Allow the timegraph widget to be used as a callstack.
The main difference with existing timegraphs is that the list of entries on the left side may change according to the event selection, and not only the visible time range as is currently the case. We'll have to determine if we need a completely new widget type, or if this functionality can be tacked on the existing timegraph widget.
This will require work at both the Scope and Jabberwocky levels.
When defining a time graph, the "left side" should not change according to user selection. It should be immutable, once the project is analyzed and indexed. This will be much simpler, both at the programming level and at the user-experience level.
For callstacks, this means the tree entries would become a simple list of "Depth 1", "Depth 2", "Depth 3", etc. entries. These values will be valid for all positions in the timegraph.
If we want to show the "callstack", i.e. the actual vertical list of function names at a given timestamp, then we should use a separate view. If the timegraph has labels showing those names, do we really need a separate view though?
Allow the timegraph widget to be used as a callstack.
The main difference with existing timegraphs is that the list of entries on the left side may change according to the event selection, and not only the visible time range as is currently the case. We'll have to determine if we need a completely new widget type, or if this functionality can be tacked on the existing timegraph widget.
This will require work at both the Scope and Jabberwocky levels.