Audacious 2.1.0 (2009-07-06) ============================ Enhancements ------------ * Newui was moved into a plugin and is now called 'gtkui'. It is still not very complete, but has progressed some. * Sorting by album added. * Completely new, written from scratch alsa-ng output plugin, which fixes many of the problems our old ALSA output plugin had; including high CPU load with Intel-HDA. * Listing of available interfaces via 'audacious -i list'. * Added bs2b (Bauer stereophonic-to-binaural) headphone crossfade filter effect plugin. It is designed to prevent listener fatigue by reducing the super-stereo effect that headphones normally produce. * rootvis plugin was dropped, as it does not work on most systems anymore. * Some updated translations. Bugfixes -------- * Mouse wheel settings where removed (for now) from properties, because they didn't actually do anything. * Fixed problems with seeking that could occur under certain conditions. * Many fixes in newui/gtkui. * Fixed frenzied looping, 100% CPU when loading metadata is disabled. * Run visualization process only when a shown and playing. * Quit cleanly (and safely) on SIGINT/QUIT/TERM. * Fixed several race conditions and NULL-pointer dereferences. * New playlist scanner, which, related to above, is safer and more effective. * Fixed MPRIS interface's GetStatus, StatusChange and Quit functionality to match the official MPRIS spec. * Bunch of LADSPA, cdaudio-ng, PulseAudio fixes and cleanups. * Remove bogus and unnecessary end-of-playback logic in vorbis plugin. * Fixed NEON transport plugin error reporting. * SID-plugin was synchronized with XMMS-SID codebase, with number of enhancements and fixes. * Fixed SSE2 instruction set support checking in configure. * JACK output plugin got cleaned up and checking for JACK availability was changed to use pkg-config. * And bunch of other bugfixes, cleanups, etc. Audacious 2.0.1 (2009-05-13) ============================ Bugfixes -------- * Math dealing with the over-commitment of the equalizer values has been corrected. However, please note that the equalizer goes from literal -12dB to +12dB, which can be a very major boost. So yes, the highest settings causing distortion on really loud audio is *not a bug*. We may drop this down to 6dB in both ways in 2.1, depending on how many complaints we get. Audacious 2.0.0 (2009-05-12) ============================ This branch of Audacious represents a major restructuring of how the player itself works. Notably, the skinned UI is no longer part of core. There is an experimental 'skinless' UI which can be launched with 'audacious2 -i newui', but it is incomplete at this time. Patches accepted! Other notable enhancements include switching to GIO/GVFS from using stdio, which while raising our requirements to GLib 2.16, gives us the ability to do many things out of the box. This is a good thing. Enhancements ------------ * Everything between 1.5.0 and 1.5.1. * Lots of other stuff. * Playing streams no longer edits the config preferences. This is a behavioural trait from XMMS. * Many plugins have been added, including experimental support for PSF2. * The projectM plugin now works with libprojectM 1.2. Bugfixes -------- * The equalizer preamp setting has been fixed. * --play-pause actually works correctly now, it broke a long time ago when we started using DBus * MPRIS exported data now includes the track number * Behavioural issues while seeking when paused have been corrected. * `audtool help playback-seek-relative` no longer segfaults * The file information dialog is no longer destroyed when you close it, but is hidden instead. * Many legacy XMMS behaviours have been corrected. * Handling of non-bitmap fonts have been corrected. * Tons of others which are less notable. New bugs -------- * Probably a lot! Test and report. Audacious 1.5.0 (2008-03-13) ============================ This is the last branch of Audacious that will support in Winamp2 skins in core. We are working on a new default UI for Audacious2 which provides a more metadata rich environment (while still be lightweight and fast). However, it is planned to implement support for Winamp2 skins as a plugin in the next version. No promises, though. It has been fun being seen as "rescuing XMMS" by the community; but this has never been our goal. Our goal has always been to develop the best possible player on Linux. This means that we are audiophiles, and we want our audio to sound as nice as possible. This has been our primary goal; not maintaining a Winamp2 skinned player. When we determined that the current UI implementation was getting in our way, we decided to rip it out. As such, the future of Winamp2 skins in Audacious is undefined, but they will probably be supported via a plugin. Enhancements ------------ * Everything between 1.4.1 and 1.4.7. (duh) * Ability to disable GTK+ theme suggestions provided by skins. * Entirely from scratch audio engine (libSAD) which supports: - 16-bit audio - 24-bit audio - 32-bit audio - IEEE floating point audio - 28-bit fractional audio - many weird obscure "audiophile" formats and: - high quality dithering/conversion/resampling to produce as high quality audio as possible for your sound system. - bit-exact passthrough of decoded audio (compare to "kernel streaming" in foobar2000) As a result, support for oldschool 8-bit soundcards has been _dropped_. It's 2008, they haven't made such hardware since the early 90's. Additionally, codecs now send audio to audacious using their native format; removing the need for costly implementation-specific conversion functionality. * New unified metadata editing for most formats. Ability to see "raw metadata"; the metadata which can be used in tuplez scripts. * Removal of the XMMS audio conversion engine in it's entirety. * Support for "virtual plugins"; the ability for a single module to export up to thousands of plugins. This will be used more in Audacious2. * Many adjustments to the winamp2 skinengine. * Faster jump-to-file. * Implement support for session management. Bugfixes -------- * playlist_compare_artist(): Avoid NULL dereferences. (Bugzilla #172, Bugzilla #168, Bugzilla #35). * Tons of memory leaks fixed. Audacious 1.4.1 (2007-11-17) ============================ Enhancements ------------ * The skin code has been tweaked a little to fix loading of more substandard winamp skins. (http://boards.nenolod.net/viewtopic.php?f=5&t=859) * Playlists now carry an atomic serial number in order to avoid unnecessary reloading of metadata. * The command names used by audtool have been sanitized in order to make some level of sense. * Add many new calls to libaudclient which were unimplemented after DBus transition. * Add error reporting when Tuplez scripts fail to compile. * Use a hashtable of extension to codec candidate lists to speed up codec detection. * Make on demand codec detection use the hashtable to filter out unplayable files. * Codec detection is now done on demand by default. Due to the recent rework of the codec detection system, this is the recommended setting. (Audacious can process a large directory in only a few seconds of blocking this way, and streams are entirely processed in background this way.) * Add codec and quality fields to MPRIS TrackInfo tuples sent over DBus. Bugfixes -------- * [Bugzilla #53]: UTF-8 validation gives false positives (Reported by Kosma Moczek; Fixed by Yoshiki Yazawa) * [Bugzilla #20]: status icon plugin breaks the 'always on top' feature (Reported by Dorian Scholz; Fixed by Tomasz Moń) * playback_get_length() was returning the typecode of FIELD_LENGTH, not it's value * audctrl.h was dependent on #ifdef __cplusplus being a valid macro, now we use G_BEGIN_DECLS/G_END_DECLS to handle that. * Fix doublesize issues on startup. (Gentoo #199071) * Fix possible glitches that could be triggered by a scrolling textbox, when the skin is switched. * Fix improper handling of integer constants in the Tuplez compiler/evaluation engine. * Fix misleading assertion notice triggers when a URI fails to load. * Cope with Winamp2 skins which do not have dashes in numbers.bmp. Audacious 1.4.0 (2007-11-01) ============================ * Everything ;) Audacious 1.3.0 (2007-03-02) ============================ Enhancements ------------ * The playlist code has been rewritten using conditional variables, which has resulted in speed improvements for the playlist code overall. * The behaviour for detecting filemagic has been redone and is now more scalable. * DoubleSize support has been improved significantly. * Race conditions in produce_audio() have been corrected. * The playlist editor is now properly drawn in shaded mode. * g_utf8_validate has been replaced with a more strict utf8 validator, which solves the encapsulated UTF-8 phenomenon properly. * The controlsocket code no longer uses the stack in an improper manner. * Behaviour regarding conversion of %20 in the playlist has been corrected. * Window decorations can now be enabled in preferences. * The translation templates have been simplified. * Support for translating plugins has been re-added. * The Jump-to-File code has been rewritten and now searches using regular expressions. In addition, the search is done without GTK queueing redraws for each entry, which has resulted in interface speedups. * The overall UI code has been reworked significantly and should be of better quality. Notable enhancements include support for UIManager xml file declarations. * A facility for searching and selecting/acting upon entries in the playlist has been added. * Support for multiple playlists has been added, and a playlist manager has been added to manage them. * Audacious now uses libmcs to store it's settings, which allows for tighter desktop environment integration. * Audacious now supports the XDG BASEDIR standard. * A lot of XMMS and BMP code has been either removed or rewritten so that it is less confusing. This should allow for easier maintainability in the future. * Debugging tools have been added for tracking memory allocation and freeing. * A safe signaling implementation has been added. * A voiceprint visualizer mode has been added. * Support for changing the tint in the graphics of skins has been added. * InputPlugins now use an InputPlayback context to track activity instead of blindly pushing data down a potentially undefined pipeline. * Audacious now uses GOption to parse the commandline. Bugfixes -------- * Too many to count.