-- News -=
 -------=

 0.1.13:
	Fixed another bug that could lead to a segfault. Also, started
	to write the ChangeLog in a format that is easily parsed by
	a script.

	Probably this is the last release before 0.2.0. As far as I know
	there are no grave bugs, so I can consider this release stable.
	Not to mention that I must rewrite the whole thing, this time with
	a much better design. This is the only way to get rid of the nasty
	hacks introduced in earlier releases, and to make this beast
	extensible.

 0.1.12:
	Fixed a segfault bug, added path resolving and minor clean-ups.
	Getting ready for a stable release...

 0.1.11:
	Thank God I didn't release 0.1.10. It was pretty broken in many
	ways. I hope I cured them all...

	This time I got rid of a bunch of bugs that could result in
	segfaults, including the pop-up-the-about-box-many-times-to-crash
	bug. Damn GTK! The widget was destroyed, but wasn't set to NULL,
	it was in a weird state, so GNOME_IS_ABOUT thought it is Ok,
	but it wasn't, so the program crashed.

	Another nice thing is the id3xml utility which can be used to
	extract ID3 tags from MP3 files, and store them in the M3X XML
	format.

 0.1.10:
 	Fixed a nasty bug, started to comment up the sources, and added
	a new database element type: object. This is not used yet, but
	will be.

 0.1.9:
	Added sounds, and fixed a typo.

 0.1.8:
	Made the main list (on the main window of the GNOME app) accept
	drops, so now if you pick a file from GMC for example, and drop
	it onto the list, m3x will try to import it.

	As usual, some bugs are fixed too, including some of the
	memory-leakers.

 0.1.7:
	I should test everything before releasing... but hey! These are
	development releases! All right, guilty feeling gone... :)

	Fixed some bugs again, added GNOME applet, and made the install
	process even more intelligent.

 0.1.6:
	A mainly bugfix release. Many bugs fixed, probably opened new ones too...

	Also I modified m3x-text and configure.in, because I want to allow
	full control over what is built and what is not. Check out the new
	options!

 0.1.5:
	Heavily reorganized the command parser of m3x-text. It is now smaller
	and faster.

	Also fixed a bug that could cause a segfault in some situations
	(try 'get /entry/' with the m3x-text in 0.1.4).

	In this version, the support for the user_data structure is implemented.
	Take a look at data/user_data.xml for more information. Or you can access
	it from the GUI by clicking on 'Change user specified data' in the 'Edit'
	menu. NOTE: The support for this in the GUI is in beta state!

	For more information, see the Changelog.

 0.1.4:
	Well, 0.1.3 wasn't free of bugs... Hacking on m3x-text revealed many
	weeknesses of the database handling routines. They worked correctly
	in sane situations, but when something goes wrong (say /entry/id1
	is a string, not a GHashTable), they'd do nasty things.

	This is now fixed.

	The other big improvement comes from a rewrite of the internal
	database handling. Now we walk through the hash table, and free
	up every bit of memory. This solves at least some of the memory
	leak problems.

 0.1.3:
	This version is the first fully working version. I think there are
	no serious bugs, except those memory leaks.	

	Also, this is the start of a quite strange program: m3x-text, a text
	mode interface.

 0.1.2:
	Rewrote the internal database handling, so the program uses less memory,
	the source is easier to read, and more memory is freed up upon object
	destruction.

 0.1.1:
	Decided not to announce m3x for a while. I'll announce when it reaches
	a stable state.

	A bunch of fixes, and a few new things in this release.

 0.1.0:
	First public release. Almost everything is working, although there's
	much to improve.

 0.0.3:
	Another unadvertised prerelease. It's good that noone downloaded the
	previous two versions yet...

	Now working:
		+ Entry operations (add/edit/remove)
		+ Database saving
		+ Database closing, and creating a new one

 0.0.2:
 	This is still an unadvertised prerelease. The UI looks fine, some
	functions appear to work.

	I need to work on the API, because the current one doesn't handle
	the 'Edit' request, it adds the new element in all circumstances.

	Remove shouldn't be a big problem...
