In the Qt4 reimplementation of the RiverWare Simulation Workspace, there was a change in the automatic scaling of the In-View Locator Mode and the Locator View window.
In these views, the workspace is now scaled to approximately fill the visible area with the extent of the existing Simulation Object icons.
This was a change from the Qt3 implementation which instead scaled the workspace to fill the visible area with the configured canvas size (width and height, in pixels) -- which can be quite a bit larger than the extent of the existing Simulation Object icons.
The new behavior works well for models in which the locator views are most useful -- large models -- in terms of being able to locate a particular geometric region of the workspace. The larger zoom factor (based on existing objects rather than the configured canvas size) more effectively shows the areas of interest.
However ...
... this is confusing for small models which already fit within -- or mostly fit within -- the actual size of the window.
In the screenshot to the right, the "draggable rectangle" covers the entire visible area ... compare with the prior image.
In the old Qt3 implementation, this was much less likely to occur, since the configured canvas size (the bases for the automatic-zoomed scale in the locator views) is typically significantly larger. (The default canvas size is 6450 x 6450 pixels).
We could consider ...
... adding a Checkbox, or a pair of Radio Buttons to switch between these two automatic-zoom modes, and default to the zoom mode most suitable for the current model content.
... or we could consider restoring the old Qt3 locator view automatic-zoom implementation (based strictly on the configured canvas size).
