connections are now correct (at least, I was pretty careful), so routing is hopefully ok. I had to move the platform to make it this far, so I was hesitant to do more.
I'm not sure what restrictions reittiopas/opentripplanner respects, but it seems like it might be worth adding a one-way to prevent routing down the lift?

This is a rather major edit, but it turns out the layout of Tapiola station was substantially wrong. Lift for the Ainoa indoor path from the ticket level should have been the same lift as on the bridge into the station. The position of the first lift inside the building is obvious from the building layout, but was located much further in. Escalators similarly needed to be moved inside the building outline, and entrance A onto the building outline. I'm sure there are still errors, but hopefully it now makes more sense (if you step through the layers one by one).

Sure. In this case there was a clear separation (including parking, which still isn't marked on the map).
The error in connecting to the road is simply because I was mapping on my phone while pushing a pram ...

I was only visiting temporarily.
The only sidewalks on the map were mapped separately, and I mapped the ones I could opportunistically (expecting them to be completed over time).
If this is your local area, feel free to edit as you see fit
(Personally I much prefer mapping sidewalks separately)

Why was the way deleted that defines the whole area of #7?
It doesn't do any harm, and helps nominatim users recognise that it's not just a single building/entrance

Thanks for the tip. I didn't realise history was updated by area as well as by object. I suppose I could just submit one changeset per object, when they are not geographically related.

These were all corrections to errors in tags reported in http://tools.wmflabs.org/wiwosm/wiwosmlog/broken.html
Also see:
osm.wiki/WIWOSM

Tag should be:
#hotosm-project-1113 #PeaceCorps #PCMozambique source=Bing

Tag should be:
#hotosm-project-1113 #PeaceCorps #PCMozambique source=Bing