linx-simulator2/node_modules/lame/deps/mpg123/NEWS.libmpg123
2019-09-18 11:11:16 +03:00

134 lines
3.5 KiB
Plaintext

Changes in libmpg123 libtool interface versions...
36.0.36
- Extended MPG123_RESYNC_LIMIT to initial header search.
- Not cutting decoder delay unconditionally anymore (only in combination with known encoder delay / padding).
35.0.35
- Added mpg123_meta_free().
34.0.34
- Added flag MPG123_AUTO_RESAMPLE.
- Changed (improved;-) outbuffer behaviour.
33.0.33
- Added MPG123_BUFFERFILL.
32.0.32
- Added mpg123_framepos()
31.0.31
- Added mpg123_framedata() and MPG123_IGNORE_INFOFRAME.
30.0.30
- Added MPG123_FEEDPOOL and MPG123_FEEDBUFFER.
29.0.29
- New decoder: ARM neon.
- Added support for 24 bit output (dumb byte-chopping of 32 bit output).
28.0.28
- Add mpg123_strlen().
27.0.27
- Implictly disable seeking on streams when client enforced ICY parsing.
This helps debugging dumps of http streams.
26.0.26
- Added mpg123_encsize().
- Added flag MPG123_SKIP_ID3V2.
25.0.25
- Version increase to mark the point where the split between normal and large-file-enabled library vanishes again. The world did not like it.
Now Thomas lost some days of recreation and sleep to give it a dual-mode libmpg123 on large-file-sensitive systems.
24.0.24
- Introduce mpg123_replace_reader_handle() and mpg123_open_handle()
... this is also in preparation for the next version which will drop the separated large-file library again, due to public display of dismay.
- Add the experimental mpg123_framebyframe_decode to the off_t-sensitive functions.
23.0.23
- Version increase to mark the point where the explicit split between normal and large-file-enabled library has been introduced.
22.0.22
- Experimental framebyframe API added.
21.0.21
- Added support of unicode file names under windows via UTF-8 argument to mpg123_open.
- Added mpg123_feature(), for example to test for the above behaviour.
20.0.20
- New flag: MPG123_PLAIN_ID3TEXT
- Corresponding text encoding handling API added:
mpg123_enc_from_id3, mpg123_store_utf8
19.0.19
- Hm, what was it exactly now... there are candidates:
- runtime dithering
- free format
- ARM optimizations
18.0.18
- new parameter: MPG123_PREFRAMES is now tunable (the number of frames to decode and skip before a seek point), also default value increased
17.0.17
- introduction optimized stereo synths
16.0.16
- introducing floating point x86-64 SSE synth
15.0.15
- first addition of x86-64 SSE optimizations
14.0.14
- the first libmpg123 with actually working MPG123_UPSPEED
- also important regression fix concerning skipping of frames
13.0.13
- The whole set of output formats is generally available (8, 16 and 32 bit integer, signed/unsigned, float)
- Many features can be absent from libary as build decision (minimize binary size).
12.0.12
- added mpg123_current_decoder
- fixed value of MPG123_ENC_FLOAT
- float output now is a real runtime option
11.0.11
- added mpg123_getstate
- run-time tunable frame index
- officially configured with largefile support where available
10.0.10
- new flag MPG123_FUZZY (along with the fuzzy seek functionality)
9.0.9
- added mpg123_tell_stream
8.0.8
- added mpg123_get_eq
7.0.7
- added mpg123_set_filesize
6.0.6
- added mpg123_icy2utf8
5.0.5
- added mpg123_feed
- input buffers now const
4.0.4
- extended mpg123_string api (mpg123_add_substring, mpg123_grow_string)
3.0.3
- Initial floating point support as compile-time option.
2.0.2
- New flag MPG123_SEEKBUFFER.
1.0.1
- Added MPG123_RESYNC_LIMIT parameter.
- Added MPG123_OUT_OF_SYNC and MPG123_RESYNC_FAIL error codes.
- Fix for uninitialized framesize value in handle.