Iván Sánchez Ortega

> have you noticed this glitch? @mourner Nope, it's the first time I notice this. Right now I'm suspecting it's due to a `L.Renderer` calling `_update` in between a CSS...

> Also, it's reprodicible on Safari, but not Firefox I can reproduce it on Firefox 98 on Linux. It's fiddly, though, and feels like a race condition somewhere.

It's always nice to find some weird piece of code to armor against an obscure edge case, asking oneself "but why did we do this?"... and then being able to...

I feel a tiny bit slightly better setting the default to `strict-origin-when-cross-origin` for **all** tile domains instead of just for OSM. After all, as we discussed at #9883, this *should*...

> I think @IvanSanchez is asking about the altitude of the point the camera is looking at. Yes, yes, yes, yes. Sorry if I was unclear in my first explanation....

@Firefishy I'm going over https://developer.mozilla.org/en-US/docs/Web/API/HTMLImageElement/referrerPolicy and https://w3c.github.io/webappsec-referrer-policy/#referrer-policy , and according to those, all browsers should be already using a default value of `strict-origin-when-cross-origin`. So for the default use case this...

@Firefishy By any chance, do you have a drill-down of tile requests with no referrer set, split by HTTP vs HTTPS? I'm wondering if HTTPS sites are using HTTP tiles,...

> I don't think setting an "unsecure" Referrer Policy would be a good idea. @Firefishy I'm not thinking about that. I'm thinking that, if the user is doing `L.TileLayer('http://tile.osm.org/....')`, the...

> Currently approximately 10% of tile.openstreetmap.org's tile traffic is from sites which are not sending a HTTP Referer. I can not believe that ~10% of Leaflet deployments out there are...

@Firefishy Apparently the behaviour depends on the browser. Firefox 141 drops the header, but Chromium 140 keeps it.