Improve workflow with large number of label sources

[igorpisarev]

The current workflow is not optimized for a large number of label sources. In the use-case where label sources are binary masks of primitive type (meaning that they have only 0 or 1 values), there are the following issues:

• When importing many primitive-type sources (for example, via cmd option --entire-container), for each of them Paintera shows a pop up saying that sparseness of objects will not be utilized since the data is of primitive type and there is no label-to-block mapping. The user has to click OK as many times as there are sources. It would be nice to have a checkbox "Do not show again".
• When the sources are added to the project, all objects have the same color, since the label IDs are the same in all sources. To assign different colors to them, it would be helpful to be able to automatically set the random seed value for each source separately.
• To visualize all objects in 3D, the user needs to go through all sources and hit Ctrl+A for each source. It would be great to have some sort of batch source operations, that would allow to select all objects in all sources.
  • Although, in general this may be more complicated because the sources may have different types. To solve this, for instance, the "batch select all" functionality could apply to all sources of the same type as the currently active source.

[igorpisarev]

Working around the second point is easy: we can simply use a unique ID in each source, i.e. instead of each source being a 0/1 binary mask, it could be 0 or some unique integer value. The first and third points remain and would need to be addressed in Paintera.