-
Notifications
You must be signed in to change notification settings - Fork 15
Drop 2D map view #390
Comments
I would vote for relegating the map view to a viewer only. I believe that the ability to view the data in different projections is an important feature, and as far as I tried playing around with plot_map operation during the IRM6, many projections didn't work as expected. We have to be careful to not introduce more work by dropping the mapview completely. |
Many OpenLayers projections I've tried so far, especially pole projections, are very buggy (at least in Cate Desktop) and can cause the App to become unresponsive. Even keeping display is sync is extra work as you can see from the source code and we haven't even really touched yet vector data layers. I will leave the code in as it was already a lot of work. But since I have the ok from Ed and Rainer, I like to remove the feature from the 1.0 release. Who knows, maybe someone meanwhile develops an integrated Cesium/OpenLayers component, ideally for ReactJS. |
Of course. I just provided my input, but you know best the amount of hassle required to even keep the map view up to date, so ultimately it's your decision! And 'it took a lot of work' is not a great reason for leaving code in :) |
I fully agree that the projections are an advantage, but remember, Cate is not a GIS and the 2D map view is just a view. Regarding projections, I guess it is much more important to users if they can
|
Interesting discussion and thank you for invitation to comment. Is there a way this discussion can extend to the Champion Users also, or perhaps that's already the case anyway? All the best, Ed |
@esacci Hi Ed, yes, it is already the case |
Just commented out GUI code that renders the switch between 2D/3D. |
I think the 2D map view is the one that scientists are used to use. I woudl give it preference over the 3D globe. |
@barsten This is a valid argument. But having the non-scientists in mind I would stick to 3D. The 2D is still available within the "plot_map"-operation (in a way). |
@barsten You can still go from 3D to planar 2D (geographic lat/lon "projection ") and 2.5D perspective by clicking the upper right corner of the World View, the globe icon button. We may be able to add mercator projection for the 2D view later. The feature that shall be dropped here is a 2D map view (using OpenLayers) that uses a given projection / CRS which b.t.w. as often buggy and caused the display to hang. In addition, from a technical point of view, the 3D Cesium globe offers us a lot more visualisation options for the future (e.g. point cloud visualisation, terrain, stacks of atmospheric layers). EDIT: From both Frank and Rainer (note, both scientists!) we got the ok already. The argument was, that there is no urgent need to use a particular (EPSG) projection while looking at the data. For creating maps, we can still use |
About to remove dependency to OpenLayers and proj4 and to remove modules that use it: MapView, OpenLayersMap, ProjectionDialog, ProjectionField, projection-util. Here is the tag that still contains all the code: https://github.com/CCI-Tools/cate-desktop/releases/tag/1.1.0.dev1-with-ol |
For Cate Desktop, it is becoming too much work to implement and maintain a harmonized abstraction API on top of 2D map (OpenLayers API) and 3D globe (Cesium API) so that GUI actions perform in a similar way.
As suggested and agreed upon at IRM6 we'd like to drop the 2D map view from Cate Desktop so that users will no longer be able to switch from default 3D globe to 2D map in the VIEWS panel.
The 2D map offered the possibility to set the projection which is in particular useful for polar data products. However, this can still be achieved by the
plot_map()
operation.For the future, note that also Cesium supports WMS and WMTS imagery sources as OpenLayers does. For Cate, the most beneficial OpenLayers features are the build-in GIS tools that also allow for editing points, lines, polygons. We'll have to search for a Cesium equivalent.
Specifications
Cate 0.9.0.dev7
The text was updated successfully, but these errors were encountered: