> Charles,
> Somehow I think holding a write lock for minutes will
> not be a good
> thing. The UI would not be able to even read the
> model for this whole
> period of time so had better not even try. That's
> why my sense is that
> something that takes five minutes should produce no
> visible changes
> until it's done. Let the UI view an old plot until
> there is a new one
> and then switch it when its done.
>
Yes, I am planning to simply change the background color of the plot to a light red or something so the user knows it is out of sync. So I need to send a message to the view indicating that a change has occur. Likewise, when the plot is being calculated, I need a way signal the editor to disable all editing capacity (but still allow viewing).
I am little confused by your comment through--I want the plot to hold a "write" lock--I would expect the GUI could still "read" the model and display features of it.
I assume a simple mutex lock would work for this. Still, i would like to know how much of the framework I can leverage now.
I did not think to look on the command stack to see what is pending, but that is a great idea, thanks.
Charles