MichaelBuessemeyer

>Should @markbader review this, maybe? Yeah maybe. to me it looks like a state for a first review :)

> Using the term "remote" would be more clear as that terminology is used in other parts of the API as well (i.e. remote datasets). What do you and @markbader...

> I'm not sure whether I fully understood, but are you saying the method you introduced add_remote/foreign_layer can be used to add a layer from the local file system to...

> To do so, I still find the term add_remote_layer/mag well aligned with the current mental model. In this specific scenario, I agree with you. But I think the add...

> @MichaelBuessemeyer @markbader Thank you for your work on this! I just saw that webknossos.utils includes an is_fs_path method. Could you clarify whether that's the exact opposite of the is_remote_path...

> But I think we only use the few implementations that are checked in is_fs_path and is_remote_path. So I'll give it a try to define is_remote_path as the bool'sche negation...

> I also think, this might be related to vcr-py. All the more reason to get rid of it. Can we skip these tests for now and move on with...

Now `FAILED tests/dataset/test_add_layer_from_images.py::test_test_images[8909407-embedded_NCI_mono_matrigelcollagen_docetaxel_day10_sample10.czi-kwargs4-uint16-1-size4]` fails. I tested locally and it works :shrug: I'll give it a retry and if it doesn't work I'll have to disable this as well I guess...

Notes from my talk with @normanrz: - The main motivation is that datasets are no longer referenced by their name but by their id. This enables to have multiple datasets...

Ok I think I know whats going on: This is independent of the hdf5 mapping and the reuploading. Each volume layer has a bounding box attached to it. That's how...