[pulseaudio-discuss] [ANNOUNCE] PulseAudio 0.9.22

[Lennart Poettering]

Heya,

http://0pointer.de/lennart/projects/pulseaudio/pulseaudio-0.9.22.tar.gz

Yes, it has been far too long for this release, my apologies!

You might wonder why I was difficult to reach the last months. In case
you haven't been following Linux news recently: Kay Sievers and I have
worked on getting 'systemd' off the ground, a project which I spent a
substantial amount of time on in the last months, and which will replace
the outdated init systems like Upstart and sysvinit in the big distros
sooner or later. The upcoming Fedora 15 release will ship it as we will
an upcoming OpenSUSE release, and many smaller distros already include
it in one way or another. 

But enough about systemd. Now that that project is pretty good in shape
and relatively complete I can divide ma time better and hope to do more
PA work again. And this release is the first bit of it.

So, again, my apologies for not doing my best maintenance work on PA the
recent months, but hey, it was all for the good cause, to bring Linux
ahead! ;-)

Lennart

Antti-Ville Jansson (1):
      combine: Handle reappearing slave sinks in non-automatic mode.

Arun Raghavan (2):
      Mark shared variables as volatile
      Add a configure option to change 'udevrulesdir'

Colin Guthrie (20):
      alsa: cover Input Source:Int Mic
      alsa: Cover the 'Int Mic Boost' element.
      core: Fix macro typo - PA_SINK_IS_LINKED -> PA_SINK_INPUT_IS_LINKED
      intended-roles: Do not pick monitor sources when doing automatic role-based device selection
      rtp: Fix bracketing in pa_rtp_recv.
      alsa: Fix assertion on mmap_write (triggered via a52 plugin)
      x11: Partially convert to XCB.
      alsa: Set the rewind safeguard proportionally to sample spec
      alsa: Only set the 'first' flag to false when we actually call snd_pcm_start()
      xcb: Ensure the XCB connection is valid before using it.
      xcb: xcb_get_setup() can return 0, so make sure we check it before using
      x11: Use the default screen for X11 properties.
      stream-restore: Clear the save_sink/save_source flags on apply_entry.
      augment-properties: Search for .desktop files in subfolders too.
      device-manager: Ensure that sinks/sources populate the device manager lists in order of their priority.
      suspend: Do not assert when checking for device suspended status and a stream is not linked.
      augment-properties: Fix debug messages and statement bracketing.
      intended-roles: Mark devices with a form factor of 'headset' as being appropriate for 'phone' streams
      sink-input: Fix comment
      combine: Only check if the sink is h/w etc. in automatic mode

Daniel Mack (2):
      alsa-mixer: add profile for Traktor Kontrol S4
      alsa-mixer: add profile for Native Instruments Korecontroller

Daniel T Chen (7):
      threaded-mainloop: Properly initialise m->n_waiting_for_accept to prevent deadlock
      udev: Use SOUND_CLASS instead of SOUND_FORM_FACTOR when checking for modem
      More src/pulsecore/cpu-arm.c FTBFS fixes
      Fix the following warnings (which now cause buildd failures in Ubuntu 10.04):
      Add missing profile and alsa-mixer/paths to src/Makefile.am
      Handle 'Digital Mic' as an 'Input Source'
      Handle 'Internal Mic 1' as an 'Input Source'

Daniel T. Chen (1):
      udev: handle sound cards with both modem and audio properly

David Fries (4):
      doxygen: Fix documentation typos
      doxygen: Fix the "all" comments regarding volume helper functions.
      doxygen: Documentation improvements
      doxygen: Add 'See also' linking to the overview page

David Henningsson (4):
      Fix crash on jack server shutdown
      jack: Prevent crash on jack server shutdown
      SSE/MMX/ARM: Fix high frequency noise with unusual number of channels
      Add "Rear Mic" to alsa mixer paths.

David Kågedal (1):
      alsa: add profile set for M-Audio FastTrack Pro USB

Jan Kratochvil (1):
      pulse: make sure legacy_dir is not static

Jez Austin (1):
      socket-client: properly handle asyncns failures

Kees Cook (1):
      core-util: ensure that we chmod only the dir we ourselves created

Lennart Poettering (59):
      dbus: remove filter functions only if they were actually set before
      native: fix request counter miscalculations
      core: make sure we always return a valid memblock in sink_input_pop() callbacks
      bluetooth: destruct stream only if it is not already destructed
      bluetooth: don't hit an assert if latency is queried for a stooped recording stream
      client: detect forking in sample cache API, too
      client: verify connection state in pa_stream_connect_upload()
      udev: don't forget to unref devices we are not interested in
      once: make once related variables volatile
      bluetooth: fix invalid memory access
      log: add an easy way to disable log rate limiting
      udev: make sure we get events only for sound devices
      alsa: ignore volume changes from the hw if we are not on the active console
      cpu: check for CMOV flag before using this intsruction in assembly
      udev: rework modem detection a bit
      daemon: first take name on the bus, then return in starter process
      alsa: cover bass boost mixer element
      udev: use ID_MODEL_ENC instead of ID_MODEL if it is set
      pacat: allow configuration of latency in msec
      client: implement $PULSE_LATENCY_MSEC
      client: include dolby channel names in comments
      native: rework handling of seeks that depend on variables the client does not know anything about
      alsa: cover "Desktop Speaker" mixer elements
      alsa: cover 'Shared Mic/Line in', 'Analog Source'
      alsa: cover "Internal Mic" elements
      alsa: use default output port names
      build-sys: add gobject to build dependencies
      padsp: emulate /dev/audio, too
      dbus: first restart timer, then dispatch it
      fdsem: be more verbose when reading from eventfd fails.
      pacat: always fully fulfill write requests
      pacmd: store away fd type
      pacmd: don't enter busy loop when reading from stdin very early
      shm: don't complain about missing SHM segments
      vala: fix definition of INVALID_INDEX
      vala: fix definition of the GLib mainloop adapter
      channelmap: Use "Subwoofer" as pretty name for LFE
      vala: fix wrapping of port setting calls
      proplist: explicitly mention a role 'test'
      stream-restore: be a little bit more verbose why we don't reastore a sink
      sample-cache: use the sample name as unmodified fallback in the properties when playing a sample from the cache
      scache: when playing a sample from the cache make sure not queue them up when the sink is suspended
      pacat: pass buffer_attr to recording streams too
      suspend-on-idle: resume audio device even for initially corked streams
      native: improve logging for buffer_attrs
      alsa-util: strip spaces from ALSA card/pcm names
      alsa: reset max_rewind/max_request while suspending
      core-util: introduce generic function pa_strip()
      esd,simple: use pa_memblockq_pop_missing()
      core: rework how stream volumes affect sink volumes
      Revert "Fix crash on jack server shutdown"
      thread: name all threads so that the names appear in /proc/$PID/task/$TID/comm
      udev: fix hex decoding
      alsa: don't make use of tsched related variables when tsched is disabled
      alsa-mixer: make a few strings translatable
      alsa-source: fix enum types
      build-sys: bump soname
      i18n: update POTFILES.in
      build-sys: simplify things for building on fedora

Pierre-Louis Bossart (3):
      alsa: disable rewinds when using ALSA plugins
      add rewind-safeguard parameter
      alsa: remove redundant call to snd_pcm_nonblock()

Stephen Moehle (1):
      upnp: Implement the MediaServer2 D-Bus interface

Tanu Kaskinen (2):
      libpulse: Store pa_stream pointers to hashmaps instead of dynarrays.
      alsa-sink: Get rid of a compiler warning regarding rewind_safeguard type.

Wanming.Zhang at tieto.com (1):
      man: Remove repeated lines

Wim Taymans (7):
      alsa: resume smoother after unsuspend
      alsa: increase the smoother window
      smoother: avoid losing precision
      alsa-source: refactor smoother and device start
      alsa-time-test: make test usable for capture too
      alsa: work around slightly broken _delay implementations
      alsa: make defines for smoother configuration

Lennart

-- 
Lennart Poettering - Red Hat, Inc.