Permalink
Commits on Mar 8, 2012
  1. Fix FireWire compilation on mac when compiling in 64 bits mode.

    jyavenard committed Mar 1, 2012
    Also, darwin firewire headers were being included also for non-darwin platforms
    
    Conflicts:
    
    	mythtv/libs/libmythtv/firewiredevice.cpp
    [cherry-picked from 3196c2c]
  2. Fix missing headers when compiling under mac 10.5

    jyavenard committed Mar 8, 2012
    Why it worked without them, not sure.. They certainly should have been there to start with
    
    Conflicts:
    
    	mythtv/libs/libmythupnp/mmulticastsocketdevice.cpp
    
    [cherry-picked from b522d60]
  3. Fix compilation when using 10.5 SDK.

    jyavenard committed Mar 8, 2012
    For some reasons, loading the OpenGL headers fail when using the 10.5 SDK. This is due to some data types not being defined.
    
    So add the framework where those types are defined.
    
    I couldn't find an elegant way to query what the current SDK is, so instead if compiling under 10.5 SDK, you define an extra -D_USING_105SDK=1 to CFLAGS and CXXFLAGS
    
    [cherry-picked from d65a742]
  4. Fix compilation on OS 10.7

    jyavenard committed Mar 7, 2012
    Missing header
  5. Delay first call to QTextCodec until after main QApplication has been…

    ctreleaven authored and jyavenard committed Mar 4, 2012
    … created
    
    Calling QTextCode before QApplication has been created, causes Qt to ignore local plugins and use system-wide instead. Not a problem on linux because they are one and the same, but on mac they are different: resulting in loading two sets of Qt libraries in RAM, ultimately leading to a crash
    
    Fixes #9447
    
    Signed-off-by: Jean-Yves Avenard <jyavenard@mythtv.org>
    [cherry-picked from 4392c16]
  6. Look in more than one location for libdvdcss on mac.

    jyavenard committed Mar 5, 2012
    We look into:
    Contents/Frameworks
    Contents/PlugIns <- proper spelling
    Contents/Plugins <- spelling used in old packager
    
    libdvdcss used to be store in the Plugins directory, and was loaded by mythtv while searching for myth plugins, generating warnings about invalid plugin.
    The library should be in Frameworks instead, but we don't want to break the old behaviour
    Add both PlugIns and Plugins, so the bundle can work on case sensitive file systems
    [cherry-picked from 18fdd67]
  7. Fix universal package build with new osx-builder

    jyavenard committed Mar 5, 2012
    For darwin/OSX, ffmpeg's dynamic libs were built with an id including the full installation path. This meant that upon linking to those libs, mythfrontend would use an absolute path to the libav* ; making the universal build invalid.
    
    Change the linker flags for creating dynamic libraries to remove the absolute path and add headerpad_max_install_names so we can later update the path while creating the application bundle and not get an error that there's no space left.
    
    Conflicts:
    
    	mythtv/configure
    [cherry-picked from d1a6cc8]
  8. Fix universal architecture compilation of myth on mac.

    jyavenard committed Mar 2, 2012
    Following 456083c, found that qmake handle compilation arguments differently according to them being between quotes or not
    
    [cherry-picked from e34c8d7]
  9. Add --firewire-sdk configuration switch to configure

    jyavenard committed Mar 1, 2012
    The configure script only looked into /Developer to find the mac FireWire SDK. Unfortunately the FireWire SDK available on Apple developer site is for i386 and ppc archicture only. As such compilation in 64 bits would fail on mac
    The --firewire-sdk allows to point to an alternative SDK, one that would be 64 bits compatible (as new mac build script allow)
    
    Conflicts:
    
    	mythtv/configure
    [cherry-picked from 21ad34b]
  10. Fix mac compilation when Qt Frameworks are in use.

    jyavenard committed Mar 1, 2012
    Qt has a special handling for mac frameworks that includes header files. Resetting QMAKE_LIBDIR_QT breaks qmake from generating proper makefile with correct include path.
    Fix various other issues such as QMAKE_LFLAGS being never properly set. For the time being limit the change to mac compilation in order to minimise the risk of breaking something else elsewhere.
    HD homerun compilation failed due to sig_t being previously defined in sys/signal.h
    
    Conflicts:
    
    	mythtv/libs/libmythmetadata/libmythmetadata.pro
    [cherry-picked from 5dccbd5]
  11. Merge branch 'fixes/0.24' of github.com:MythTV/mythtv into jyavenard/…

    jyavenard committed Mar 8, 2012
    …backports/fixes/0.24
    
    Conflicts:
    	mythtv/libs/libmyth/audiooutputbase.cpp
Commits on Mar 7, 2012
  1. Correct issue passing python library name for dependency checking.

    wagnerrp committed Mar 4, 2012
    (cherry picked from commit 366d96d)
Commits on Mar 6, 2012
  1. AvFormatDecoder: add FORCE_DTS_TIMESTAMPS environment variable

    tralph committed Mar 6, 2012
    This allows a user to force the player to use video dts timestamps for av-sync
    purposes in the event they have material with busted pts timestamps.
    
    backport 40c69e1
    
    Fixes #10007.
Commits on Mar 5, 2012
  1. Increase kMaxUIWaitTime to 10 seconds. The theory is that this timeou…

    stuartm committed Mar 5, 2012
    …t was to prevent a very delayed response to an availability check causing playlists to resume playing long after the user had moved on. The present timeout was very short though and didn't account for delays caused by disks needing to be spun up or just the latency of checking multiple storage groups for a file. Almost certainly fixes #8316
    
    (cherry picked from commit 6c3a799)
Commits on Mar 2, 2012
  1. Fix 0.24 build after backporting dead key support. The log macros wer…

    stuartm committed Mar 2, 2012
    …e renamed in 0.25, they aren't really needed in 0.24 so just remove them.
  2. Add basic dead key support to mythuitextedit. Implemented from scratc…

    stuartm committed Mar 2, 2012
    …h because I couldn't see a way of getting QT to handle it. To start with the following characters are supported, but if patches are provided I will add more - áâàãäå éêèë ç íîìï ñ óôòõö úûùü ýÿ . Fixes #8304 (cherry picked from commit d84d5c9)
Commits on Feb 21, 2012
  1. Do not attempt to play audio with more than 8 channels if not support…

    jyavenard committed Feb 21, 2012
    …ed natively.
    
    While trying to play an encrypted video over airplay, all nasty things would happen. So add a safeguard so we never try to play more than 8 channels audio unless the hardware supports it (e.g. not requiring downmixing)
    
    [cherry-picked from c5b6e92]
Commits on Feb 17, 2012
  1. Fix reccount value for the Default group

    stuartm committed Feb 17, 2012
    (cherry picked from commit 4588c0e)
Commits on Feb 15, 2012
  1. Fix ToggleUpmix return value.

    jyavenard committed Feb 14, 2012
    The boolean returned would have only been valid if audio was configured as 5.1. For 7.1 system, it would have always returned false
    
    [cherry-picked from 584647e]
  2. Fix CanUpmix()

    jyavenard committed Feb 14, 2012
    Following caafff6, CanUpmix would have only reported if we were currently upmixing
    
    [cherry-picked from c85b33c]
  3. Size in resampling output buffer is in frames, not samples

    jyavenard committed Feb 14, 2012
    [cherry-picked from 0ef4b58]
Commits on Feb 14, 2012
  1. Alter the watched flag fix slightly to take into account the need to …

    stuartm committed Feb 14, 2012
    …ignore post-roll for recordings, something I forgot before.
    
    (cherry picked from commit b758ea8)
  2. Fix videos not being marked as watched. We had some code to handle in…

    stuartm committed Feb 14, 2012
    …-progress recordings where we didn't have the total number of frames, this used the start/end time of the recording which can't be trusted for video because they are based on the length from downloaded metadata e.g. An episode of Frasier might only be 22 minutes long without adverts but according to the metadata it's 30 minutes. We now trust the totalFrames to be accurate for anything which isn't an in-progress recording.
    
    (cherry picked from commit f53a00f)
Commits on Feb 12, 2012
  1. Fix mono audio playback.

    jyavenard committed Feb 12, 2012
    Follow-up 8adf652, which was incorrect
    
    [cherry-picked from ce8844a]
  2. Minor code cleanup.

    jyavenard committed Feb 12, 2012
    Replace occurrence of 4 with sizeof(float) where appropriate
    
    [cherry-picked from ce8844a]
  3. Fix mono audio playback.

    jyavenard committed Feb 12, 2012
    Follow-up 8adf652, which was incorrect
    
    [cherry-picked from 699918f]
Commits on Feb 11, 2012
  1. Fix issue that could occur when playing mono content upmixed to stereo.

    jyavenard committed Feb 11, 2012
    Stupid typo
    
    [cherry-picked from 8adf652]
  2. Fix issue that could occur when playing mono content upmixed to stereo.

    jyavenard committed Feb 11, 2012
    Stupid typo
    
    [cherry-picked from 8adf652]
  3. Minor code cleanup in audio

    jyavenard committed Feb 11, 2012
    [cherry-picked from 375e78c]
  4. Fix a couple memory leaks. cppcheck.

    daniel-kristjansson authored and jyavenard committed Feb 2, 2012
    [cherry-picked from ca59bf3]
  5. Mythtranscode: Fix compilation following AudioOutput changes.

    Mark Kendall authored and jyavenard committed Nov 23, 2011
    [cherry-picked from f8f90e5]
  6. Fix type of outsize and locality of outsize and data_size. cppcheck.

    daniel-kristjansson authored and jyavenard committed Feb 1, 2012
    We were using an unsigned for outsize but setting it from a function that returns -1 on error and then checking the unsigned value for -1.
    
    [cherry-picked from 74ace05]