Add release notes for 11.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
This commit is contained in:
Anton Khirnov 2014-08-18 18:04:47 +00:00 committed by Diego Biurrun
parent 4d55e9de27
commit 12f0388f9c

View File

@ -1,51 +1,50 @@
Release Notes
=============
* 10 "Eks"
* 11 "One Louder"
General notes
-------------
One of the main features of this release is the addition of reference-counted
data buffers to Libav and their use in various structures. Specifically, the
data buffers used by AVPacket and AVFrame can now be reference counted, which
should allow to significantly simplify many use cases. In addition,
reference-counted AVFrames can now be used in libavfilter, avoiding the need
for a separate libavfilter-specific frame structure. Frames can now be passed
straight from the decoders into filters or from filters to encoders.
With this release we are trying to answer the numerous calls from our users for
shorter development cycles. From now on we will aim for approximately two major
releases per year.
These additions made it necessary to bump the major versions of libavcodec,
libavformat, libavdevice, libavfilter, and libavutil, which was accompanied by
dropping some old deprecated APIs. These libraries are thus not ABI- or API-
compatible with the previous release. All the other libraries (libavresample
and libswscale) should be both ABI- and API-compatible.
Libav 11 is API-, but not ABI-compatible with the previous major release. This
means that the code using our libraries needs to be rebuilt, but no source
changes should be required. Note however, that a number of old APIs remain
deprecated and will be dropped in the near future. All users are strongly
encouraged to update their code as soon as possible. The doc/APIchanges file in
the Libav source tree and the migration guide on the wiki should help with
migration to the new APIs. If those are not sufficient, do not hesitate to
contact us on IRC or through the user mailing list.
Another major point is the inclusion of the HEVC (AKA H.265, the successor of
H.264) decoder in the main codebase. It was started in 2012 as a Libav Google
Summer of Code project by Guillaume Martres and subsequently completed with
the assistance of the OpenHEVC project and several Libav developers.
One specific API issue in libavformat deserves mentioning here. When using
libavcodec for decoding or encoding and libavformat for demuxing or muxing,
the standard practice was to use the stream codec context (AVStream.codec) for
actual decoding or encoding. There are multiple problems with this pattern
(the main one is that the decoder/demuxer or encoder/muxer are not necessarily
synchronized and may overwrite each other's state), so it is now strongly
discouraged and will likely be deprecated in the future. Users should instead
allocate a separate decoding or encoding context and populate it from the
demuxing codec context (or the reverse for encoding) with the
avcodec_copy_context() function.
As usual, this release also contains support for other new formats, many smaller
new features and countless bug fixes. We can highlight a native VP9 decoder,
with encoding provided through libvpx, native decoders for WebP, JPEG 2000, and
AIC, as well as improved WavPack support with encoding through libwavpack,
support for more AAC flavors (LD - low delay, ELD - enhanced low delay), slice
multithreading in libavfilter, or muxing chapters in ASF. Furthermore a few new
filters have been introduced, namely compand, to change audio dynamics, framepack,
to create stereoscopic videos, asetpts, to set audio pts, and interlace, to convert
progressive video to interlaced. Finally there is more fine-grained detection of
host and target libc, which should allow better portability to various cross
compilation scenarios.
The main highlights of this release include native Opus, VP7, OpenEXR, and On2
AVC decoders, HEVC encoding through libx265, new APIs for exporting ReplayGain
and display transformation metadata and countless bug fixes. A large effort was
also expended on internal cleanups which are not very visible to our users,
but should make the codebase cleaner, safer and easier to maintain and extend.
One point worth mentioning is refactoring the large monolithic framework for
architecture-specific codec optimizations into small blocks, which reduces the
size of configurations that selectively enable or disable certain codecs.
See the Changelog file for a fuller list of significant changes.
Please note that our policy on bug reports has not changed. We still only accept
bug reports against HEAD of the Libav trunk repository. If you are experiencing
issues with any formally released version of Libav, please try a current version
of the development code to check if the issue still exists. If it does, make
your report against the development code following the usual bug reporting
guidelines.
The avserver streaming tool, which has not been maintained for many years and
was mostly broken, was removed from the tree. It was decided that it is a
significant maintenance burden and that we do our users no service by pretending
to support it, while we in fact do not.
See the Changelog file for a more extensive list of significant changes.
API changes
-----------
@ -54,31 +53,23 @@ A number of additional APIs have been introduced and some existing functions
have been deprecated and are scheduled for removal in the next release.
Significant API changes include:
[libavutil]
+ added the reference-counted buffers API (buffers.h)
+ moved the AVFrame struct to libavutil and added a new API for working with
reference-counted AVFrames (frame.h)
[libavcodec]
+ added an API for working with reference-counted AVPackets (av_packet_*)
+- converted VDPAU to the hwaccel framework; the old way of using VDPAU is no
longer supported
- old audio encoding and decoding APIs removed
- old video encoding API removed
- deprecated enum CodecID removed (enum AVCodecID should be used instead)
- deprecated audio resampling API removed (libavresample should be used
instead)
+ Added the avcodec_copy_context() function that must from now on be used for
freeing codec contexts.
+- Added a new VDA hardware acceleration API, since the old one was broken and
not fixable in a compatible way. Deprecated the old VDA API.
[libavfilter]
+- replaced AVFilterBufferRef with AVFrame; AVFilterBufferRef and everything
related to it still exists, but is deprecated
+ converted all filters to use the AVOptions system for configuration, it is
now possible to query the supported options, their values and set them
directly with av_opt_*
+ added a slice multithreading framework
+- merged avfiltergraph.h to avfilter.h, using AVFilterGraph is now explicitly
mandatory (it was implicitly required even before); added new API for
allocating and initializing filters
[libavformat]
+ Added support for exporting stream-global (as opposed to per-packet) side
data. This feature is now used by some demuxers to export ReplayGain or
display transformation matrix (aka rotation) or stereoscopic 3D mode.
+ Added an API for live metadata updates through event flags.
+- Changed the way to provide a hint about the desired timebase to muxers.
Previously it was done by setting AVStream.codec.time_base. Now callers
should set AVStream.time_base.
[libavresample]
+ Added an API for working with AVFrames.
Please see the file doc/APIchanges for details along with similar
programmer-centric information.