Opencyclemap renders two Helsinki bays as ground and I can't figure out why

Hi folks,

I use opencyclemap.org a lot, and it’s bugging me that these two bays don’t look like water on it. I can imagine the situation as confusing for people without local knowledge.
Töölönlahti
Pikku Huopalahti

I’ve only touched the second bay. It used to be a way as the first still is. Some ages old topic from the bygone days of the help.openstreetmap forum suggested to change the way into an area by adding area=yes. Ok, I did that, but it did not change anything. After that I tried to copy how other bays nearby are mapped - that is, by using relations. I’m a relative newbie, but anyway I ventured to change the second bay into a multipolygon relation to see if that would look normal. That was about 15 days ago, I think it should’ve updated on Opencyclemap by now if that was what the right thing to do, but alas, nothing.
Can anyone figure this out? Thanks in advance.

Does the rendering on the Standard, Humanitarian, and Tracestack Topo layers on the osm.org home page look right? (It does to me but I don’t know the area). If those are OK, maybe OpenCycleMap has not updated yet.

Coastline is generally handled separately in rendering, so updates often propagate slower (in the OSM org carto layer, that’s several days instead of minutes). Possibly, OpenCycleMap updates coastline even more seldom.

1 Like

There’s an old thread (from 2022) on the Finnish section of the community forum on this subject.

It indeed seems that there’s a problem with the coastline data updating, apparently, very slowly on OpenCycleMap. Currently, most of the other map layers render both Töölönlahti and Pikku Huopalahti correctly (even the ‘Humanitarian’-layer that used to have problems with them).

1 Like

Thanks for the responses. I had no clue slow coastline updates could be a thing that exists, good to know.
Tolstoi, thanks for the link. It didn’t occur to me there would be a discussion in Finnish - did all I could to search for english discussions on the matter (expected that since the reach for finding answers is greater).

To answer your question, Alan, the issue exists on three different maps available on the menu (in web browser) - these are CyclOSM and Public transport in addition to Opencyclemap.

Is there a way to verify that slow coastline updates are indeed the cause? I don’t know how to check historial records of how the coastline has looked previously, for example.

A slow way to check is to refine the coastline somewhere else, and check in on that later. If your refinements appear but the bays are still broken, there might be something else.

That’s a good point! Also, I just compared the coastlines of OSM and OCM in a few places where I know land reclamation has been done. Comparing e.g. the coastline of Hernesaari on OSM and on OCM, OpenCycleMap is clearly out of date. The history of that section of the coastline shows that that particular edit was made in February '23, so almost one and a half years ago!

The fact that e.g. Töölönlahti has not been rendered corretly for two years is surprising, though. Maybe OCM is having some topical problems? That Töölönlahti and Pikku Huopalahti render correctly on (most) other layers suggests that the data itself is hunky-dory & kosher. Perhaps edits on those waterways were just made at an inopportune moment, when the coastline data had just been updated on OCM.

area=yes is only needed if the tag is valid for open and closed ways and it is unclear if the way should be treated as linear or as an area which is not the case for bays.
Mapping the area as an multipolygon relation with only a single member is discouraged as it makes it more complicated without benefits and is definitely not needed in your case.

1 Like

If this is hosted by OSM France, someone has suggested https://github.com/osm-fr/infrastructure/issues as a way of reporting these issues with the French infrastructure.

Perhaps try there?