2003-09-08 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Complete reqork of perspective tilting and scroll based tilt for 
	3D and spike modes.  Now it looks much better and works well in nearly
	any orientation. (though most tilt disappears around a vertical 
	orientation, but is better than previous renditions).  Added arbritary
	zooming for high res modes (only at the moment other modes will have
	this feature in a future release). 
2003-08-18 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Some work on land3d.c to alter the tilt algorithms.  Still not perfect
	but fixes some previous issues.  The implementation is difficult since 
	we're not working in true 3D. (OpenGL). Added smooth3D button into
	the options to get rid of the black grid in Land-3D mode.  Looks cool
	with some audio sources...  Released 1.8.11
2003-06-14 Dave J. Andruczyk <djandruczyk@yahoo.com>
	More cleanups, released version 1.8.06 from Sourceforge...

2003-05-08 Dave J. Andruczyk <djandruczyk@yahoo.com>
	A lot of cleaning.  removed dead functions/variables. Started
	moving a lot of the old cruft from globals, to their respective files.
	Removed the backing pixmap option. (not worth it anymore..) Fixed
	several small bugs with the markers,axis placements, consolidated a lot
	of legacy code to streamline things making it easier to work on...

2003-04-20 Dave J. Andruczyk <djandruczyk@yahoo.com>
	More work in audio_processing.c  split the audio copy routines
	so it isn't doing things for the scope buffer when the scope ain't
	being drawn,  vice versus when running the fft's. no point in moving
	data to the scope buffers it it ain't being drawn.  Seemed to be good
	for a couple percent lower CPU usage when using the scope, and only
	about a 0.25 percent savings on the fft based displays, but I guess
	every little bit helps..

2003-04-19 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Modified the audio_processing function to update the buffer window
	with a rectangle showing the amount of data processed. This shows
	when you overstep the buffer with decimation.
	
2003-04-16 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Nearly complete rewrite again of the audio_processing functions. 
	The last incarnation was significantly slower due to a lot of 
	memcpy()'s so I incorporated the best of both the oldest and newer
	implementations, giving decimation ability and decent processor
	usage with no memcpy() calls.  Also fixed a bug in the sound.c
	input where it failed to fill the audio_ring before looping
	when the size of the fft was larger than 4096. Added decimation
	factors of 6-8 and increased the global audio ringbuffer to 2 seconds
	from the previous 1 second length.
	Known bugs:
	 - markers are still screwed when using decimation. 
	 - no over-wrap check on decimation, bounds-checking IS in place 
	and functional.
2003-04-14 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Major rewrite of the audio_processing functions to accomodate 
	decimation (used to get better low-freq resolution). Seems to be 
	fully functional except the markers on the spectragrams are 
	incorrect when using decimations.  The left-click markers on 
	the graphic EQ are also wrong as well when using decimation.  
	The scope can see much lower freqs easily and a side effect is
	it looks cooler with higher decimation values. (decimation acts 
	in a similar manner to the sweep rate control of a conventional 
	scope). Files modified: scope.c, audio_processing.c, options.c, 
	globals.h, protos.h, init.c

	Known bugs:              
        - FFT's larger than 4096 have a buffer problem, causing unwanted
	harmonics in the display that are NOT in the signal.
        - The decimator can segfault when using large FFT's if it runs out
	of bounds. (no bounds checking in place yet..)
        - The markers are all wrong when using decimation.

2003-04-08 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Rewrote a section of the land3.c file to reduce redundant calcs. 
	reduced CPU usage a swidgeon. Code cleans throughout that file, 
	fixed indentation issues, etc... Added a 0Hz border to the flying
	landforms view, that was previously missing since initial release.
	 -- later on --
	Rewrote most of scope.c for speed.  Can't believe I wrote so poorly
	last time.  Insteadof drwing things practically by the pixel, made
	use of GTK functions to handle it in bulk batches. Over 3-5 times
	faster now.  CPU usage dropped from 13% to 3% !! (scope modes only)
	NOTE: this is only forthe line/dot scope modes.  The pixmap mode
	still is inefficient and I have not found a way around it.
2001-11-17 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Fixed a very longstanding bug in the audio I/O routines.  When the 
	buffer wrapped, it had an improper offset in an memcpy routine
	leaving a variable sized hole in the display, no more than nsamp/2
	bytes wide.  (a real bitch to find too...)
	CVS commit.
2001-10-22 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Fixed time/frequency markers from clearing the screen.  Also adds
	a little breaker bar to the scrolled portion when switching freq
	axis's so as to show when the user changed the display.
	CVS commit.

2001-10-14 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Fixed numerous bugs in reducer.c so that it no longer craps out
	when bins_per_pip=1.0000.  Also fixed the bug of half data missing
	when bins_per_pip=0.5000.  Changed lag control to actually be in 
	milliseconds.  will make it easier when I add the proper lag
	compensation code later. (to allow auto-display stop when the
	audio stops.  (saves CPU, among other things..))  Changed the audio
	ring-buffer to be of a fixed size. (4 seconds of audio). Fixed
	the lag compensator s othat it adjusts properly when changing
	the FFT length, so the fft display stays in sync with the audio.
2001-10-07 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Added variable bandwidth for hi-res displays as per users' requests.
	Did some rework on the reducer algoritm and the marker functions.
	Lots of other little fixes here and there.  Too many to mention.
	Released 1.6.3
2001-09-11 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Incorporated all of Rolf Muellers Colormaps that he sent over. 
	Thanks again... So now everyone else can enjoy them too.  Began
	the workings of the new colormap file format.  eXtace currently
	understands both formats (the new format is not complete yet)
	Don't try to use the new format yet, until I get it written. (unless
	you write it before I get a chance to... )
	Released 1.6.1

2001-09-01 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Spelling fixes (Rolf Mueller). Colormap suggestions for better 
	display ability..
	Released 
2001-08-24 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Major cleanups in the code.  Trying to remove the redundnat
	functions to make the code cleaner and tighter. Added two new header
	files to clean up the DEF's used in the code..  Fixed some minor
	bugs in the display code (And prolly added some new ones).
	Reorganized to options panel to make more sense and look much
	cleaner. Various little bugfixes and code reorgs.  Options controls
	now in a seperate files.  modularized some of the event handlers to
	remove the excess code.  Added the buffer latency monitor to
	visualize how the data comes in and is processed.  should make it
	easier to adjust the Lag factor for visual sync to the audio being
	heard.
	
	eXtace will warn about options being out of date and will blindly
	delete them, but it will warn you. It is better off to delete them
	and spend a couple minutes resetting the controls to get used to the
	different layout.  You have been WARNED!
	Released 1.6.0

2001-05-17 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Rewrote colormap generator section to use pointers instead of 
	static declarations.  Fixed a lot of little things here and there.
	Too many to list here..
	Released 1.5.0
2001-01-20 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Code cleanups, added new header with defauts for options easily set.
	Released 1.4.6
2001-01-01 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Minor code cleanups here and there. 
	Released 1.4.5
2000-12-14 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Fixed the segfault when changing sample widths. Added an auto-lag
	adjust that should eliminate the meed to modify the lag when 
	changing fft widths.  
	Released 1.4.4
2000-12-11 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Fixed some bugs in the convolve code with segfaults on small FFT
	sizes.  Fixed the lag control. (It was backwards). Now readjusting
	the lag when changing fft width is not necessary. convolve code
	disabled on fft sizes of 1024 points and smaller, due to buffer
	overruns, making the convolve not worth it for the small window.
	Use 2048 or larger fft widths for convolve to work normally
	Released 1.4.3 
2000-12-10 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Fixed a major buffering problem. The harmonics should be much cleaner
	now, as the previous version was chewing the sound input all to hell
	the blips in the scope will also be gone..
	Released 1.4.2
2000-12-07 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Removed mixdown function, and combined its functionality into GetFFT
	to help improve performance and make the code easier to read. 
	Fixed some severe memory leaks, (And probably added some more...)
	Found that using gtk_timeouts to give the variable frame rate is
	CPU costly. Know of any other high-resolution timer on the system?
	(need 1 millisecond resolution MINIMUM, versus the 10 msec or the 
	kernel jiffies timer)
	Released 1.4.1
2000-11-29 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Complete Sound I/O rewritten from scratch. Now fully multi-threaded
	in both Esound, and ALSA. (thanks to tips/code from ecasound).
	The display now runs as a seperate thread at a constant speed.
	The fft size is now configurable (via config file only so far).
	Tested it works properly up to 32768 point FFT's (though it eats my
	processor completely) and the display is rather like slow motion,
	due to the releative lack of data to makea 32768 point fft run at 
	20 frames per sec, so data is somewhat reused to allow for smooth
	screen motion, instead of the jerky motion of other large point
	fft's.
	Release 1.4.0
2000-11-04 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Fixed Colormap code somewhat. Made eXtace auto-generate 5 
	colormaps on first startup... Other misc little fixes here and
	there...
	Released 1.3.11
2000-10-31 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Between funerals, sick cats and the project managers from hell, 
	I've managed another release. This is a simple twist (literally)
	on the good ol' Spectragram, a new mode has been added, Vertical
	spectrogram, which includes a hi-res 2D spike FFT with axis markers
	Performance should be as fast as any other modes. The frequency
	marker bar can be dragged to almost any position (within 125 pixels
	from the bottom to make sure there's space for the 2D display.
	Released 1.3.10
2000-10-16 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Fixed various coding bugs on the 2D Graphic EQ Analyzer.  Added
	clickable markers to see what pip is what frequency in the 2D EQ. 
	Fixed the ALSA detection so that it actually works properly. Added 
	RPM spec file from Alexander Skwar. (Thanks again, Alexander...)
	Released 1.3.9
2000-10-11 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Cleanups in scope.c. Removed a whole lotta redundant code. Dropped
	cpu usages by a whopping 2% (woohoo!). Much easier to read anyway,
	so it was worth the effort.
2000-10-07 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Imprevements to 2D EQ scaler. Markers now auto-scale resonably.
	Looking for suggestions to make the markers look "right"...
	Thread fixes,  (still a deadlock in the sound-switching functions)
2000-10-02 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Made 2D Graphic Eq Auto-scale much better than before.  Fixed a
	bad code problem and potential memory leak. Preparing for Axis
	markers for 2D EQ.  Added Peak and hold for trailers a-la xmms.
	Looks cool. Added slider to control panel to adjust peak hold, and 
	falloff speed.

2000-09-26 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Incorporated patch from Eric Lassauge which fixes colormap bugs,
	ALSA related bugs, and adds the version to the titlebar.
	Released 1.3.8
2000-09-22 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Fixed several threading bugs in the code. should be less likely
	to lockup on close. (something to do with gdk_threads_enter/leave)
	Broke out all drawing routines into their own files.  This makes
	it a whole lot easier to add new routines in the future.  Fixed
	colormap saving (had my own home dir hard-coded into the code, oops.)
	Tested the conditions where ifa user deletes the colormaps so 
	that eXtace doesn't freak with a scrambled colormap...
	Released 1.3.7
2000-09-21 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Added flip Y-axis button, to invert display. Fixed several math
	core dump bugs with sound source. (Thanks to Eric Lassauge for 
	the patch). 
	Released 1.3.6
2000-09-20 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Added ability to select ALSA card, device and subdevice interactively
	The changes take effect IMMEDIATELY, which I think is just plain cool
	Fixed several thread related bugs.  eXtace should no longer freeze
	up when momentarily pausing it in ALSA mode. The lockups  that
	sometimes happened on exit no longer should happen, if so let me 
	know.  There is stil the posibility of freezing when switching the 
	ALSA card/device/subdevice while its running. It is a timing issue
	somewhere, and I haven't got it yet. if you don't try to make it 
	show you'll prolly never see it. Hopefully I can have it fixed by
	the next release.
	Released 1.3.5
2000-09-16 Dave J. Andruzyk <djandruczyk@yahoo.com>
	Added ALSA loopback support.  Extace now requires pthreads, due to
	alsa's kernel callbacks.  Added choice in options panel to choose
	ALSA or ESD for the audio source.. Its getting close to the point
	where a full reqrite should be done to make it easier to thread for
	multiple displays and other cool stuff. There are the slight chances
	for race conditions or segfaults due to my complete lack of knowledge
	of using threads the Right Way...
	Released 1.3.4

2000-09-05 Dave J. Andruczyk <djandruczyk@yahoo.com>
	NEarly complete proper perspective tilting of the 3D spike axis'
	to a give a much more correct and eye-pleasing display.  The
	axis can now be flipped over properly and dragged to any direction
	with the proper tilting or the y axis with respect to the X and
	Z axes.  Still needs a little work, but much better that ever.
	Hopefully the extra trig won't hurt performance..
	Relased 1.3.3
	    
2000-08-27 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Fixed fftw tests due to change in fftw RPM compilation made in
	fftw 2.1.3.  Thanks to Alexander Skwar <ASkwar@DigitalProjects.com>
	for pointing out the changes to fftw.
	Released 1.3.2
2000-08-26 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Rewrite of 3D landform routines.  Added semi-perspective tilting.
	Works fairly well, though the math ain't right... Split off the
	main draw routine into its own file to make for easier maintenance.
	Will prolly do this to all major graphical routines as draw.c was
	over 1000 lines long. The display over-writing itself is FINALLY
	fixed. (very very long-standing bug). it should no longer draw 
	over itself corrupting the display anymore no matter what direction
	you aim it..
	Released 1.3.1
2000-07-22 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Added colormap read/write support. (Asked for feature). eXtace 
	remebers your last one you used and defaults to it when running the
	next time.  An arbritary number of colormaps may be saved and re-used.
	If you run into problems, close eXtace delete your config file;
	~/.eXtace/config adn rerun extace to rebuild theconfig file.
	Released 1.3.0
2000-06-11 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Removed GSL support as it seems to no longer be used much, and it
	has three major versions all of which act differently making it a 
	rela bitch to support.  Redid config routine to handle the various
	possible installations of fftw (double/single precision versions)
	Should make it work on other systems who have isntalled the BINARY
	release of fftw 2.1.3 properly.  The configure script also CORRECTLY
	recognizes the missing libs and reacts accordingly.
	Thoroughly tested with single/double precisions fftw libs and the
	none versioned one (normally double precision) 
	Released 1.2.25 (1.2.24 never made it to the server...)
2000-05-03 Dave J. Andruczyk <djandruczyk@yahoo.com>
	REALLY fixed the memory leak in convolve code, thanks to Ralph Loader
	(author of convolve code) for the fix.  Added backing pixmap to all
	scroling displays. It may or may not eat lots fo extra cpu with it
	turned on.  Seems to depend on how much accelleration X has for your
	vid card plus the amount of ram avail on the card (for pixmaps and 
	such)
	Released 1.2.23
2000-04-24 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Fixed the memory leak in the convolve code thanks to mike perry
	of njamd fame, and Ralph Loader who originally wrote hte convole code.
	Increased the window used to match patterns for scope stabilixer to
	triple its original size, making the scope stable down to 60hz. 
	Below that I run out of samples.. and the display will tend to want
	to scroll. (use the "sine" program to see for yourself)
	Released 1.2.22
2000-4-09 Dave J. Andruczyk <djandruczyk@yahoo.com>
	Todays my birthday, Happy Birthday to me.. :)  Added graticule to
	the scope. It auto-sizes too.  Fixed several memory leaks, Found 
	a small bug in convolve code. still working on that one.  
	Released 1.2.21 
2000-4-04 Dave J. Andruczyk <djandruczyk@yahoo.com>
	* Added trailing peaksto Graphic EQ, as per users request.  Added
	more controls to the control panel to turn the peaks on or off as 
	desired and control their decay speed.  As usual all settings are 
	stored between sessions. Added trace stabilizer control to the scope
	controls to see the difference it makes...
	Released 1.2.20
2000-3-31 Dave J. Andruczyk <djandruczyk@yahoo.com>
	* Added convolve routine into the scope from Alsaplayer designed by
	Andy La Foe.  Now the scope is a million times more stable making
	it far easier to read.  
	Released 1.2.19
2000-3-20 Dave J. Andruczyk <djandruczyk@yahoo.com>
	* Been busier than I'd like.  Found some severe code problems, which 
	have been resolved and created some new ones. (4096 point mode goes
	twice as slow now. (I was only crunchin on 1/2 the data as I was 
	supposed to.  Fixed now.)  Scope now goes to 1024 points instead of 
	512 and it will self-truncate is you switch number of points as it
	should have previously.   Fixed the hopefully LAST bug in the window
	function code, due to the above 1/2 data problem..  The window 
	functions now work better than I ever expected... (use the "sine" 
	program to see what I mean (sine |esdcat)
	
	The 3D Spike mode now has semi-perspective axis tilting, which sorta-
	works, but not quite the way I want.  If ANYONE knows howto get a 
	3D object to look write in 2D space teach me howto do it...
	Released 1.2.18
2000-2-6 Dave J. Andruczyk <djandruczyk@yahoo.com>
	* Been busy working for money....  Several minor fixes.  Moved the 
	project over to sourceforge.net (extace.sourceforge.net)  Cool site!
	Released 1.2.16
1999-12-12 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c.  Stereo scopes completed. Seems to be working fairly well, 
	though X intensive. (non-accelerated X servers may be kinda sluggish)
	Released 1.2.15 (.14 had severe bugs)
1999-12-9 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* all files. Rewrote audio gathering routine to be stereo freidnly for
	soon to be made changes. (synaethesia sort of modes coming)
	Added and internal mixdown routine for the ffts (as they still want
	mono data until I figure out how to make em faster.) Also ALSA ready
	thanks to the new mixdown routine.
	
1999-11-22 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c, globals.h main.c  Rewrote the options panel to use GTK's 
	notebook widget. (looks a LOT better). Added new scope modes. 
	Rewrote scope routine so its faster, and uses less CPU. Now there
	are three scopes, selectable via the options panel.
	Getting ready for an Alsa capable version. (won't require ESD)
	Released 1.2.13
1999-11-05 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c reducer.c  Wrote a newer far more functional reduction algo
	rithm that actually does what I intended it to.  Now you can reduce
	the number of samples in ~./eXtace/config (nsamp) to 1024 or less and
	still get a decent display. (less resolution and faster of course.)
	Added "-Wall" to CFLAGS to clean up the code.
	Released 1.2.12
1999-11-04 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* dir.c main.c	STUPID bug fixed that cause GDK bad window error
	message. (missing safety check.)
	Released 1.2.11
1999-11-04 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* dir.c main.c globals. protos.h.  Additions so that gradient
	window and direction window remeber their positions when they 
	are moved, and then disappear when changing modes..
	Released 1.2.10
1999-10-25 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c globals.h color_win.c Various bugfixes to color picker
	to make changing colors a bit easier.. Added a new mode "Wire_3D"
	which is pretty close to LAND_3D.
	Released 1.2.9
1999-10-17 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* globals.h main.c  Updated number of buffers to 12 to compensate
	for the newer esound (0.2.14) which seems to have greater lag (RH 6.1)
	Released 1.2.7 and 1.2.8 (quickie fix)
1999-10-15 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c, misc.c, main.c, globals.h. Added a time scale to the 
	spectrogram. (user suggestion.) Various other optimizations to 
	improve performance.
	Released 1.2.6
1999-10-13 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* all files.  Major rewrite of the display engine. Squashed MANY
	bugs.  Thanks again to Rolf Mueller for helping me debug the window
	function code. That part is fully functional and working properly.
	Fixed DnD (crag n drop) code bug in 3D spike mode.  Display 
	optimization code fixed. (should be much faster) even with more 
	data points in the fft (up to 4096). I noticed redhat incorporated
	eXtace into redhat 6.1 (woohoo!!). 
	Released 1.2.5

1999-10-05 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* Removed Gnome dependancies, and other unneeded cruft. Thanksa million
	to Andy Lo A Foe, for the patch, and the advice.  Other configure
	cleanups.  Should configure about 4 times faster...
	Released 1.2.4
	
1999-10-04 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* window functions acting screwy. Looking for assistance...
	Minor code improvements to spectrogram for moving the axis smoothly
	Release 1.2.3
1999-09-20 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* window functions improved slightly.  Now are user selectable on 
	the fly from the options menu.  Settings are saved between sessions.
	Found a bug/featurwith the window function where it ONLY has the 
	desired behavior when setup for 1024 points.  who the hell knows...
	the display is still cool, and accurate, so what the hell...
	Released 1.2.2
1999-09-16 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* Window Functions FIXED!!. Thanks to Rolf Mueller for the insight
	into window functions. It was mistakenly used on the freq domain
	data along with the time domain data giving a rather weird view.
	7 Window functions are avail, and will be added to a menu soon.
	
	misc.c Signifcant speedups accomplished by moving some code around.
	Now uses 10% less CPU time as prior releases.  Scope mode now uses
	50% less cpou time. (fft was being computed when it didn't need to be)
1999-09-15 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* misc.c globals.h.  Upped the number of samples per fft block to
	4096. (gives a 2048 point fft (i think :)) Need to rewrite display
	sections for spike and spectrogram to reduce redndant writes to the 
	same pixel location. (who's running wiht 2048 pixels across?) Should
	dramatically reduce CPU especially on slow machines (I hope).
	Fixed a bug in the audio buffering routine, where old data seemed
	to slip in and mess up the display. Now the buffer is zeroed before
	being refilled.
1999-09-07 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* color_win.c misc.c main.c, and others. Wrote color picker routine
	so you can change colors on the fly. (slow as hell when it has to 
	regenerate the entire colormap each time you change it). Any 
	suggestions to improve this (lousy) method are welcomed. NOTE:
	color choices are NOT savced on closing (yet). Hopefully the next
	release will address that. 
	Released 1.2.0
	
1999-08-29 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
        * dir.c, misc.c Fixed color routine so you doin't get duplicate color
       	spectrum when over 64 bands. Minor changes to direction control 
       	widget to make it less sensitive and easier to control.
	Released 1.1.16
	
1999-08-16 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* init.c globals.h events.c misc.c Changed spec_start to a float
	so it doesn;t mess up when you shrink the window. (stored as a 
	percentage instead of an pixel value)
	Released 1.1.15
1999-08-10 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* misc.c events.c. Minor fixes, the new spectrogram mode is now
	draggable to anywhere on the window. Axis markers are now standard
	on the spectrogram. Full autoscaling as usual...
	Released 1.1.14
1999-08-05 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c main.c.  Fixes for a stupid error that caused eXtace to
	segfault due to doing a "bad thing". Added a new mode called
	Spectragram. submitted by Gregor Leusch 
	<gleusch@Pool.Informatik.RWTH-Aachen.DE> 
	Released 1.1.13

1999-07-28 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* init.c Damn axis type was set to "linear" instead of log, made
	the landscape view and graphic eq look like crap..  fixed
	Released 1.1.12
1999-07-28 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* main.c globals.h, init.c:  Now window position is saved between 
	sessions. so it comes up whereever you had it last.  Thanks to the
	Alsaplayer code for that tidbit..
	Released 1.1.11
1999-07-22 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c misc.c init.c: Fixes in init.c for not saving/restoring the
	sensitivity setting.  Fixed window function routines, and added
	six more window functions from the "freq51" package.  Will add them
	to the menus in hte next release. (you can manually pick them 
	by editing the "config" file in ~/eXtace/ )
	Released 1.1.10
1999-07-19 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* configure.in events.c: Added additional debugging code with 
	regards to the DND code.  Minor fixes causing the 3D Pointed
	fft to be undraggable fixed. (typo :( )
	-- later that evening....
	main.c Fix for display coming up and not doing anything.
	cause was leaving it in the "stars mode" (about button). 
	missing function so it wouldn't restart on rerunning now fixed.
	Released 1.1.9
1999-07-16 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* configure.in Fix to accomatate missing esd-config scripts
	Released 1.1.8
1999-07-15 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* configfile.c configfile.h  Copied and hacked from the xmms
	tree. (thanks guys, saved me tons of work!!). Now extace
	auto-saves options upon close, and restores them at next run.
	Released 1.1.6
1999-07-14 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* events.c Reqrote DND routine for moving the trace around.
	Now implements locking so you can move as fast as you want.
	Also fixed  numerous  positional bugs when grabbing and
	moving the the traces. If any bugs are noticed email them
	to me ASAP. (I need the start and end points for X and Y and
	the X and Z speeds (from the control panel))
	Released 1.1.5

1999-07-13 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* dir.c buttons.c logo.xpm Redid the logo to include
	Raster and Michael Fulbright of Redhat (original authors of
	extace). minor bugfixes..
1999-07-12 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* main.c draw.c minor fixes to the decay routines to prevent
	"yellow spot syncdrome" on the 3D landscape view. Extended
	decay amount slider up to 10, (makes the Graphic eq more 
	like my home stereo). 
	buttons.c Made the direction controller disappear when its
	note needed. (i.e. when graphic eq, scope, or "about" is up)
1999-07-12 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* configure.in  MAJOR BUGFIX!!.  Now compiles PROPERLY again with
	with FFTW library. missing library check in the script.
	Released 1.1.3

1999-07-11 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	*  configure.in misc.c: Fixed broken configure tests in regards
	to the fftw and gsl libraries and headers. Got some bug reports
	regarding particular version of the gsl library.  need to do more
	tests.	Released 1.1.2
1999-07-09 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* dir.c gc_maker.c Wrote 3D direction controller.  Works
	REALLY well.  just hold the button down and drag in the 
	direction you wanna go.  Auto-udates the control panel
	as well.  A touch sensitive on the 3D spike view.
	
1999-07-09 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c main.c events.c  Fixed broken defines for gsl due to
	working in the wrong build tree :) Put stubs in place
       	for 3D directional control window.	Released 1.0.4
	
1999-07-09 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c events.c Fixed DND routine so that its more lenient
	as to how close you have to get to get a lock on it.
	removed persepctive tilt stuff as it will require a rewrite
	of the main drawing routines, due to the lack or proper trig 
	calculations. :(	Released 1.0.3
1999-07-09 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c main.c Fixe improper defines, added tests for older
	version of gsl which seemed to put the headers in a different
	location. Released 1.0.2
	
1999-07-07 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* draw.c Working on proper slant of the trace and leading edge.
	Sorta works, but proportions are way off.  Time to dig out 
	the trig books... :(  Released 1.1alpha1
	
1999-07-06 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* events.c Minor fixes in DND code.  Disbaled triangle-old-data
	remover. too many bugs in it.. :( Released 1.0.1
	
1999-07	06 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* Numerous fixes in configure scripts.  Added DND for the axis's
	Now the axis's can be dragged and placed wherever you want on the 
	window. You have to "go slow" as I haven't made it "lock onto"
	the trace while you're moving it. (so it ain't really true DND) but
	its still a cool feature... :) Released 1.0

	Configured configure.in so that the whole thing is now standalone.

1999-07-02 Dave J. Andruczyk <dave@techdev.buffalostate.edu>
	* Now the official maintainer of extace.  Broke it out of the gnome-
	media tree and made it standalone as its beginning to grow.  
	Has a self configuring "automake" system now which detects the 
	gsl, and fftw libs, and configures appropriately. Works well so far