Sync dev branch to master

…rted-gpx

PB-492: Add description to exported GPX file.

PB-666: Upgrade to OpenLayers v10.

supporting API3, WMS and WMTS base URL overrides.

Changing calls to functions using these base URLs from the config.js so that they now prioritize the debug value if defined.

named like they were in the mf-geoadmin3's code

it otherwise creates a circular dependency and some part of the tests can't load properly (LV95 is imported from config.js, which requires something in utils.js, which requires something in config.js, etc...)

Trying to gather common config together in a meaningfully named config file.

Adding function to get base URLs, instead of direct access to constants. So that we have the opportunity to override base URLs. I didn't want to go through the store to do that. Doing so would create a lot of circular dependencies between the store, the config and components.
As it's a debug tool, that's "good enough" this way.

moving slash enforcing function back to where other base URLs are in the config

and only export/share function to access them.

Also clearing the layers config cache whenever a base URL changes. Changing the API3 URL would not trigger a new layer config load without that.

PB-871 : add debug UI and URL params to override URLs used by data integration team - #minor

…nt size.

PB-807: Do not use conflict resolution in text printing and adjust font size.

PB-735: Replace dl=0 to dl=1 for dropbox direct download

There's no RM version of the page, so we fall back to DE for that link

…r-link

PB 898 Fix rumantsch footer link to geo.admin.ch - #patch

Make it more concise in order to use it also on smaller screen. For device
orientation we will need a new button, so we need to make some room to avoid
having the button off screen.

Also removed useless 3d debugs, they are now in the regular menu.

The initial geolocation might be invalid which creates a crash error, so we avoid it.

Not all devices uses the same orientation events, therefore added both.

Due to the moveend event that set the center when the user moved the
map by pressing/touching the map and a geolocation arrived during the move operation
the disabled tracking did not work as expected and the map was re-centered
canceling the move action of the user.

So now we are listening to the pointerdrag that happens as soon as we start to
drag and not when dragging is finished. Also moveend happened not only when the
using moved the map, but has soon as the map moved even programmatically.

When the goelocation tracking is disabled a dot in the icon is added in order
to provide a user feedback on how to re-center the map

This improved reactivity and performance

We had some case with infinite loop due to heading wrapping in 3D where
cesium reported heading 360 which was wrapped to 0 and triggered a dispatch
and a fly to which reported then back 360....

This kind of annoying especially if you want to recenter but keep your zoom.

Re-centering on position in 3D is more complex as in 2d because the of the camera.
If the pitch is not 90° then we need to compute the camera position in order
to have the position centered on the screen. For this moment we don't do this
computation and simply don't re-center the view on geolocation by disabling
geolocation tracking in 3D.

and move compass feedback to geolocation button component, linking them together visually on the UI

Keep the compass on in auto-rotation even if the map is north.

Auto rotation rotate the map using the geolocation as center and it doesn't make
sense to rotate if the location is not centered (this is also the same behavior
 as on the swisstopo app)

…le to recenter

…ation

PB-154: Geolocation device orientation (gyroscope) - #minor

we were using values from epsg.io website, but our backends use different values. I found out a project in our geoadmin organization that contained all the required values, so I copied them from there.

This should fix the delta (on the Y axis) that we were experiencing when showing WMS/WMTS data with OpenLayers while showing vector tiles under

…acement

PB-131 : fix WebMercator extent/bounds

KML are well able to set that themselves if they require it, and some use cases need this to be false (flight tracks)

…r-3d-kml

PB-865 : remove forced clampToGround for KML polyline in 3D

it was dispatching any time any URL param changed, leading to the part of the code reloading the topic/layers config being called way too often

When switching to 3D, some undefined values were set, leading to a lot of bugs (and no background tiles being added in 3D)

PB-871 : only dispatch if something has changed

Update develop with Master