Hey Guys,
I think we’ve lost a large area of mapping for Canada - if you look here:
You can see that suddenly a large amount of detail has been lost on the right along a straight line. I was using this area’s mapping last year and there was full detail available, however this year this data has been lost.
I’ve literally no idea how to go about fixing this, or even if it is possible! But I wanted to make sure that I’d raised it…
Matt.
I’m not seeing any clear recent changesets in the area that would indicate any mass removals of data.
@jmarchon Do you know if this could be related to the Canvec imports in Canada?
At least in the last 12 months or so I can’t find any deletions here:
https://simon04.dev.openstreetmap.org/whodidit/?zoom=12&lat=50.53971&lon=-89.88164&layers=BTT&age=1000
This is an overpass query looking for any historical nodes in part of your blank area. You can change the date to search earlier or later. I can’t see anything there 1, 4 or 7 years ago.
I choose one of the objects on the sharp line randomly:
Only the western part of this lake is mapped.
I do not think that anything got lost there. The sharp line probably came due to a data source that did not proivide data for the rest of the area. So the missing data was never added.
It is more likely that it hasn’t been imported yet. See Canada - OpenStreetMap Wiki section “What’s with the forests in Canada?”
I’m on my phone at the moment so I can’t examine it too well, but I can say for sure that I have never been active in that area (so far) and have not deleted any data. It appears to be a fairly standard Canvec tile border to me. I’ll take a closer look later today or tomorrow and can import some more data there soon.
I feel like I should update my original post to provide more details. I was using a map for this area last year - THROUGH GARMIN EXPLORE - which had full details available. This year the data is lost. I raised a ticket with Garmin and they reported that the issue comes from Open Street Map - hence I assumed that the data was available in OSM last year, but maybe this is not the case?
Garmin is reporting to me that the lack of data in their mapping from last year to this year is caused by changes in your dataset, but it is entirely possible that this is not the case.
I have already reported back to Garmin that some of their products (namely Earthmate and Connect) do not show any drop in fidelity. So far they have not replied on this and I will report back to them your response that there was no data for this area last year.
Of course, it doesn’t change the fact that there is a large area of mapping missing - I have no idea how to fix that…
Thanks for looking - appreciate that everyone here is a volunteer. Matt,
It might be useful if you could post a screenshot of a Garmin map that has reasonable data in this area and one that doesn’t. I find Garmin’s use of map data quite confusing - just looking at the Connect website in my area, maps in some contexts use data from Here, while other contexts use Google, for no obvious reason. I wonder if the specific app you are using switched sources since last year.
Unless Garmin can explain how the issue comes from OSM, it sounds like they are just shifting blame so they don’t have to look into what the problem in their data pipeline may be. Multiple people have investigated and found no loss of OSM data in this area so I’d say the ball is pretty clearly in Garmin’s court.
Indeed, it looks as if Garmin has switched to OSM as its map source at this location and OSM still has only limited (imported) data available there so far.
I can’t comment on Canada, but in the UK Garmin switched to using OSM for e.g. POI data a few years ago.
It’s perhaps also worth adding that my experience of Garmin’s helpdesk people suggests that they’re a bit less knowledgable around data and device use than Garmin users in various forums elsewhere 
.
100% - and I have raised it with them, with evidence that this is a. a new issue (I reported it to Garmin about 4 or 5 months ago - but it wasn’t the case a year ago) and b. that it is inconsistent across Garmin products. They asked me to raise a ticket with you guys, which I’ve done and found that there is no data loss and nothing has happened to this area of the world.
I suspect that Garmin has a hierarchy of mapping provision and what has happened is that they have changes that ordering.
OK. I think Garmin might be onto something with this. If you look here:
You can see the area where the mapping abruptly stops (you can trace the lines along - it is a very large area). If you change the layer to tracestacktopo, the lakes and extra detail then appear:
I totally appreciate that you’re saying that there have been no changes to this mapping in the last 7 years, but I am absolutely certain (though unable to prove it with screenshots) that last year this mapping displayed in Garmin and therefore I suspect on OSM too…
Any ideas? M>
If you zoom in past zoom level 11, it’s empty also in Tracestrack Topo:
Another Tracestrack Topo example from northwestern Ontario:
I particularly notice the different colour of the lake part that does not appear when zoomed in.
If I had to guess, Tracestrack gets landcover information from non-OSM sources when more zoomed out (before zoom 12), either in areas where OSM data is missing, or in general.
Right. I hadn’t noticed that and you’re correct. Is there a solution to this? Or do you think the mapping is just not available? As mentioned last year I relied heavily on this mapping and it was working (through Garmin Explore). M>
Thereʼs a subforum here for Canada. You could ask there about importing more from Canvec or possibly about mapping there by other means.
I can do some more Canvec importing in the area. Won’t be an instantaneous solution but give me a while and I can get a pretty significant area done. In areas like this with relatively large amounts of existing data, conflation can really slow down the process.
Er - looks like you already have
Alas no. I don’t use Garmin’s online services so I can’t say how frequently those are updated, but the on-device POI data on a Garmin 65s I bought in early 2023 dated from between 19/8/2019 and 18/10/2019.
I found that out by looking at OSM features added after a particular date and seeing if they occurred in the Garmin data. It would be possible to do something similar with Canvec - try and find the “oldest thing in OSM that is not in Garmin” (using a version of the overpass query that I posted above).
