The problem of the non-persistent PyMOL pipe was discusssed at
https://mail.gna.org/public/relax-devel/2006-10/msg00168.html (Message-id:
<7f080ed10610290135p7ebce039tda7cd622b7ca30ff@...>) and
https://mail.gna.org/public/relax-devel/2006-10/msg00169.html (Message-id:
<BA58D16C-56EF-4AC3-B415-34A87A0B7FAB@...>).
The solution came from 'DeLano Scientific'. The details can be seen in my post at
http://sourceforge.net/mailarchive/message.php?msg_id=37186991 and the responses within the thread
located at http://sourceforge.net/mailarchive/forum.php?thread_id=30844748&forum_id=60.
