Closed PaulUithol closed 8 years ago
Could you send me a link to the exact OSM dataset you have loaded here? I'd like to reproduce this.
Hi Nick, I get this with any dataset we've tried; different base layer, different POI data, even no POI data present at all but a couple of forms/points already surveyed. But for example, try https://drive.google.com/open?id=0Bx3_hrMYdNwLY3VGaXZrOVRwUDQ
Occurs with the same data as #121: one form, no osm data, one mbtiles for background.
Form: https://drive.google.com/open?id=0Bx3_hrMYdNwLUV9RZXF6U0pKb1U Mbtiles: https://drive.google.com/open?id=0Bx3_hrMYdNwLN0taTUJUbXZWdTg
With two, which one gets selected still seems somewhat logical. However, that quickly diminishes when you get more (like the screenshot above).
@hallahan , any news on this one? It's biting us quite a lot.
I'll take a look. This is a map renderer issue, and we will be replacing the legacy renderer, most likely soon.
This may be a quick fix, however.
Note to self - this would require some sort of buffer around POIs with an intersection check. If this is high priority, pinpoint where that buffer would be put in and figure out LOE (probably higher than we'd like).
Don't really know how to call this; but what seems to happen:
I'm putting random in quotes because there does seem to be some logic to it. I can't discover what logic that would be though. It's definitely not the closest one, can be miles away. But it is often the same one