Azara 2.0 changes

1.	Notable changes from 23 Nov 1996 release to ** *** **** release

1.01	process : added varian option

	A new command has been added which allows the processing of
	data acquired on Varian machines.  The syntax is

		varian <dimension ordering>

	where the <dimension ordering> specifies the storage order
	for the data (which can depend on the experiment).

1.02	plot2 : fixed pivot bug

	The pivot used in the (approximate) phasing of slices was one
	less than it should have been.

1.03	plot2 : added popup to script actions

	Scripts can now be used to popup individual modules.

1.04	unblock : added int and swap options

	The output can now optionally be stored as integer (using the
	'int' key word) and with a swapped byte order (using the
	'swap' key word).

1.05	plot2_nog : fixed region stats bug

	The 'region.stats' command did not work because a preliminary
	check was returning TRUE instead of FALSE.

1.06	plot2 : modified 'whole' command

	The region specified by the 'region.whole' command now goes
	from point 0.5 to n+0.5 instead of 1 to n, where n is the
	number of points of the spectrum.

1.07	plot2_nog : added region stats output option

	The statistics produced by 'region.stats' can be sent to a
	file instead of the screen by using

		region_stats_open <file>

	and the file can be closed by using

		region_stats_close

1.08	contours : made output block size 4096

	The block size for the output was the same as the input
	block size, but this has now been changed to be 4096,
	because that is what Ansig requires.

1.09	process, plot1 : added mirror options

	Complex (time) data can now be mirrored with either zero
	dwell ('mirror_zero') or half dwell ('mirror_half').

1.10	slides : fixed bounding box parsing bug

	Parsing the bounding box in the (correct) case when the
	numbers did not have parentheses was not skipping over
	the text before trying to parse the numbers.

1.11	peak_find : fixed exclude bug

	The calculation of the difference in ppm values from point
	values between two dimensions was incorrect because the
	points started from 0 rather than 1.

1.12	plot2_nog : fixed data non-registration bug

	The data sets were not being properly registered, so some
	data operations (e.g. data.all_off) were not working.

1.13	process : fixed blocking bug

	In a non-blocked data set, if the number of points in some
	dimension was greater than the block size and that dimension
	was not processed then the work space to do the blocking was
	not allocated enough space, hence a memory trample happened.

1.14	process : fixed scale2 bug

	The imaginary part of the scaling was using the real part.

1.15	process : fixed interlace bug

	The wrong index was being used to find matched rows.

1.16	process : fixed initialisation bug

	In the second and later passes of the data the input and
	output files are the same, and the data is processed from
	the beginning or end depending on whether the input or output
	is bigger.  The place where the input and output file names
	were being checked occured before the input and output file
	names were being set (in those particular variables), so the 
	check was not happening correctly.

1.17	process : added Varian functionality

	16 bit Varian data can now be processed, and the Varian
	header is now used to determine whether the data is integer
	or floating point.

1.18	process : fixed initialisation bug

	A structure being used for output was no longer being properly
	initialised for byte size because of the above Varian changes.