





/**
 Many small changes (not recorded, because the library is still
 a part of kmpg)
*/


0.2 b
-----
* code cleanup
* added more devices (volumedevice in software, mixerdevice,..)
* moved examples out of the yaf directory, because the examples
  depend on kde 
* thought about a port to gtk. It's possible, but for now I stay 
  with kde.
  (If you want to port it to the gtk signal mechanism->email me)

0.2 a
-----

* major downgrade (back to the roots!)
  removed various command for backend-mpg123 decoder :
           
  -no buffersupport anymore
  -no pre-decoding (including decoding messages)
  -no configuration options (scalefactor,quality,resync,range_check,...)
  -no realease /dev/dsp (if possible)
  -no runtime infos of typ 1/2 anymore
  -all bug-workarounds removed

* moved command "jump" to the xplayer interface
* added new command to xplayer interface: outputfile <file>
  (Note: outputfile is _not_ a device. No ioctl should
  be performed. Its simply a file or fifo where the output
  which normally goes to /dev/dsp is written)
* added interface for kmpg GUI
* new directory structure


0.1 c
-----

* workaround: when pre-decoding doesn't work (because the mpg differs
  in  the samplerate) the mpg is (later) played with the "speed"
  of the old mpg.
  


0.1 b
-----

* bugfix in jump routine


0.1 a
-----


* checked if it possible to close /dev/dsp when the player is "off"
  This can be done. Every time when the player if "off" the
  child is killed. This frees the device.

* added faster "rewind" when the stream is a file

* fixed minor bugs

* added configuration options.
  see config-help
  	
* added equalizer support

* checked whether http://... works.(Yes)
  (just enter "open http://<path to mp3>")

* checked if a yaf port for splay is possible.(not yet)



initial release 0.1 (mpg123 support)
------------------------------------


* now buffering works.
  Its seems that due to a design flaw in mpg123 
  pre-buffering (and thus pre-decoding)
  is only possible if the following mp3 file has the same rate,frequencies,...
  Pre Buffering is not possible if the following mp3 differs in these
  things.

  The frontend must handle this.
  yaf now sends "decoding messages" (currently only necessary
  for the mpg123 decoder). If everything works fine it sends

  E.g:
  
  Command:0 Msg:decode_start ok

  If decoding is not possible:(because of the wrong format) it sends:

  Command:0 Msg:decode_start cannot prebuffer


  Thus the frontend must take care that it re-opens the mp3
  after it got the Message that the decoder if "off"
  

* for better debugging support:
  if in the current directory a file with the name:
  yaf.script 
  exists it reads first these commands in the script and then
  wait for input on stdin.

  Example for a yaf.script:
  runtime off
  open my.mp3
  play
