relax SVN archive

Made the cpmg_frq() function accept only None for the reference spectrum and corrected a typo.

Corrected a few mistakes.

These were spot by Ed in a post at:
https://mail.gna.org/public/relax-devel/2009-01/msg00037.html
(Message-id: <7f080ed10901081128v5b8fe1edu7f654fb2e87d7bd1@...>)

Corrected remaining frq instances to cpmg_frq when appropriate to avoid confusion and corrected a
few related things in the system test script.

Added something in the user function cpmg_frq().

Yes, I had forgotten to put some code after the docstring..!

Added the user function cpmg_frq() and added examples to the user function cpmg_delayT().

Made a few changes so the cpmg_delayT() function now works.

Removal of the __init__() methods so that the model-structure data structure is not automatically created.

Fixed a few typos.

Added the user function cpmg_delayT() which allows setting the CPMG constant time delay T used for
the analysed dataset.

This follows a post at:
https://mail.gna.org/public/relax-devel/2009-01/msg00027.html
(Message-id: <496583C7.6050607@...>)

Fixed many formatting errors and made the 'relax_disp' code accessible (pipes, interpreter, etc).

These changes also include a coming back to using the C code 'math_fns/relax_fit.py' since there is
still no such code associated to relaxation dispersion. This will allow workingin the code without
relax crashing and complaining about the lack of a C module name 'relax_disp.py'.