relax SVN archive

Merged revisions 14428-14429,14431,14433,14436-14442,14446-14447 via svnmerge from
svn+ssh://semor@.../svn/relax/1.3

........
r14428 | bugman | 2011-08-25 11:41:18 +0200 (Thu, 25 Aug 2011) | 5 lines

Deleted the STDOUT and STDERR capture mess in MF_minimise_command!

This was a horrible design breaking both the GUI and the test suite!!!
........
r14429 | bugman | 2011-08-25 11:42:03 +0200 (Thu, 25 Aug 2011) | 5 lines

The uni-processor call to capture_stdio() has been removed.

No IO redirection will be performed in uni-processor mode.
........
r14431 | bugman | 2011-08-25 11:49:34 +0200 (Thu, 25 Aug 2011) | 6 lines

Removed the IO redirection from the uni-processor return_object() method.

IO redirection in the 'multi' package is a bad, bad idea and destroys all IO redirection any Python
program is attempting to do.
........
r14433 | bugman | 2011-08-25 11:55:56 +0200 (Thu, 25 Aug 2011) | 6 lines

Eliminated the restore_stdio() method from the multi package.

The redirection of IO streams is totally incompatible with an IO redirection performed by the
underlying Python program.
........
r14436 | bugman | 2011-08-25 12:12:25 +0200 (Thu, 25 Aug 2011) | 5 lines

Fix for the RelaxError system - the status singleton was not being imported.

This was a result of a faulty merger of the gui_testing branch.
........
r14437 | bugman | 2011-08-25 12:13:07 +0200 (Thu, 25 Aug 2011) | 5 lines

Removed the _sys_exit module variable from the mpi4py processor.

This is unnecessary.
........
r14438 | bugman | 2011-08-25 12:18:40 +0200 (Thu, 25 Aug 2011) | 3 lines

Removed a pile of unused code from the mpi4py processor module.
........
r14439 | bugman | 2011-08-25 12:19:59 +0200 (Thu, 25 Aug 2011) | 6 lines

Reverted r14438 as the code actually was used.

The command used was:
svn merge -r14438:14437 .
........
r14440 | bugman | 2011-08-25 12:22:17 +0200 (Thu, 25 Aug 2011) | 3 lines

Some cleaning up of the mpi4py processor module.
........
r14441 | bugman | 2011-08-25 12:24:18 +0200 (Thu, 25 Aug 2011) | 6 lines

Reverted r14437 as _sys_exit is necessary for the mpi4py processor module.

The command used was:
svn merge -r14437:14436 .
........
r14442 | bugman | 2011-08-25 12:26:31 +0200 (Thu, 25 Aug 2011) | 3 lines

Fix for the pymol.write user function - the file name can be set to None!
........
r14446 | bugman | 2011-08-25 12:53:45 +0200 (Thu, 25 Aug 2011) | 3 lines

Fix for the GUI test mode - the mode variable should be a class variable.
........
r14447 | bugman | 2011-08-25 12:55:11 +0200 (Thu, 25 Aug 2011) | 5 lines

Some more fixes for the 'mode' class variable.

These problems were due to the multi-processor code merger.
........

Added missing delays in the list.

The duplicates had been omitted...

Fixed a bug in the inversion-recovery test whereas volumes were asked, but height were returned.

Some more fixes for the 'mode' class variable.

These problems were due to the multi-processor code merger.

Fix for the GUI test mode - the mode variable should be a class variable.

Renamed the inversion-recovery test data files to match the style of the exponential decay files.

Corrected the placement of the deselect.read() function.

The system tests below were failing otherwise.
* Relax_fit.test_curve_fitting_height_exp_2param_neg
* Relax_fit.test_curve_fitting_volume_exp_2param_neg

Is this a bug ?

Prepared further the test for the inversion-recovery (exp_3param_inv_neg) equation.

Fix for the pymol.write user function - the file name can be set to None!

Reverted r14437 as _sys_exit is necessary for the mpi4py processor module.

The command used was:
svn merge -r14437:14436 .