Fine-grained control over debug output

[hannobraun]

Right now, there exists an option to enable debug output from the triangulation algorithm. This has been helpful again and again in debugging issues, or sometimes just understanding what's happening with a model. I expect that this concept will be extended to other algorithms in the future, as required.

However, the debug output can already be pretty busy, especially if the model is not very simple, and adding more types of debug output will only make that worse. It should be possible to control the debug output in two ways:

• By type of debug output. It should be possible to only enable debug output from specific algorithms.
• By part of the model. It should be possible to enable debug output for the whole model, part of the model, or even just a single operation.

The way I imagine this happening is the following:

• Once a model tree (see Display model structure #13) exists, each entry should have a drop-down menu for debug options.
• The expanded drop-down menu can be a scrollable list of types of debug output, each with a check box.
• By default, the check box will inherit the value from the parent in the model tree. For the model as a whole, debug info will be off by default.
• It is possible to override the value for each node in the tree.

This is blocked on #13.

[hannobraun]

This issue is no longer applicable. See A New Direction for context.