- Added cDevice::DeviceNumber() to get the number of a device, not counting any
gaps that might be in the index count.
- Fixed fetching the current/next information to handle cases where the duration
of an event is set wrongly and would last beyond the start time of the next
event.
- Adapted type names to the new HEAD version of the driver (which the previous
NEWSTRUCT branch has been merged into). Note that to use this driver version
you still need to add NEWSTRUCT=1 to the make call when building VDR. You
need a HEAD version of the LinuxDVB driver dated 2002-10-11 or later to compile
VDR with NEWSTRUCT=1.
- Fixed radio channels in channels.conf.cable (thanks to Robert Schiele and Uwe
Scheffler).
- Fixed switching the video format in the Setup/DVB menu (thanks to Uwe Scheffler
for reporting this one).
- Reactivated full handling of second audio PID (even in 'Transfer Mode').
- Fixed a crash when closing down with remote control plugins (thanks to Oliver
Endriss helping to debug this one).
- Commands in the file 'commands.conf' can now have a '?' at the end of their
title, which will result in a confirmation prompt before executing the
command.
- Changed a few leftover 'new char[...]' to MALLOC(char, ...).
- If a command executed from the "Commands" menu doesn't return any output, the
OSD will now be closed automatically.
- The SVDRP command PUTE now triggers an immediate write of the 'epg.data' file
(suggested by Gerhard Steiner).
- The new configuration file 'reccmds.conf' can be used to define commands that
shall be executed from the "Recordings" menu; see MANUAL and 'man vdr(5)' for
details (suggested by Gerhard Steiner).
- Adapted VDR to the NEWSTRUCT driver. To use the new driver, compile VDR with
'make NEWSTRUCT=1' (thanks to Holger Wächtler for some valuable advice).
By default it currently still uses the old driver.
- Added some missing #includes (thanks to Martin Hammerschmid).
- Changed the log error message "can't record MPEG1!" to "error in data stream!",
since the mentioning of MPEG1 has irritated many people.
- Consistently using malloc/free and new/delete (thanks to Andreas Schultz).
- Temporarily made cDevice::ProvidesCa() virtual (Andreas Schultz needs this
in his DXR3 plugin).
- cDevice no longer exposes a file handle to cPlayer. A derived cPlayer class
can now call DevicePoll() to see whether the replay device is ready for
further data. A derived cDevice class must implement Poll() and shall
check if any of its file handles is ready for data.
- Implemented several replay modes to allow players that play only audio (thanks
to Stefan Huelswitt).
- Improved cCondVar::Wait() and implemented cCondVar::TimedWait() (thanks to
Stefan Huelswitt).
- VDR no longer gives up if there is no DVB device. It continues to work if
there is at least one device, either a DVB device found by the core VDR code
itself, or a device implemented by a plugin.
- Re-visited the race condition fix in the cDvbPlayer (thanks again to Andreas
Schultz).
- Changed the VFAT handling to allow users who normally use it but have forgotten
to set it when compiling a new version of VDR to at least see their recordings
made with VFAT enabled (thanks to Christian Rienecker).
- Added some missing teletext PIDs (thanks to Joerg Riechardt).
- Fixed PID handling for cReceiver.
- Added a missing #include to ringbuffer.c (thanks to Martin Hammerschmid).
- Now using CC, CFLAGS, CXX and CXXFLAGS in Makefile.
- Changed the cDevice class to allow plugins to implement their own devices (see
PLUGINS.html for details).
- Added Norwegian language texts (thanks to Jrgen Tvedt).
- Increased the usleep value in cDvbOsd::Cmd() to 5000 in order to work on
systems with the KURT/utime-patch (thanks to Guido Fiala).
- Changed the check whether the driver is loaded in runvdr to check for the
'dvb' module (the last one loaded).
- Fixed repeat function with LIRC (thanks to Stefan Huelswitt).
- Increased the upper limit for the symbol rate to 30000 (thanks to Ulrich
Rder).
- Made the position of the channel display configurable (thanks to Stefan
Huelswitt).
- Made the width and height of the OSD configurable (thanks to Stefan Huelswitt).
- DiSEqC support can now be generally enabled/disabled in the Setup menu. This
may be necessary if your multiswitch gets irritated by the default DiSEqC
codes '0' (thanks to Markus Lang).
- Fixed replaying in case there is no index file.
- Fixed jumping to an editing mark when replay has been paused.
- Avoiding unnecessary code execution in the replay progress display (thanks
to Guido Fiala).
- When entering time values the digits that still have to be entered are now
shown as '-' (as in "1-:--").
- When setting an editing mark while the progress display is not active, the
display will now be turned on for a short while to indicate the successful
setting of the mark.
- Updated 'channels.conf' for Premiere World (thanks to Helmut Schchner).
Check your timers if you use this channels.conf file, since the sequence of
several PW channels has been changed.
- Changed the color of "Info" messages to "black on green" and that of the
confirmation messages (like "Delete...") to "black on yellow".
- Fixed display with DEBUG_OSD (it still crashes sometimes, esp. when replaying,
but I can't seem to find what causes this... any ideas anybody?).
- Avoiding audio/video distortions in 'Transfer Mode' by no longer actually
tuning the primary interface (which can't receive this channel, anyway).
Apparently the driver gets irritated when the channel is switched and a
replay session is started immediately after that.
- Increased timeout until reporting "video data stream broken" when recording.
- Explicitly switching back to the previously active channel after ending a
replay session (to have it shown correctly in case it was in 'Transfer Mode').
- Fixed SVDRP commands LSTC and LSTT to make them return an error message if
no channels or timers are defined.
- Enhanced 'channels.conf.cable' (thanks to Hans-Peter Raschke).
- Fixed switching to another channel via the EPG while a recording is being
replayed.
- Fixed a memory leak in the EIT processor that happened when the system time
was set.
- Removed some redundant code from the cListBase destructor.
- Fixed internationalization of some Main menu texts.
- Updated 'channels.conf' after the recent changes of Premiere World (thanks
to Axel Gruber).
- Redesigned the ring buffer to make it work with two separate threads for
input and output (also prepared for using a remultiplexer).
- Fixed setting system time from transponders.
- Fixed a segfault in the Schedule menu in case there is no EPG information.
- The 'runvdr' script now kills any leftover vdr threads before restarting it.
- Fixed a problem with Daylight Saving Time when displaying the times of
recordings.
- Added Dutch language texts (thanks to Arnold Niessen).
- The new command line option -t can be used to set the controlling terminal
(thanks to Jrgen Sauer). This is especially useful when starting VDR through
an entry in /etc/inittab (see INSTALL).
- Since the CAM module only works if it is installed in the "highest" DVB card,
recordings now search for a free DVB card from lowest to highest index (as
opposed to the previous "highest to lowest" search) in order to not use the
CAM card for FTA recordings unless necessary. This is only important for
systems with three or more DVB cards.
- Added the "statdvb2vdr" tool from Hans-Peter Raschke.
- Fixed a segfault that sometimes happened when killing VDR.
- VDR now returns an exit status of '2' in case of an error at startup, instead
of terminating with 'abort()' (which caused a core dump).
- SVDRP now also works with clients that don't do line buffering (like the
Windows 'telnet').
- Empty lines in config files no longer cause error messages.
- New SVDRP command LSTE to list the EPG data.
- The SVDRP HELP command now prints the topics in several columns.
