rjk@greenend.org.uk [Sun, 16 Sep 2007 16:18:55 +0000 (17:18 +0100)]
correct backend docs
rjk@greenend.org.uk [Sun, 16 Sep 2007 15:50:58 +0000 (16:50 +0100)]
mention broadcast, broadcast_from in disorder_config(5)
rjk@greenend.org.uk [Sun, 16 Sep 2007 15:20:45 +0000 (16:20 +0100)]
alsa code builds again
Richard Kettlewell [Tue, 4 Sep 2007 21:53:55 +0000 (22:53 +0100)]
Install disorderd under launchd in Mac OS X.
disorder-speaker can now transmit RTP over the network. Effectively
untested and not yet fully documented.
New playrtp client. Totally untested, might not even build.
Richard Kettlewell [Sat, 25 Aug 2007 16:19:02 +0000 (17:19 +0100)]
rescan reports progress every 1000 tracks
Richard Kettlewell [Sat, 25 Aug 2007 14:29:28 +0000 (15:29 +0100)]
fix scratch path
Richard Kettlewell [Sat, 25 Aug 2007 14:28:54 +0000 (15:28 +0100)]
only report volume change fail once
Richard Kettlewell [Sat, 25 Aug 2007 12:58:47 +0000 (13:58 +0100)]
refer to README.raw
Richard Kettlewell [Sat, 25 Aug 2007 12:58:00 +0000 (13:58 +0100)]
mention plugin install locations
Richard Kettlewell [Sat, 25 Aug 2007 12:57:52 +0000 (13:57 +0100)]
mention raw players
Richard Kettlewell [Wed, 22 Aug 2007 19:42:44 +0000 (20:42 +0100)]
make sure tests is last directory
Richard Kettlewell [Wed, 22 Aug 2007 18:14:03 +0000 (19:14 +0100)]
gc.h might be in /sw/include/gc
Richard Kettlewell [Wed, 22 Aug 2007 18:13:52 +0000 (19:13 +0100)]
put the newline in the right place
rjk@greenend.org.uk [Sun, 19 Aug 2007 15:55:30 +0000 (16:55 +0100)]
README hint for debian users
Richard Kettlewell [Sun, 19 Aug 2007 15:54:53 +0000 (15:54 +0000)]
DB_RECORDCOUNT has been deprecated for ages and has gone in db4.6
Richard Kettlewell [Tue, 31 Jul 2007 19:17:18 +0000 (20:17 +0100)]
skeletal test infrastructure
rjk@greenend.org.uk [Sun, 29 Jul 2007 18:48:47 +0000 (19:48 +0100)]
don't leak cn->container. FINALLY.
rjk@greenend.org.uk [Sun, 29 Jul 2007 18:40:54 +0000 (19:40 +0100)]
further horrid memory debugging stuff
rjk@greenend.org.uk [Sun, 29 Jul 2007 17:10:55 +0000 (18:10 +0100)]
report cache size with MDEBUG
rjk@greenend.org.uk [Sun, 29 Jul 2007 16:56:54 +0000 (17:56 +0100)]
use clear_children() more consistently
rjk@greenend.org.uk [Sun, 29 Jul 2007 16:55:11 +0000 (17:55 +0100)]
don't leak list of children
rjk@greenend.org.uk [Sun, 29 Jul 2007 15:59:48 +0000 (16:59 +0100)]
more verbose menu items
Richard Kettlewell [Sun, 29 Jul 2007 14:48:34 +0000 (15:48 +0100)]
quieten compiler
rjk@greenend.org.uk [Sun, 29 Jul 2007 14:36:28 +0000 (15:36 +0100)]
,commit
Richard Kettlewell [Thu, 26 Jul 2007 19:32:52 +0000 (20:32 +0100)]
remember to dist disorder-speaker.8
Richard Kettlewell [Wed, 18 Jul 2007 19:11:16 +0000 (20:11 +0100)]
a couple of missing pcre -> gc bindings
Richard Kettlewell [Tue, 17 Jul 2007 21:44:51 +0000 (22:44 +0100)]
Don't link against -lgc if not using garbage collection.
Richard Kettlewell [Sun, 15 Jul 2007 21:38:58 +0000 (22:38 +0100)]
-liconv where needed
rjk@greenend.org.uk [Sun, 15 Jul 2007 16:43:59 +0000 (17:43 +0100)]
disorder-rescan now obsoletes tracks that belong to no collection;
handy when you remove some collection. The code should be relatively
robust against such tracks but they'll annoy the random track picker.
trackdb_rescan() is thus modified to be able to return all tracks.
find_track_collection() maps a track to its collection, or 0 if it
doesn't have one.
rjk@greenend.org.uk [Sun, 15 Jul 2007 15:09:56 +0000 (16:09 +0100)]
minor tidying following last change
rjk@greenend.org.uk [Sun, 15 Jul 2007 15:07:16 +0000 (16:07 +0100)]
Abolish shared libdisorder; now we use a static one.
New server/exports.c ensures that all the right symbols are exported
to plugins.
Richard Kettlewell [Sun, 15 Jul 2007 14:05:32 +0000 (15:05 +0100)]
disobedience: don't crash if no password
rjk@greenend.org.uk [Tue, 10 Jul 2007 23:28:18 +0000 (00:28 +0100)]
update README.streams to be bidirectional. (don't cross the streams\!)
rjk@greenend.org.uk [Tue, 10 Jul 2007 23:14:31 +0000 (00:14 +0100)]
disobedience more robust against server restart
rjk@greenend.org.uk [Tue, 10 Jul 2007 23:09:01 +0000 (00:09 +0100)]
updated CHANGES
rjk@greenend.org.uk [Tue, 10 Jul 2007 23:08:49 +0000 (00:08 +0100)]
New queue_pad option defines how big to keep the queue (by adding
random tracks). User-chosen tracks are added before the final block
of random tracks.
Also updated CHANGES a bit.
rjk@greenend.org.uk [Tue, 10 Jul 2007 23:08:22 +0000 (00:08 +0100)]
note bug
rjk@greenend.org.uk [Tue, 10 Jul 2007 22:30:41 +0000 (23:30 +0100)]
more speaker mode care
rjk@greenend.org.uk [Tue, 10 Jul 2007 22:23:39 +0000 (23:23 +0100)]
speaker process is now more consistent about whether it's using ALSA
or writing to a subprocess.
rjk@greenend.org.uk [Tue, 10 Jul 2007 21:53:42 +0000 (22:53 +0100)]
New configuration option sox_generation to choose between different
sox syntaxes.
Richard Kettlewell [Tue, 10 Jul 2007 20:38:54 +0000 (21:38 +0100)]
ignore tarballs
rjk@greenend.org.uk [Sun, 8 Jul 2007 14:19:56 +0000 (15:19 +0100)]
docs note that sox is required for speaker_command
rjk@greenend.org.uk [Sun, 8 Jul 2007 14:15:17 +0000 (15:15 +0100)]
missing break in soxargs()
rjk@greenend.org.uk [Sun, 8 Jul 2007 14:05:50 +0000 (15:05 +0100)]
build speaker even if no alsa
rjk@greenend.org.uk [Sun, 8 Jul 2007 13:44:43 +0000 (14:44 +0100)]
document speaker_command and sample_format
rjk@greenend.org.uk [Sun, 8 Jul 2007 13:35:16 +0000 (14:35 +0100)]
Merge from Mark's branch.
revno: 9
committer: mdw@distorted.org.uk
branch nick: disorder
timestamp: Sat 2007-06-09 12:58:16 +0100
message:
server/speaker: Wake up on POLLERR on kidpipe too.
Ooops. The theory was that the speaker process would notice an EPIPE on
its kid pipe and respawn its kid. Unfortunately I don't understand
poll(2) enough, and failed to listen for POLLERR, so in fact the speaker
goes into a tailspin if its kid dies.
------------------------------------------------------------
revno: 8
committer: Mark Wooding <mdw@distorted.org.uk>
branch nick: disorder
timestamp: Wed 2007-05-30 11:56:52 +0100
message:
Support streaming to external process.
Introduce two new configuration variables:
* speaker_command: Shell command to pipe audio to instead of using an
ALSA device.
* sample_format: The sample format expected by this process in the
form BITS/RATE/CHANNELS -- each integers, except that BITS may be
suffixed by `b' or `l' for big- or little-endian respectively. The
default is 16/44100/2.
If speaker_command is unset, everything is as it used to be. If it's
set, however, disorder-speaker will resample its input (using sox) to
conform to the desired format if necessary, and pipe the result to the
command's stdin. Only one speaker_command is run at a time (respawned
automatically if it quits), and it receives the audio data of many
tracks.
An example command (and the reason I did this grim hack):
lame -h -r -s44.1 -b128 -x -mj --preset standard - - |
ices -c /etc/disorder/ices.conf >/dev/null
re-encodes as an MP3 stream and feeds the result to my IceCast server.
rjk@greenend.org.uk [Sun, 8 Jul 2007 13:21:58 +0000 (14:21 +0100)]
Merge from MDW's branch:
revno: 10
committer: mdw@distorted.org.uk
branch nick: disorder
timestamp: Wed 2007-06-27 09:38:12 +0100
message:
templates: Kill slash bogus @url@/? entries in templates.
rjk@greenend.org.uk [Sun, 8 Jul 2007 13:16:16 +0000 (14:16 +0100)]
make speaker logging less verbose
rjk@greenend.org.uk [Sun, 8 Jul 2007 13:03:38 +0000 (14:03 +0100)]
Handle underrun when detected during poll() setup as well as from
snd_pcm_writei(); previously the speaker would just get stuck in this
case. Add some extra logging of ALSA parameters as part of the
debugging process.
rjk@greenend.org.uk [Sat, 7 Jul 2007 16:48:53 +0000 (17:48 +0100)]
more mm care in disobedience/choose.c
rjk@greenend.org.uk [Sat, 7 Jul 2007 16:11:22 +0000 (17:11 +0100)]
copyright dates
rjk@greenend.org.uk [Sat, 7 Jul 2007 15:57:35 +0000 (16:57 +0100)]
disobedience: Clear out the root before refilling it.
rjk@greenend.org.uk [Sat, 7 Jul 2007 15:10:40 +0000 (16:10 +0100)]
compute_alias() is now also hardened against tracks outside any
collection. It takes the view that such never have an alias, no
matter what.
rjk@greenend.org.uk [Sat, 7 Jul 2007 15:07:09 +0000 (16:07 +0100)]
find_track_root() now issues a log message if it finds a track outside
any collection. Should make it clearer what's going on if something
believes the answer and crashes.
track_to_words() now includes a bodge for such tracks to avoid a
crash. Ultimately the answer is to expunge such tracks from the
database but while my db is in a bizarre state it's worth making the
code a bit more robust.
rjk@greenend.org.uk [Sat, 7 Jul 2007 14:30:48 +0000 (15:30 +0100)]
Disable use of libgc on libglib and descendants. I believe that
somewhere the latter is still allocating memory outside of libgc's
ability to detect it, but I cannot determine where.
GtkScale format-value callbacks fixed to use g_strdup().
rjk@greenend.org.uk [Sat, 7 Jul 2007 14:06:08 +0000 (15:06 +0100)]
insist on GC_all_interior_pointers
Richard Kettlewell [Tue, 3 Jul 2007 23:39:33 +0000 (00:39 +0100)]
Enable gc_mem_gc_friendly ad G_SLICE_CONFIG_ALWAYS_MALLOC, the latter
turning out to be critical for running GTK+ and libgc together.
Richard Kettlewell [Sun, 1 Jul 2007 22:07:20 +0000 (23:07 +0100)]
note about configuration required by disobedience
Richard Kettlewell [Sun, 1 Jul 2007 21:51:52 +0000 (22:51 +0100)]
More Debianization stuff:
* depend on debconf
* fix indenting of description
* postrm does init.d and debconf purge
Richard Kettlewell [Sun, 1 Jul 2007 21:29:37 +0000 (22:29 +0100)]
use dist-hook to ship bzr changelog
Richard Kettlewell [Sun, 1 Jul 2007 19:37:14 +0000 (20:37 +0100)]
Remove some wacky leftover debugging.
check_suitable() now rejects tracks that don't belong to any
collection. This situation can arise if you remove a collection from
the configuration.
A remaining bug is that even reproducible data for the tracks remains
in the database long after the collection has been removed.
Richard Kettlewell [Sun, 1 Jul 2007 19:33:25 +0000 (20:33 +0100)]
debianization makefile hackery
Richard Kettlewell [Sun, 1 Jul 2007 19:33:07 +0000 (20:33 +0100)]
ignores for local build
Richard Kettlewell [Sun, 1 Jul 2007 16:16:50 +0000 (17:16 +0100)]
more ignores
Richard Kettlewell [Sun, 1 Jul 2007 14:50:16 +0000 (15:50 +0100)]
Include bzr changelog in ChangeLog.d.
Richard Kettlewell [Sun, 1 Jul 2007 14:45:16 +0000 (15:45 +0100)]
Fixes for 64-bit Linux support. Reported by Ross Younger.
Richard Kettlewell [Sun, 1 Jul 2007 13:37:53 +0000 (14:37 +0100)]
change version to reflect development status
Richard Kettlewell [Sun, 1 Jul 2007 13:30:11 +0000 (14:30 +0100)]
Debianization fixes from Colin Watson.
Richard Kettlewell [Sun, 1 Jul 2007 13:12:13 +0000 (14:12 +0100)]
Fix sedfiles.make to cope with changed make syntax. Reported by Colin
Watson and Michael Stevens.
Richard Kettlewell [Sun, 1 Jul 2007 13:09:29 +0000 (14:09 +0100)]
Remove redundant words in configure output. Reported by Colin Watson.
Richard Kettlewell [Mon, 28 May 2007 17:43:03 +0000 (18:43 +0100)]
reference bzr
rjk@greenend.org.uk [Sun, 28 Jan 2007 12:01:25 +0000 (12:01 +0000)]
configure.ac: Correct definition of BUILD_SPEAKER for active case.
Richard Kettlewell [Sun, 28 Jan 2007 11:54:39 +0000 (11:54 +0000)]
configure.ac: Only look for ALSA libraries on Linux. On other platforms,
we don't try to build the speaker subprocess.
server/speaker.c: Don't build on platforms without ALSA.
These two files are tied together via the BUILD_SPEAKER define, so that
when the speaker process supports other APIs, it can be turned back on.
rjk@greenend.org.uk [Tue, 23 Jan 2007 22:51:34 +0000 (22:51 +0000)]
vertical-align:: wants 'middle', not 'center'
See: http://www.w3.org/TR/REC-CSS2/visudet.html#propdef-vertical-align
rjk@greenend.org.uk [Tue, 23 Jan 2007 21:45:45 +0000 (21:45 +0000)]
Remove arch tags throughout
Add some ignores to keep 'bzr status' happy
rjk@greenend.org.uk [Tue, 23 Jan 2007 21:23:45 +0000 (21:23 +0000)]
Import from Arch revision:
rjk@greenend.org.uk--2004/disorder--mainline--0.1--patch-328