[vdr] VDR developer version 1.3.13

[Klaus Schmidinger <Klaus.Schmidinger@cadsoft.de>]

VDR developer version 1.3.13 is now available at

    ftp://ftp.cadsoft.de/vdr/Developer/vdr-1.3.13.tar.bz2

*** NOTE THAT THE WARNINGS REGARDING THE USE OF VPS CONTROLLED
*** TIMERS FROM PREVIOUS RELEASE NOTES STILL APPLY!

Well, it's been a long time without a new developer version.
First I have spent several weekends trying to get VDR to run on
kernel 2.6 with the dvb-kernel driver, and then I was hunting
buffer overflows for several weekends. So the main focus in this
new release is overall stability when doing several parallel
recordings. This now works a lot better than before:-)

There are still several things that have been postetd on the VDR ML
which I haven't been able to attend to, yet. These will be considered
in future versions. I hope to be able to release new versions more
often now ;-)

The changes since version 1.3.12:

- Fixed checking for the presence of NPTL (thanks to Jouni Karvo).
- Making sure section filters are only set if the device actually has a lock
  (thanks to Andreas Share for pointing this out).
- Fixed a possible NULL pointer assignment in cMenuText::SetText() (thanks to
  Marco Schlüssler).
- Fixed a crash in case the last line in channels.conf is a group separator and
  that group is selected in the channel display (thanks to Dick Streefland).
- Added cRingBufferLinear::Read() to read directly from a file handle into the
  ring buffer.
- Using timeouts in ring buffers to avoid 'usleep()'.
- Clearing the 'Transfer Mode' ring buffer after clearing the device to avoid
  an "almost full" ring buffer.
- Removed locking from cRingBufferLinear for better performance under high load.
- Using a cRingBufferLinear in cRemux to avoid unnecessary copying of data.
- Using a cRingBufferLinear in cTSBuffer and filling it in a separate thread
  to avoid buffer overflows. Plugins using cTSBuffer will need to remove the
  call to the now obsolete Read() function (see cDvbDevice::GetTSPacket() for
  the new usage of cTSBuffer).
- cRemux::Process() has been split into Put(), Get() and Del() to allow for a
  better decoupling of the remuxing and disk writing process. Plugins using
  cRemux will need to be modified accordingly.
- The actual disk writing in recordings is now done in a separate thread to
  improve the overall throughput.
- Changed cRemux so that it returns the maximum available amount of data with
  each call, not just 2048 byte.
- Added a visual display of all cRingBufferLinear buffers for debugging. To
  activate it, define DEBUGRINGBUFFERS in ringbuffer.h.
- Instead of cCondVar now using the new cCondWait (which also avoids a possible
  "near miss" condition; thanks to Sascha Volkenandt for pointing out this one).
  cCondVar is still present for plugins that use it (and VDR itself also still
  uses it in cRemote).
- The cRingBuffer now does EnableGet()/EnablePut() only if the buffer is more than
  one third full or empty, respectively. This dramatically improves recording
  performance and reduces system load (thanks to Marco Schlüßler for doing some
  testing regarding buffer performance and giving me some hints that finally led
  to finding out that this was the basic problem causing buffer overflows).
- Improved Transfer Mode (thanks to Marco Schlüßler for suggestions and testing).
- Fixed a possible crash with inconsistent SI data (thanks to Marcel Wiesweg).
- Fixed showing the replay mode if the OSD is currently in use (thanks to Kimmo
  Tykkala for pointing out this problem).
- cOsdProvider::NewOsd() now always returns a valid pointer, even if the OSD is
  currently in use (it will then return a dummy cOsd object and write a message to
  the log file).
- Added Estonian language texts (thanks to Arthur Konovalov).
- Fixed 'newplugin' and libsi/Makefile to use the compiler defined in $(CXX) for
  generating file dependencies (thanks to Andreas Brachold).
- Moved the initialization of aPid1 and aPid2 to the beginning of cDvbDevice::cDvbDevice()
  to have them set in case a patch references them (thanks to Wayne Keer for pointing
  this out).
- Completed the Russian OSD texts (thanks to Vyacheslav Dikonov).
- Avoiding unnecessary section filter start/stops (thanks to Marco Schlüßler).
- Made the "Channel not available!" message and mtInfo instead of mtError (suggested
  by Wayne Keer).
- Made volume control more linear (thanks to Emil Naepflein and Udo Richter).
- Now skipping code table info in SI data (suggested by Milos Kapoun).
- Added missing Czech characters to fontosd-iso8859-2.c (thanks to Milos Kapoun).
- Fixed a crash in the time search mechanism (reported by Reinhard Nissl).
- If one PID can't be added, the whole cDevice::AttachReceiver() will now fail
  and all PIDs added so far will be deleted (thanks to Marco Schlüßler for
  pointing out this one).
- Now only saving channels.conf after a modification made by the user (avoids
  lots of disk access due to automatic channel updates). Automatic channel
  modifications will be saved every 10 minutes if no recording is currently
  active.
- Removed the 'Log' parameter from the cChannel::Set... functions. Instead
  checking if the channel has a non-zero number.
- Updated 'channels.conf.terr' for Hannover (thanks to Sven Kreiensen).

Have fun!

Klaus