Question: How do I flush out the data memory build up?

Can someone explain to me how I flush the memory build up that occurs as Maple keeps track of historical execution?

If I, for example, write a FOR loop that iterates a huge amount of times and don't include the colon after "end do", but rather use the semicolon, the memory usage keeps creeping up and then finally Maple just goes off into never land and never comes back.

I tried this in Windows and Mac OS X.  Same results.

I thought I could add an interface(historysize=?) statement to limit the recording length, but the interface command does not accept historysize as a option.  It does accept other ones though.

While using the colon solves the immediate problem, I can see a future problem building up because even small historical records just keep adding up a few kbytes at a time till Maple loops off into infinity at about 500Kb.  Then the only way to recover is kill the job, restart Maple, and work some more till it happens again.