1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
|
2025-02-24 Petter Reinholdtsen
* Version 0.3.0.
* This is a maintenance release of the current 0.x series,
containing documentation, stability and security (CVE-2009-3378
CVE-2009-3388) fixes.
2009-08-17 Viktor Gal <viktor.gal at maeth dot com>
* Version 0.2.0
* Fix for freeing uninitialized pointer in oggplay_seek
(ticket: 429)
* Fix semaphore handling on FreeBSD, Solaris and AIX.
* Unify the return value of semaphore handling functions
as they does not necessarily mean the same on various
OSes, e.g. Unix-like systems vs Micrsoft
* Check for EOF when reading initial Ogg Data
(ticket: 435)
* Added new function oggplay_get_duration that returns
the length of the Ogg content.
* Re-factored the YUV to RGB conversion functions.
Changed the API of the conversion functions in order to better
reflect what the functions will really do:
- oggplay_yuv2bgr -> oggplay_yuv2bgra
- oggplay_yuv2rgb -> oggplay_yuv2rgba
- oggplay_yuv2argb for supporting ARGB packaging.
- Fixed the vanilla implementation of the conversion functions.
- Added SSE2, SSE, MMX and AltiVec implementations of the
conversion functions.
- Added run-time CPU-extension detection.
- Added support for for YUV 4:2:2 and 4:4:4 pixel format support.
- Fixed the conversion functions for odd sized frames.
* Added libtiger support
* Mozilla bug 482461: Fix the presentation time for oggz-chopped
content.
* Mozilla bug 480014: Added integer overflow checks + input argument
validations in different functions. Two new error codes has been
introduced in OggPlayErrorCode:
E_OGGPLAY_TYPE_OVERFLOW: integer overflow detected.
E_OGGPLAY_OUT_OF_MEMORY: ran out of memory.
* Removed the use of the deprecated 'fish_sound_set_interleave' function.
* Fixed file closing in oggplay_file_reader_destroy: only close the
file handle if it was successfully opened.
* Mozilla bug 480063: stop the loop for requesting more chunks
if we received a shutdown event in the meanwhile.
Patch by Chris Double.
* Mozilla bug 492436: a track with UNKNOWN content type shall not be
ever active, as we do not know what codec shall be used.
Added dummy support for Dirac codec in 'callbacks' struct in order
to prevent out-of-bound access to the array.
A new error code has been introduced E_OGGPLAY_TRACK_IS_UNKNOWN
* Fix for liboggplay not correctly handling Theora files with offsets.
Patch by Gregory Maxwell <greg at xiph.org>
* Mozilla bug 493224 & 463726: enforce Ogg's requirement that
granule position cannot decrease in any part of the stream.
* Added cpu.c, ppc/oggplay_yuv2rgb_altivec.c, x86/oggplay_yuv2rgb_x86.c,
x86/yuv2rgb_x86.h to noinst_HEADERS in order to create a
proper archive of the current release.
* Added error checking for data callback function
If the OggPlayDataCallback is not set for the OggPlay handle
return with E_OGGPLAY_UNINITIALISED when calling oggplay_step_decoding
as there is no function that would process the decoded frames. Thus
there's no point in decoding.
* Added solution files for Visual Studio 2008
* Revisited the initialisation of the Ogg streams.
The oggplay_initialise function will try to read the headers of each
streams before returning, in order to avoid to be able to activate
streams with bogus headers.
The oggplay_set_track_active will return E_OGGPLAY_TRACK_UNINITIALISED
error code if the given stream was not properly initialised.
Changed the behaviour of the various callback functions (OggzReadPacket),
made them more liberal: if an error occures don't stop the whole
decoding of the Ogg content, but skip the packet, and hope that
there some valid packets of the streams are still about to come.
And in worst case scenario simply disable the given bogus stream, i.e.
stop decoding the rest of it.
* Mozilla bug 496051: Fix presentation time for theora streams.
From theora bitstream version 3.2.1 onwards, frame granule numbers
are relative to the end of the frame, i.e. frame granule numbers
start at 1. The presentation time is calculated by
(frame_number * granule_period), which is only correct if the
frame number starts at 0 (theora bitstream versions older than 3.2.1)
credits for Chris Pearce.
* Update the minimum required version of liboggz to 0.9.8
* Update the minimum required version of libfishsound to 0.9.1
* Updated shared version info to 2:0:1
2008-08-05 Silvia Pfeiffer <silvia at annodex dot net>
* Preparing for first release 0.0.1
|
