Oct 21st 2005 build

[Core]

Windows builds can now run on multi-processing machines using (included) RedHat pthreads-win32.
The build has been tested on Dual Xeons (2 threads) and Dual Dual Opterons (4 threads) and
can achieve excellent performance boosts.

[Batch Language]

1). Included the ability to profile HyPhy batch code (see Examples/BatchLanguage) for usage
examples

2). Can now delete elements from Associative arrays using the '-' operator.

3). DeleteObject command introduced (currently works for likelihood functions only).

[Standard Analyses]

1). DataFileTools: SeqAlignment.bf. A simple but very fast multiple sequence alignment for
protein-coding nucleotide sequences. Sequences are checked for reading frames (with single
frame shifts detected and corrected), aligned in protein space to a reference sequence
(e.g. one of the sequneces in the input file), mapped back to codons and merged into a MSA using
the star topology. This works well for (single-time point) HIV sequences, e.g. clones from
a single patient at a given time point.

2). Compartmentalization:SlatkinMaddison.bf can now support more than 2 subpopulations.

3). A collection of speed tweaks for other analyses.

[GUI]

1). Object Inspector now displays likelihood functions along with previously supported object
types.

2). Previously posted addition of a database widget.