Large bug clean out, regressions, and a new version of relax. -- July 19, 2006 - 10:38

Hi,

I've finally put my thesis on a temporary hold and have cleared out a
large number of bugs in relax.  A number of new bugs surfaced while
fixing others already in the bug tracker
(https://gna.org/bugs/?group=relax).  The fixed bugs include:

Bug #5746 (https://gna.org/bugs/?func=detailitem&item_id=5746)
Bug #5806 (https://gna.org/bugs/?func=detailitem&item_id=5806)
Bug #6288 (https://gna.org/bugs/?func=detailitem&item_id=6288)
Bug #6383 (https://gna.org/bugs/?func=detailitem&item_id=6383)
Bug #6384 (https://gna.org/bugs/?func=detailitem&item_id=6384)
Bug #6388 (https://gna.org/bugs/?func=detailitem&item_id=6388)
Bug #6389 (https://gna.org/bugs/?func=detailitem&item_id=6389)

Some of these bugs required disruptive changes to the source code and
regressions (bugs in previously working code) may now appear.  The
relax test-suite now passes however as the number of tests are small
many regressions will not be caught.  One of the big changes is with
the 'local tm' parameter.  Previously both the global correlation time
'tm' and the 'local tm' parameters were both referred to as 'tm'.
This caused problems so now they are respectively referred to as 'tm'
and 'local_tm'.  The other change was within the internal model-free
model_statistics() function.  Hence regressions may occur through the
use of the local tm parameter or with model selection.

A bug fix release of relax (1.2.5) will soon be available and will
only be announced on the relax-announce mailing list
(https://mail.gna.org/listinfo/relax-announce/).  The 1.2 line in the
subversion repository also contains all the bug fixes.

Edward