relax SVN archive

Fix for bug #7277 (https://gna.org/bugs/?7277).

This bug was reported by Alex Hansen. The problem was that the
'self.relax.generic_fns.palmer.create()' function was skipping spins when no relaxation data was
present whereas the complementary funtion 'self.relax.generic_fns.palmer.extract()' was not skipping
these spins. Hence the second of these functions was expecting more spins than were in the
modelfree4 generated 'mfout' file.

Created the JOBS file. This file describes the roles of the relax manager positions.

Ported r2600 from the 1.2 line.

The command used was:
svn merge -r2599:2600 svn+ssh://bugman@.../svn/relax/1.2

This fixes bug #7297 (https://gna.org/bugs/index.php?7297).

Ported r2599 from the 1.2 line.

The command used was:
$ svn merge -r2598:2599 svn+ssh://bugman@.../svn/relax/1.2

This changes 'Sconstruct' to 'SCons' in the relax manual.

Splitting of the 'pdb()' user function into 'pdb.read()' and 'pdb.vectors()' (task #3838).

These changes complete task #3838 (https://gna.org/task/?3838). The details of the changes were
discussed at https://mail.gna.org/public/relax-users/2006-10/msg00002.html (Message-id:
<7f080ed10610032153o4571935bq138d2739f55ca3bd@...>).

The only difference between the post proposing the changes and the actual changes is that the
proposed user function 'pdb.set_vectors()' is actually now called 'pdb.vectors()'.

Fix for bug #7297 (https://gna.org/bugs/index.php?7297).

This bug was reported by Alex Hansen. The problem was that relax was being too strict with the
checking of the Da and Dr diffusion tensor parameters. I have added a margin of error to these
checks to allow machine precision errors and optimisation precision issues (for the Method of
Multipliers constraint algorithm which allows the parameters outside of the limits).

Changed all references to Sconstruct in the manual to SCons. The program is really called SCons!

Completion of task #3630 (https://gna.org/task/?3630).

The unmodified Hessian and turning the line search off are no longer both specified by the string
'None'. Now they are specified by the regular expressions '^[Nn]o [Ll]ine [Ss]earch$' and
'^[Nn]o [Hh]essian [Mm]od' respecively.

This change implements half of task #3630 (https://gna.org/task/?3630).

The docstring in the generic optimisation function has been modified to clear up the ambiguity cased
by supplying the option 'None' together with Newton optimisation.

Ported the changes of r2595 of the 1.2 line into the 1.3 line.

The command used was:
$ svn merge -r2594:2595 svn+ssh://bugman@.../svn/relax/1.2

This changes the CSA value in the sample scripts to -172 ppm.