]>
git.tue.mpg.de Git - paraslash.git/log
Andre [Sun, 28 May 2006 18:59:39 +0000 (20:59 +0200)]
audiod: kill close_writer()
This is better handled by the wng event handler.
Andre [Sun, 28 May 2006 18:51:01 +0000 (20:51 +0200)]
audiod: kill some dead code
Andre [Sun, 28 May 2006 18:49:05 +0000 (20:49 +0200)]
integrate writers into para_audiod
Seems to work, but might still be full of bugs.
Andre [Sun, 28 May 2006 05:35:13 +0000 (07:35 +0200)]
configure.ac: add writer objects to para_audiod.
Andre [Fri, 26 May 2006 14:48:15 +0000 (16:48 +0200)]
para_write: Own commandline parsers for each writer
Andre [Thu, 25 May 2006 21:57:48 +0000 (23:57 +0200)]
convert para_audiod to the new scheduler.
As expected, this was a bit harder. It works, but I'm sure there are
plenty of bugs left. Fortnunately, this doesn't matter much right now,
because the next step will be the integration of the writers into the
para_audiod executable. This will lead to major code simplifications,
so let's defer the bug-hunting until this integration is complete.
Andre [Thu, 25 May 2006 13:53:00 +0000 (15:53 +0200)]
introduce input_eof and ouput_eof
para_filter/para_audiod needs this. For example, it is pointless to
convert more audiod data if the writing application is no longer active.
Andre [Thu, 25 May 2006 12:49:03 +0000 (14:49 +0200)]
convert para_filter to the new scheduler
This was again straight forward.
Andre [Wed, 24 May 2006 01:45:28 +0000 (03:45 +0200)]
rename struct filter_chain_info to filter_chain
Andre [Wed, 24 May 2006 01:29:27 +0000 (03:29 +0200)]
Make para_recv use the new scheduler.
This was pretty straight-forward, but it broke audiod badly. The plan
is first to convert para_filter, then para_audiod. It this works out well,
para_server will be converted as well.
Andre [Tue, 23 May 2006 22:23:21 +0000 (00:23 +0200)]
error.h: kill some unused errors
Andre [Tue, 23 May 2006 21:49:31 +0000 (23:49 +0200)]
Rename error handler() to event_handler()
Andre [Tue, 23 May 2006 21:45:42 +0000 (23:45 +0200)]
simplify sched: nuke PRE_EOF_IS_ERROR
Much too simple to get it wrong. Instead, treat zero as success and
make all tasks return negative if they wish to have their error handler
called. A subsequent patch will rename error_handler to event_handler.
Andre [Tue, 23 May 2006 18:47:31 +0000 (20:47 +0200)]
check_wav: return error on premature end of file.
Andre [Tue, 23 May 2006 18:27:23 +0000 (20:27 +0200)]
Make the wng->eof a pointer.
The receiver/stdin_reader already have a corresponding field.
This patch also fixes a bug in the alsa_writer which was introduced
by the previous patch. This bug caused a playback delay of one second.
Andre [Tue, 23 May 2006 17:27:08 +0000 (19:27 +0200)]
first version of the universal paraslash scheduler
Andre [Fri, 19 May 2006 12:12:27 +0000 (14:12 +0200)]
fix serious typo
That max should actually be a min..
Andre [Thu, 18 May 2006 15:07:43 +0000 (17:07 +0200)]
error.h: new macro SS_NAME
expands to the name of the given subsystem.
Andre [Thu, 18 May 2006 12:42:05 +0000 (14:42 +0200)]
afs.c: fix documentation of afs_preselect()
Andre [Wed, 17 May 2006 20:41:51 +0000 (22:41 +0200)]
mp3_afh.c/ogg_afh.c: fix doxygen file documentation
Andre [Wed, 17 May 2006 19:47:27 +0000 (21:47 +0200)]
rename mp3.c to mp3_afh.c and ogg.c to ogg_afh.c
Andre [Wed, 17 May 2006 19:38:52 +0000 (21:38 +0200)]
split afs.h
move the audio format handler stuff to the new afh.h header file
which isn't included by any of the senders.
Andre [Wed, 17 May 2006 19:22:59 +0000 (21:22 +0200)]
major afs cleanup
This patch makes afs.c the only user of the audio format handler
arry afl[] in order to completely separate audio format handling from
audio file sending.
This is achieved by providing the new helper functions afs_get_header()
and afs_chunk_time(). audio file senders may call these functions
without knowledge of struct audio_format. This allows to get rid of
the "af" parameter for the pre_select and post_select functions of
struct sender.
Related to this cleanup is the removal of match_audio_file_name() from db.c
which also depended on afl[]. This function was essentially identical to
afs.c's guess_audio_format(), so make the latter public and use it from
within db.c.
Andre [Wed, 17 May 2006 19:00:26 +0000 (21:00 +0200)]
aac audio format handler: fix end of file timeout
20 x chunk_time is plenty, but lets be conservative here.
Andre [Wed, 17 May 2006 15:39:31 +0000 (17:39 +0200)]
fix audiod command line option handling.
The previous patch broke this.
Andre [Wed, 17 May 2006 13:57:58 +0000 (15:57 +0200)]
update grab_client_new() documentation
Andre [Tue, 16 May 2006 19:38:04 +0000 (21:38 +0200)]
grab: use grab_client_cmdline_parser_string()
No more need to use split_args().
Andre [Tue, 16 May 2006 16:38:18 +0000 (18:38 +0200)]
Make aac work with any bitrate.
By replacing the hardcoded 23ms timeout by a timeout
computed from the audio spec config.
Andre [Sun, 14 May 2006 02:04:01 +0000 (04:04 +0200)]
afs.c: Fix 'initialization from incompatible pointer type warning'
These were missed by the previous patch.
Andre [Sun, 14 May 2006 01:48:07 +0000 (03:48 +0200)]
struct audio_format_handler: use proper pointer type for the init function
Andre [Sun, 14 May 2006 00:42:56 +0000 (02:42 +0200)]
web/download text update
Andre [Sat, 13 May 2006 23:45:21 +0000 (01:45 +0200)]
improved info status items for the playlist selector
With this patch, the three "dbinfo" status items contain the current
dir, the number of entries in the playlist, and the previous/next
entry in the list.
Andre [Sat, 13 May 2006 22:06:02 +0000 (00:06 +0200)]
improved info text for the random selector.
It now reports random_dir, the dir of the current audio file and
the number of audio files found in random_dir.
Andre [Sat, 13 May 2006 21:18:44 +0000 (23:18 +0200)]
random/mysql selector: make it find m4a files
This is achieved by adding an array of typical suffixes
for the audio format to struct audio_format_handler.
Andre [Sat, 13 May 2006 20:46:34 +0000 (22:46 +0200)]
Rename struct audio format to audio_format_handler
Andre [Sat, 13 May 2006 20:26:55 +0000 (22:26 +0200)]
fix audiod decoder flags for aac
Do not try to encode the audio format in a single digit. Just use
0: off
1: receiver activated
2: writer started
3: receiver activated and writer is running
regardless of the audio format.
Andre [Sat, 13 May 2006 20:20:54 +0000 (22:20 +0200)]
fix init message of the playlist selector.
The prefix "dbinfo1:" was missing.
Andre [Sat, 13 May 2006 20:01:36 +0000 (22:01 +0200)]
add a close function for the aac audio format handler.
Without this, we're leaking one fd per audio file...
Andre [Sat, 13 May 2006 19:55:51 +0000 (21:55 +0200)]
Kill duplicate para_fread()
This was contained in aac_afh.c as well as in mp3.c. Move it
to fd.c and add documentation.
Andre [Sat, 13 May 2006 19:38:39 +0000 (21:38 +0200)]
documentation update
mention aac and clarify the audiod section in INSTALL
Andre [Sat, 13 May 2006 17:19:28 +0000 (19:19 +0200)]
Minor aac cleanup
Use size_t and ssize_t rather than a mixture of several other types.
This also makes gcc on Darwin STFU.
Andre [Sat, 13 May 2006 16:37:45 +0000 (18:37 +0200)]
aac afh: do not send the huge audio file header
Andre [Sat, 13 May 2006 05:09:52 +0000 (07:09 +0200)]
aac: error cleanup
put each aac-related error message to its proper subsystem
Andre [Sat, 13 May 2006 04:35:34 +0000 (06:35 +0200)]
aac: add GPL headers to all files
and some other minor cleanups.
Andre [Sat, 13 May 2006 04:20:17 +0000 (06:20 +0200)]
aac: move aac_find_stsz from aac_common.c to to aac_afh.c
It is only used there.
Andre [Sat, 13 May 2006 04:07:22 +0000 (06:07 +0200)]
aac: make find_stco() return the first offset
rather than the number of offsets which is rather uninteresting.
Also, rename this function to aac_find_entry() and return the
correct number of frames to para_server.
Andre [Sat, 13 May 2006 03:35:14 +0000 (05:35 +0200)]
Some more aacdec cleanups
rename the "decoder" field to "handle" as it is of type NeAACDecHandle.
Fix some typos in comments and make two more variables local.
Andre [Sat, 13 May 2006 03:23:31 +0000 (05:23 +0200)]
aacdec: Rename some functions
Andre [Sat, 13 May 2006 03:21:56 +0000 (05:21 +0200)]
aacdec: "consumed" is only used in mp4dec()
Remove it from struct private_aacdec_data.
Andre [Sat, 13 May 2006 03:14:45 +0000 (05:14 +0200)]
aacdec: inbuf and inbuf_len are only used in mp3dec()
No need to keep them around in the private_data struct.
Andre [Sat, 13 May 2006 03:03:48 +0000 (05:03 +0200)]
aacdec: no need to read the offset table.
We only need the first entry.
Andre [Sat, 13 May 2006 02:38:26 +0000 (04:38 +0200)]
First version of the aac audio format handler that kinda works.
ATM, only 2 channels, 44100Hz is supported. Still needs more work.
This also fixes two bugs in the error path of aacdec, where positive
values are returned due to a missing "-".
Andre [Fri, 12 May 2006 18:38:45 +0000 (20:38 +0200)]
First draft of the aac audio format handler
Not working yet. aacdec still works.
Andre [Fri, 12 May 2006 15:00:10 +0000 (17:00 +0200)]
Merge branch 'master' into aac
Conflicts:
Makefile.in
Makefile.in looks fine. WTF is git complaining about?
Andre [Fri, 12 May 2006 14:31:58 +0000 (16:31 +0200)]
new codename, reset version to git
Andre [Fri, 12 May 2006 14:28:42 +0000 (16:28 +0200)]
the paraslash-0.2.12 release tarball
Andre [Fri, 12 May 2006 14:26:34 +0000 (16:26 +0200)]
paraslash 0.2.12
Andre [Thu, 11 May 2006 17:33:43 +0000 (19:33 +0200)]
Typo
Andre [Wed, 10 May 2006 23:08:50 +0000 (01:08 +0200)]
Merge branch 'master' into aac
Conflicts:
configure.ac
configure.ac looks fine. Why the hell is git complaining?
nothing to commit
Andre [Wed, 10 May 2006 23:03:48 +0000 (01:03 +0200)]
configure.ac: Fix mysql configuration
The --enable_mysql_libs option had two cut-and-paste bugs.
Andre [Wed, 10 May 2006 22:59:19 +0000 (00:59 +0200)]
add configure options for libfaad
This patch also fixes two issues in the mysql part of configure.ac.
Andre [Wed, 10 May 2006 17:19:12 +0000 (19:19 +0200)]
ignore zero bit when checking for frame start
Andre [Tue, 9 May 2006 22:41:24 +0000 (00:41 +0200)]
Merge branch 'master' into aac
Andre [Tue, 9 May 2006 22:41:15 +0000 (00:41 +0200)]
aacdec: Do not try to feed invalid buffers to libfaad
Valid entry points start with 0x21, so look for the next 0x21
and skip everything in between.
Andre [Tue, 9 May 2006 17:26:47 +0000 (19:26 +0200)]
cosmetics
Andre [Tue, 9 May 2006 17:25:12 +0000 (19:25 +0200)]
check barriers also in afs_send_chunk()
As this function is called even if the select timeout didn't expire,
we'll have to check the eof_barrier and the data_send barrier also
early in afs_send_chunk() and return if there's time left.
Andre [Tue, 9 May 2006 16:18:44 +0000 (18:18 +0200)]
ortp_send: Fix EOF packet sending
Andre [Mon, 8 May 2006 22:29:33 +0000 (00:29 +0200)]
Merge branch 'master' into aac
Andre [Mon, 8 May 2006 22:23:17 +0000 (00:23 +0200)]
server: call afs_send_chunk() even if no timeout occured.
afs_send_chunk() checks the current time anyway.
Andre [Sat, 6 May 2006 22:49:11 +0000 (00:49 +0200)]
ortp_send: kill useless gettimeofday()
Andre [Sat, 6 May 2006 18:25:07 +0000 (20:25 +0200)]
Merge branch 'master' into aac
Andre [Sat, 6 May 2006 15:24:16 +0000 (17:24 +0200)]
fix error log message
Also, make recv_buffer() write a zero even if the underlying recv_bin_buffer()
failed.
Andre [Sat, 6 May 2006 15:21:10 +0000 (17:21 +0200)]
Send correct http message.
Screwed up since the 1.0 vs. 1.1 fix.
Andre [Sat, 6 May 2006 14:09:16 +0000 (16:09 +0200)]
grab: init gc->audio_format_num to -1
Previously it was 0, which happens to be the mp3 audio format number.
Therefore, grab without args only grabbed mp3 files. This patch makes
it grab any audio format.
Andre [Sat, 6 May 2006 14:06:31 +0000 (16:06 +0200)]
audiod: fix minor memory leak
The default filter setup code (which is only called at startup)
leaks four bytes.
Andre [Sat, 6 May 2006 13:33:20 +0000 (15:33 +0200)]
aacdec: Try do decode raw packets if header was not found
Andre [Fri, 5 May 2006 14:20:09 +0000 (16:20 +0200)]
Add M. Bakker to CREDITS
Andre [Fri, 5 May 2006 13:28:11 +0000 (15:28 +0200)]
aacdec: Fix some signedness issues
Andre [Fri, 5 May 2006 12:31:24 +0000 (14:31 +0200)]
Merge branch 'master' into aac
Andre [Fri, 5 May 2006 12:30:57 +0000 (14:30 +0200)]
compress: Adjust forumla for clipping
Use (3 * old + min) / 4 instead of (old + min) / 2 as the new gain
if clipping occures.
Andre [Fri, 5 May 2006 12:06:10 +0000 (14:06 +0200)]
Doxify the write subsystem
Andre [Thu, 4 May 2006 21:34:16 +0000 (23:34 +0200)]
aacdec: kill some dead code
Andre [Thu, 4 May 2006 21:33:00 +0000 (23:33 +0200)]
aacdec: raise min buffer size to 1000 bytes.
Fixes problems when used together with the wav filter.
Andre [Thu, 4 May 2006 18:08:59 +0000 (20:08 +0200)]
audiod: reset max_fileno at each run.
Kill some noisy debug messages also.
Andre [Thu, 4 May 2006 17:37:53 +0000 (19:37 +0200)]
First workable version of the aac decoder
Andre [Tue, 2 May 2006 18:16:14 +0000 (20:16 +0200)]
com_upd: always escape the filename
Fixes problems with filenames starting with "'"
Andre [Tue, 2 May 2006 14:58:32 +0000 (16:58 +0200)]
grab_client: Error out if given an invalid grab mode
Andre [Tue, 2 May 2006 00:26:07 +0000 (02:26 +0200)]
com_upd: ignore duplicates
Previously the update command aborted if files with duplicate basenames
were found. Suggested by Gerd Becker.
Andre [Mon, 1 May 2006 22:07:35 +0000 (00:07 +0200)]
fix trivial compiler warning in net.c
net.c:347:7: warning: extra tokens at end of #else directive
Andre [Mon, 1 May 2006 21:52:54 +0000 (23:52 +0200)]
configure: add two mysql-related command line options
Andre [Mon, 1 May 2006 16:02:31 +0000 (18:02 +0200)]
Add "-I ." to aclocal options
This way, one can put missing m4 files (like pkg.m4) to the paraslash
directory.
Andre [Mon, 1 May 2006 03:21:06 +0000 (05:21 +0200)]
mp3dec: skip invalid frame headers
First try to decode the header, and if that fails with a recoverable
error, return FRAME_HEADER_SIZE to skip the invalid frame.
Andre [Sun, 30 Apr 2006 00:33:30 +0000 (02:33 +0200)]
net.c: fix ifdef UCRED madness
simply use ordinary recv_buffer()/send_buffer() in case ucred is
not available.
Andre [Sat, 29 Apr 2006 23:34:43 +0000 (01:34 +0200)]
add a notice message if the expected pattern was not received
Andre [Sat, 29 Apr 2006 23:28:17 +0000 (01:28 +0200)]
1000 * 1000 microseconds is too much for Darwin
linux never complained about that, but Darwin returns -EINVAL.
Andre [Fri, 28 Apr 2006 20:40:32 +0000 (22:40 +0200)]
HTTP_GET_MSG: Do not resist on http 1.0
recent itunes for example sends 1.1. The obvious solution is to just
accept every client that sends at least 'GET / HTTP/'. Spotted by
Gerd Becker.
Andre [Mon, 24 Apr 2006 14:16:04 +0000 (16:16 +0200)]
Fix another gcc signedness warning in write.c
Andre [Mon, 24 Apr 2006 14:15:02 +0000 (16:15 +0200)]
minor mysql_selector cleanup.
make_message() never returns NULL, so remove the check.
Andre [Mon, 24 Apr 2006 14:13:44 +0000 (16:13 +0200)]
alsa_writer.c: Fix gcc-4 signedness warning
Andre [Mon, 24 Apr 2006 14:12:53 +0000 (16:12 +0200)]
NEWS: Mention the reason for size decrease of the tarball