0.16    27.12.2003
    - added support for Catweasel MK3 card
    - added support for PCI HardSID cards
    - more modular for easier support of new card types
    - updated kernel compatibility and cleanups
    - updated Makefile for 2.6 kernels

0.15a   26.1.2002
    - fixed read ioctl that Sidplay2 is going to use

0.15    25.1.2002
    - added /proc support for 2.2 kernels. The file is not
      /proc/driver/hardsid but /proc/hardsid as there is no /proc/driver
      directory in 2.2 :( For consistency moved it in 2.4 kernels too.
    - fixed a mute problem from the sidplay2 patch
    - added ability to disable filters via ioctl
    - reduced speaker pops and clicks by modifing reset ioctl
    - fixed oops with multiple chips
    - added flush, read and delay ioctls
    - all configurable things are now normal module options
    - writes cleaned up, only one function does it anymore
    - ioctl definitions put to separate header file
    - faked reads to write only registers speed up some tunes
    - added Quattro hack to play a tune through all SIDs on a Quattro card
    - compatibility with the first 2.5 kernels

0.14    23.2.2001
    - CVS repository moved to Sourceforge
    - updated sidplay2 patch, unpacked
    - added devfs support
    - added card detection override hack
    - added mute support

0.13    27.1.2001
    - SID type autodetection tested to work with 8580
    - DUH! 0.12 oopses if you try to open an unexisting device...
    - slightly updated sidplay2 patch

0.12    27.1.2001
    - fixed a problem with 2.2 kernels. It seems doing a sema_init on
      semaphore doesn't initialize it properly, but doing a init_MUTEX before
      it does... grr...
    - added newer sidplay2 patch, check README for build instructions

0.11    25.1.2001
    - tested to work on Alpha! Required compiler options are documented in
      Makefile
    - multiple SID support! Now you can use 1-16 SIDs at the same time
      using /dev/sidXX where each device maps to one SID. This has been
      tested to work with hacked Sidplay2 using stereo .MUS files and two
      normal HardSID cards (on Alpha :)
    - device numbers changed for multisid support (60 by default, change
      from source), minors are from 0 (/dev/sid0) to 15 (/dev/sid15). I 
      applied for an assigned major for this device (60 is marked as for 
      local/experimental use), let's see how it goes.
    - fixed a bug with reads and Quattro cards
    - simplified reset ioctl

0.10    24.1.2001
    - moved to Sourceforge! http://hardsid.sourceforge.net, or
      http://www.sourceforge.net/projects/hardsid, as the first URL
      doesn't seem to work yet.
    - some cleanup and initial support for Quattro cards (card type and all
      chips should be detected, only first one is used for now though).
    - added ioctl to query card type

0.09    15.1.2001
    - Added support for dummy writes (register 0x1f) for delays longer than
      0xffff cycles, fixed David Whittaker's Lazy Jones.
      
0.08    14.1.2001
    - Added CVS tags
    - Fixed reset, the previous version actually didn't work...

0.07    14.1.2001
    - #include cleanup
    - some ioctls (reset, fifo status, SID type)

0.06    5.1.2001
    - Added autodetection of the HardSID cards including checking the chip
      type (6581/8580). Both are of course untested except with my 6581 at
      0x300 :) Success/failure reports very welcome, especially from 8580
      users. The code detects all four possible cards, but uses only the
      first one for now.
    - SID reading now actually works, you have to OR 0x20 to the register
      when writing it to the chip before reading.

0.05    3.1.2001
    - Ouch, reads froze the machine, now fixed. They still are quite
      problematic as the whole buffer has to be flushed before reading.

0.04    31.12.2000
    - All kinds of statistics
    - Some timing changes

0.03    18.12.2000
    - SMP support now tested to work, some small locking changes
    
0.02    16.12.2000
    - Added some SMP locks	
    - Tested to work on 2.2.18
    - Always delay a bit (in this case two cycles) to give SID some time
      to process writes
    - Reset all registers on open, turn master volume on for some tunes
    - Wait until all writes are done before reading

0.01	15.12.2000
    - Totally untested support for 2.2 kernels

0.0     15.12.2000
    - Initial release