Azara 2.0 changes from Azara 1.0

0. Notable changes from 22 May 1995 release to 23 Nov 1996 release

0.01 process, combine : fixed reflate bug

	The output 'par' structure did not have the reflation
	(really, deflation) level initialised to zero.  On some
	computers this leads to the key-word 'deflate' or
	'reflate' appearing in the output 'par' file.

0.02 plot2, plot1, combine, unblock : checked for deflation

	These programs cannot use deflated input files, but
	this was not checked for.

0.03 plot2, plot1 : cleaned up X code

	There were lots of X compiler warnings due to loose
	function prototyping.  Also re-organised some widget
	routines.
	
0.04 plot1 : fixed polynomial baseline initialisation bug

	The initialisation for 'base_poly' had set up the actual
	processing to use the function for 'base_poly2' by mistake.

0.05 process, plot1 : fixed baseline fitting bug

	The baseline fitting was being done even if enough baseline
	points had not been found.

0.06 plot1 : fixed list initialisation bug

	The data object modules were not having the number of data
	sets being initialised, hence no data sets were being shown.
	Bug probably introduced in 0.03 above.

0.07 plot2 : fixed list initialisation bug

	The position list in the region module was not having the
	number of saved positions being initialised, hence no
	positions were being shown.
	Bug probably introduced in 0.03 above.

0.08 all : tried to make code usable for large (> 1 Gb) data sets

	The code was not designed to work on data sets of more than
	1 Gb (or so).  By changing 'int' to 'long int' in a few
	appropriate places the code should now work on larger data
	sets.  This has not been tested (not enough disk space!).
	(PS: there is no difference between 'int' and 'long int'.)

0.09 plot2 : fixed 1D display clipping bug

	The code did not clip the axes in the 1D display drawing.
	This is only relevant to the Postscipt output (the HPGL
	code does not implement clipping).

0.10 process, plot1 : added new baseline commands

	There are three new commands, 'base_points', which allows
	the user to define the baseline used in the fitting,
	and 'base_subtract' and 'base_subtract2', which allow
	the user to subtract an explicitly defined baseline.

0.11 process, plot1 : added normalised Fourier transform commands

	The existing Fourier transform commands were not normalised.
	There are now corresponding normalised (or approximately
	normalised) routines.  For every command *ft there is now
	a command *ftn.

0.12 plot2 : added baseline correction

	As well as phasing, baseline correction of the selected
	(1-dimensional) slice is now allowed.

0.13 plot2, plot1 : modified objects modules

	The default property module and data set property module
	have been modified so that multiple (instead of single)
	selections can be made, in order to make the selection
	process easier.

0.14 plot2 : display modules made iconifiable

	The 1D and 2D display modules have been modified so that
	they are iconifiable separately from the main application.
	The tracking display module has not been modified, and is
	not separately iconifiable under the MWM window manager.  

0.15 plot1 : display module made iconifiable

	The display module has been modified so that it is
	iconifiable separately from the main application.

0.16 process : added interlace command

	A new (effectively multi-dimensional) command has been
	added which allows the unwrapping of interlaced
	experiments.  The syntax is

		interlace <dim>

0.17 plot2, plot1 : allow exit when application iconified

	The application could not exit when it was iconfied
	(at least with the MWM window manager) because the
	exit question dialogue box would not pop up.

0.18 plot1 : added referencing

	Referencing has been added by assuming that the correct
	reference is that given by the most recent read_par
	command.  Alternatively, an explicit referencing can be
	set with the command

		set_reference <sw> <sf> <refppm> <refpt>

0.19 plot2, plot1 : time and date in output

	The time and date are now printed in the lower righthand
	corner of Postscript and HPGL output.

0.20 plot2 : fixed display destruction bug

	The iconifiable displays are destroyed when the display is
	closed/quit.  The code that destroyed these widgets forgot
	to unregister them from the list of all widgets in the
	application.  This list is used for cursor manipulation,
	and when the cursor was changed in the timer popup the
	program tried to access a non-existant widget.

0.21 plot2, plot1 : allow different ticks in x and y directions

	There is now allowed a different number of (minor and
	major) tick marks in the x and y directions.  If only
	one value is specified then it is assumed that the number
	of tick marks is the same.

0.22 all : work around fseek bug

	The system routine fseek (which allows a jump in a disk
	file) does not seem to work for a negative offset of more
	than 256*1024*1024 bytes.  This has been worked around by
	breaking the fseek into smaller bits.

0.23 slides : new program

	This new program allows the creation of slides using a
	selection of Postscript files.

0.24 plot2 : script files

	Script files are now allowed, which allow, for example,
	loading up of multiple planes upon start-up.

0.25 plot2_nog : new program

	This is a non-graphical (i.e. no Motif, no X) version
	of plot2, which allows output (e.g. Postscript) to be
	created just using a script file.

0.26 process : real maxent

	Maximum entropy now works on real data as well as
	complex data.

0.27 process, plot1 : added new convolution commands

	In addition to the existing function 'convolve' (now
	considered obsolote) there are now the commands 'conv_sine'
	(which is the same as 'convolve' and uses a sine function),
	'conv_box' (which uses a box function), 'conv_triangle'
	(which uses a triangle function), 'conv_gaussian' (which
	uses a gaussian function) and 'conv_file' (which uses a
	function specified in a file).

0.28 process, plot1 : added new weight command

	There is now a command 'weight_file', which uses a
	weighting function as specified in a file.

0.29 plot2 : thickness allowed to be an array

	The thickness in the extract module is now allowed to be
	an array instead of a scalar (but it can still be a scalar
	if desired).  This means that the number of planes
	extracted can be different in the dimensions orthogonal
	to the plane.

0.30 process, plot1 : modified baseline routine

	The routine that finds the baseline points now checks that
	the first and last baseline points are far enough apart
	(specifically, at least 0.5 of the number of points).

0.31 process, plot1 : added inverse cosine weighting function

	There is now a command 'inv_cosine', which has as
	1 / cosine as a weighting function.  The syntax is

		inv_cosine <frequency> <spectral width>

0.32 all : changed low-level input/output code

	The low-level input/output code has been placed in one
	module.  This has been extended to read and write
	deflated, as well as normal, data.  Most programs now
	allow deflated data as input.

0.33 plot2, plot1, slides : fixed EPS output bug

	There were a couple of bugs in the EPS output comments
	which meant that some programs would not read the files.

0.34 contours : fixed input bug

	The change in 0.32 introduced a bug in the initialisation
	of the input file if a subset of the spectrum was being
	contoured, in general (a counter was being updated which
	should not have been).

Azara help: CHANGES-2.0-0 / W. Boucher / azara@bioc.cam.ac.uk