v0.9.0-rc1, Dec 29, 2014.
 * many fixes including background extraction, compositing alignment, rejection
   algorithms, wavelets
 * threading of the heavy computations, to avoid graphical user interface freezing
   and provide a nice way of seeing what is happening in the console window
 * image rotation with any angle (not in registration yet)
 * Canon banding removing tool
 * GTK+ minimal version is now 3.6

v0.9.0-beta (new version), Nov 11, 2014: major refactoring. See more details at
http://free-astro.vinvin.tf/index.php/Siril:0.9.0_beta
 * new development team and website
 * many new image formats supported for import (converted to Siril's FITS: 8-bit
   and 16-bit BMP, TIFF, JPEG, PNG, NetPBM binary images, PIC (IRIS), RAW DSLR
   images and loaded natively: SER files and any film) and export
 * new demosaicing tool with 4 algoritms (AHD, VNG, Bilinear, Super Pixel)
 * better image sequence handling, with non-contiguous sequences, but still
   requiring file names to be postfixed by integers (like: image001.fit and not
   image001L.fit)
 * new graphical user interface based on GTK+ version 3.4 and above, and in
   general, update of dependencies to latest versions
 * new display modes added to the standard linear scaling with lo/hi bounds: log,
   asinh, sqr, sqrt, histeq, negative, false colors
 * manual translation as new registration method with two preview renderings of
   the current image with reference frame in transparency
 * automatic translation as new registration method for deep-sky images, based on
   the FWHM/PSF analysis of one star instead of DFT, more suited for planetary and
   nebula images
 * new commands available for the command line, see the list of available commands
 * a star finding algorithm with PSF information
 * new processing functions: unsharp mask, background extraction, green dominance
   removal, new method for background equalization, two-way Fourier transform used
   to remove annoying patterns in images, kappa-sigma stacking
 * new image compositing tool, allowing colour images to be created from multiple
   one-channel images by affecting user-defined colours
 * numerous bugs fixed and internal code refactoring for a better extensibility

v0.8
 * last release of the old Siril, in 2006.

v0.6
 * a intermediate release before a first beta release
	a lot of changes both in the GUI and the inner structure.
	48 bits RGB fits processing, bmp conversion

a sample session is given here :
	given a directory with a 24 bit bmp series 1.bmp... n.bmp,
	a sample session producing a registered 48bit fit stack
	of those is the following :
		change dir (go to the desired directory)
		file conversion: (defaults are ok for this example)
			->go 
		sequence processing (siril detetcs and builds the image sequence)
		select a region in one of the R,G,B layers 
			(registration will allways be computed on G layer)
		click register
		click stack
	you are done.

v0.5 
 * first light of the sequence graphic interface

v0.3 
 * 20050820; a few corrections.

v0.2
 * siril now has wavelet and wrecons
