Original commit message from CVS:
Fix some clocking issue in OSS. The issue is that if we seek forward (note: specifically forward-only), then we call handle_discont() before re-setting the clock to active. However, gstclock.c tells us that handle_discont only succeeds if allow_discont=TRUE, which is set in... set_active(TRUE). So, we first need to re-activate the clock and *then* call handle_discont(). More importantly, though, we should **NEVER EVER EVER EVER EVER** **NEVER EVER EVER EVER EVER** call clock_wait() after a forward discont without first having called handle_discont(). I don't know who added that code, but it's beyond fundamentally broken. clock_wait() **WAITS** until we're at the new given buftime, so if we do that on a forward-seek buffer, we... yes! we wait the amount of time that we seeked forward. Anyway, Apparently this code has been in here for quite a long time so I don't get how this can ever have worked...
Original commit message from CVS:
Remove ossgst... It's a crude hack (beyond ugly), it's broken and it failed to load during the last few months. If anyone wants to revive it, have fun finding it back in the CVS history
Original commit message from CVS:
Fix device probing from multiple childs. It's done once in the parent class only now, but the childs do get the correct values. Also fixes an incorrect succesful state change if we opened a v4l device that doesn't have the capabilities that are needed by the plugin.
Original commit message from CVS:
Don't change mixer if there's nothing to change. This caused a bug if the soundcard only supports one input and I call this function with rec=TRUE twice.
Original commit message from CVS:
Remove all config.h includes from header files, add it to each source file and remove duplicate config.h includes from several source files
Original commit message from CVS:
first bunch of conversions to new plugin_init. Includes libs/gst, gst/id3, sys/oss, ext/gnomevfs, gst/typefind and ext/mad.
You guessed it, everything Rhythmbox needs ;)
fixed BMP typefind and made gnomevfs one plugin instead of two while doing this
Original commit message from CVS:
Rename osselement to ossmixer and only open audio device if we have at least one pad. This makes ossmixer *only* open the mixer, which means we can open multiple mixer sessions.
Original commit message from CVS:
Some interface implementations for video4linux/video4linux2 plugins:
* a Tuner interface, with which one can select inputs and stations. Audio
work is underway here, but unfinished.
* A Xoverlay interface with which one can do simple overlay. Similar to
the API of the v4l/Xv XFree86 extension. Widget implementation for
GTK-2.0 coming up in the sandbox.
* Colorbalance - for adapting colors (brightness, contrast, etc.) - pretty
basic and maybe somewhat overdesigned. But it'll do for now.
Apart from these interfaces, there's also a loadable library 'xwindowlistener'
that listenes to X for the movement of a window and the overlap of other
windows. This is partly copied from xawtv (and thus partly GPL :(), but it's
needed for the xoverlay interface implementation in the v4l/v4l2 elements.
Lastly, some small changes to remove redundant properties from the v4l/v4l2
elements since these can be done much simpler. Comments appreciated!
Original commit message from CVS:
Make GstMixerTrack a GObject. I also want to make it emit several signals, starting work is in here but it's not fully implemented yet. for OSS, this will cause issues, but for ALSA, this is all automated.
Original commit message from CVS:
Interface implementation example: OSS mixer. Also osscommon->osselement so it can be loaded without being a source/sink (for a stand-alone mixer)
Original commit message from CVS:
* actually recurse into sndfile if we are able
* big ladspa cleanups, mainly to comply with the buffer-frames caps property, but also general
cleanups
- the samplerate prop is gone, if you want to set it explicitly (as in for get-based plugins)
you need to use a filtered connection, just like with buffer-frames
* big float2int and int2float changes for buffer-frames compatibility - I think it's quite a bit
simpler
* make the ossclock general, add it to gstaudio, and use it in sndfile as well
i need to update mimetypes, but that's coming soon. there are some other plugins that don't
support buffer-frames, i guess i need to get around to fixing them as well.
Original commit message from CVS:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
Original commit message from CVS:
implemented wait_async and unschedule ossclock, and support it in osssink -- really should make this a general clock, ill need it in gstsf
Original commit message from CVS:
compatibility fix for new GST_DEBUG stuff.
Includes fixes for missing includes for config.h and unistd.h
I only ensured for plugins I can build that they work, so if some of them are still broken, you gotta fix them yourselves unfortunately.
Original commit message from CVS:
Plugins cleanup:
* stereo2mono, mono2stereo, int2float, float2int: replaced by audioconvert.
* stereosplit replaced by oneton.
* vumeter replaced by level (and was broken anyway).
* avifile replaced by ffmpeg.
* mjpegtools duplicates functionality of jpeg. jpeg now works with jpeg-mmx,
too, which makes mjpegtools unneeded.
* allow for jpegmmx instead of jpeg.
* openquicktime replaced by qtdemux and ffmpeg. Broken anyway.
* XMMS is broken and will never be fixed.
* vga is broken and will not be fixed anywhere soon.
* videosink has never worked. If it works, add it back to replace xvideosink.
Original commit message from CVS:
- UNITS -> DEFAULT
- added chunk_size option to osssink, buffers will be written to the
devive in chunks of this size, this can increase the accuracy of the
clock on some devices.
Original commit message from CVS:
Some final fixes for the v4lsrc elements.
* remove software sync thread (use GST_ELEMENT_THREAD_SUGGESTED instead)
* make all src elements threadsafe
* fix num_buffer argument setting in v4l2src (VIDIOC_S_PARM)
* re-add bufsize (RO) for v4lmjpegsrc
* fix the A/V sync calculation in all elements (spvf=GST_SECOND/fps, not GST_SECOND*fps)
* probably some more crap....
With all this, it actually works quite well. The TODO files describes the
next steps in order to make a full-featured video recorder based on these
elements and GStreamer (bottom). Making a simple recorder should be fairly
easy now, btw.
Original commit message from CVS:
This implements filtered-caps negotiation for all the v4l*src elements, and removes the accompanying properties since they're no longer needed
Original commit message from CVS:
* implement clocking
* set clock counter back to zero on ready->paused
* move open/close to ready/null instead of paused/ready.
* add random typos
Original commit message from CVS:
another batch of connect->link fixes
please let me know about issues
and please refrain of making them yourself, so that I don't spend double
the time resolving conflicts
Original commit message from CVS:
Make the OSSSrc set the correct rate/number of channels as set in the GstCaps.
Handle state changes correctly according to docs/random/wtay/states
Original commit message from CVS:
v4l plugins:
* add open/close signals
v4l2 plugins:
* add open/close signals
* move source format enumeration from v4l2element to v4l2src
* adapt to the final v4l2 API in kernel 2.5 (patches for 2.4 on http://bytesex.org/patches)
* small tweaks
Original commit message from CVS:
This patch fixes some issues caused by design issues in video4linux, adds
some nicety to video4linux2 plugins and does some more evil stuff:
* video4linux doesn't tell us which formats are supported by a card, so
the only way to know this is by simply trying it out. This patch adds that.
* v4lmjpegsink didnt have a bufferpool yet - is integrated now.
* all copy() bufferpool functions have been removed since they're not needed.
* v4lmjpegsink doesnt have a free() function, because hen playing the frames,
all this is already handled. When the frame is not played, nothing has to
be done. In total, the function is not needed.
* adds a get_caps() function to v4l2src
* some minor crap
