In order to preserve the "prevent unintentional moving" support, perhaps this would be best performed only inside of the image command. How's this for a scenerio:

You select something in the drawing and then call the image command.
The image command tests to see if anything is selected when it starts up and, if so, asks the user (in some intelligible way) if he wants to maintain those selections in order to move them along with the background.
If the user says yes then MoI does whatever is necessary to bind the selected objects to the background, but only for the duration of that particular invocation of the image command.
The user does whatever he's going to do, and when he leaves the image command, the bind is revoked.

This would seem to me to restrict the ramifications of the binding as much as possible. What do you think?



As for learning to program things for MoI, a "diging in" starting point is what I'm looking for right now. One problem I'm going to have with this is that, although I have some training in higher level languages, my entire career involved real-time assembly language programming of microprocessors. What language is MoI written in?

Thanks,
Mark