Age | Commit message (Collapse) | Author |
|
the visual type should be checked too.
That because a plugin with the same id may support different visuals
(e.g. xv -> VISUAL_TYPE_X11, xv -> VISUAL_TYPE_X11_2).
CVS patchset: 8388
CVS date: 2006/11/23 21:05:31
|
|
allow developers to use special locking support of their plataform (like
QT library mutex).
note: only xshm has been ported.
based on Christoph Pfister's patch.
CVS patchset: 8350
CVS date: 2006/10/28 17:02:51
|
|
CVS patchset: 8344
CVS date: 2006/10/16 22:18:24
|
|
entirely, to avoid an off-by-one in decoder_map.
Found by Coverity Scan CID 122.
CVS patchset: 8304
CVS date: 2006/09/26 21:32:08
|
|
(implemented for mp3 demuxer only)
CVS patchset: 8177
CVS date: 2006/08/12 01:43:26
|
|
CVS patchset: 8137
CVS date: 2006/07/21 03:31:42
|
|
CVS patchset: 8063
CVS date: 2006/06/20 00:35:07
|
|
- rint, lrint, M_PI (ffmpeg and goom)
- avoid mkdir warnings
- check for older _GUID_DEFINED too
- ressurected generating inttypes, cleaned public os_types.h
- moved function redefiniton used only in the old MSVC to central place
Windows port update:
- build internal ffmpeg:
- workaround 'near' keyword problem
- proper using __declspec(dllimport)
Other changes:
- LIBFFMPEG_CFLAGS --> LIBFFMPEG_CPPFLAGS
CVS patchset: 7964
CVS date: 2006/04/08 16:42:23
|
|
CVS patchset: 7897
CVS date: 2006/02/14 18:47:46
|
|
"The plugin_node_comparator function used by the plugin sorted arrays
sorts the plugins by priority ASC instead of DESC."
CVS patchset: 7860
CVS date: 2006/02/01 15:11:42
|
|
CVS patchset: 7848
CVS date: 2006/01/27 07:46:09
|
|
Improved cross-compiling: detect and precedence of the host-specific libs
W32dll: enabled by platforms as originaly, user can override it
Fixed shell errors in the Makefile of the hackersguide
Detect the place of SDL header independently on platform
Fixed spelling in CDDA, include the header sys/param.h if available
Fixed format strings (fix warnings (or even bugs) on Gentoo/FreeBSD)
CVS patchset: 7815
CVS date: 2005/11/28 12:24:56
|
|
added missing mutex init in xine_ticket (fixes pause assert with win32-pthreads)
added missing destroys due to pthread
doc spells and houskeeping
use directx of older versions too, possibility to use xine w/o HW accel
CVS patchset: 7738
CVS date: 2005/09/19 16:14:02
|
|
audio can be heard with non-1X speeds if audio.synchronization.slow_fast_audio is set
CVS patchset: 7726
CVS date: 2005/09/11 22:07:48
|
|
CVS patchset: 7723
CVS date: 2005/09/11 00:11:09
|
|
CVS patchset: 7715
CVS date: 2005/08/30 15:27:58
|
|
Windows ports updates:
- finished M$VC port update
- moved xine plugins back to normal location for Windows ports,
added relative directories (used relative to ${prefix}),
make sure it's thread-safe (but it's still hacky)
- fixed ugly bug with loading plugins, when working drive was differrent from
plugins directory
directory
- directory and subdirectory separator called as platform specific
- fix crash with newer win32 pthread (missing mutex_init in vo plugin), yes!
- small doc update
CVS patchset: 7712
CVS date: 2005/08/29 15:28:16
|
|
Windows ports fixes and improvements due to my current work on toxine:
- first experiments with external win32 pthreads,
more portable code (pthread_t may be a struct)
- headers refactored
- moved dirent win32 replacement to lib/, hide it for frontends,
used system version, if found, not used non-POSIX dirent->d_reclen
(this item doesn't work in MinGW), fix memleak in dvb
- separated settings for postproc and avcodec when using external ffmpeg
- check for malloc.h in public xine.m4, used it conditionally in xine headers
- replaced random() by POSIX more common rand()
- prevent one segfault in directx vo plugin, if fails
- M$VC port update
CVS patchset: 7709
CVS date: 2005/08/25 15:36:29
|
|
CVS patchset: 7659
CVS date: 2005/07/18 01:46:02
|
|
(Based on a patch from Stefan Bruens <lurch@gmx.li>.)
CVS patchset: 7648
CVS date: 2005/07/17 16:52:14
|
|
don't stop looking for plugins just because one of them has failed to initialize.
fix the problem of xine not loading anything (first plugin was gnome-vfs and
it doesn't work here)
CVS patchset: 7536
CVS date: 2005/05/14 16:00:40
|
|
fix a couple of places were catalog->lock was not released (causing
lockup/deadlock). cleanup some functions to return at a single place
so we won't make this mistake again.
(please double check - i'm a little rusty on xine hacking ;-)
now i have to find out why my xine checkout is not able to load any input plugins...
CVS patchset: 7535
CVS date: 2005/05/14 15:41:15
|
|
snprintf(s, sizeof(s), ...) is safe; no need to waste one byte
CVS patchset: 7473
CVS date: 2005/04/20 17:08:56
|
|
was parsing the wrong description array.
CVS patchset: 7457
CVS date: 2005/04/13 19:07:17
|
|
sprintf -> snprintf replacements
CVS patchset: 7455
CVS date: 2005/04/10 16:23:05
|
|
CVS patchset: 7452
CVS date: 2005/04/10 09:31:16
|
|
CVS patchset: 7438
CVS date: 2005/03/09 22:24:20
|
|
Check for plugin init failure
CVS patchset: 7434
CVS date: 2005/03/06 13:03:33
|
|
CVS patchset: 7410
CVS date: 2005/02/13 21:21:06
|
|
CVS patchset: 7408
CVS date: 2005/02/13 16:12:35
|
|
void xine_register_plugins(xine_t *self, plugin_info_t *info);
The goal of this function is to allow a frontend to register its own statically linked plugin(s). See this thread :
http://thread.gmane.org/gmane.comp.video.xine.devel/12479
CVS patchset: 7407
CVS date: 2005/02/13 14:24:25
|
|
CVS patchset: 7402
CVS date: 2005/02/12 15:13:30
|
|
CVS patchset: 7400
CVS date: 2005/02/12 09:21:32
|
|
see my last mail about this stuff:
http://thread.gmane.org/gmane.comp.video.xine.devel/12547
and this thread:
http://thread.gmane.org/gmane.comp.video.xine.devel/11213
CVS patchset: 7395
CVS date: 2005/02/09 20:03:18
|
|
CVS patchset: 7377
CVS date: 2005/02/02 23:11:38
|
|
CVS patchset: 7296
CVS date: 2004/12/22 22:05:22
|
|
appearing earlier in the plugin path actually override later plugins;
also check the priority of video_out plugins on loading for symmetry to the
audio_out loading
CVS patchset: 7289
CVS date: 2004/12/20 21:38:24
|
|
and backwards compatible translation
Sorry, I got a litte tired proof-reading the patch, so their might be
bugs lurking around. I will give it some further examination and
(as necessary) fixing tomorrow.
CVS patchset: 7233
CVS date: 2004/12/12 22:00:47
|
|
CVS patchset: 7170
CVS date: 2004/12/01 06:03:46
|
|
CVS patchset: 7131
CVS date: 2004/11/17 11:39:24
|
|
CVS patchset: 7119
CVS date: 2004/11/13 17:01:36
|
|
up a few minor compiler warnings
CVS patchset: 7071
CVS date: 2004/10/24 00:59:37
|
|
- new flags added at the end, so the users can override them properly
- removed X11 libs and cflags from xine-config and libxine.pc
- removed replacing for setenv (thanks to Robin for the hint)
- better way for temporary disabling of w32dll module
- data directories will be different on Windows
- force installing catalogs with translation into XINE_LOCALEDIR
- timer linked in xine-engine (just central place for linking such stuff)
- adjusted load_plugins and remove reduntant WIN32 conditionals
- removed unused $(lib_thr)
CVS patchset: 6927
CVS date: 2004/09/01 18:19:49
|
|
- HAVE_SETENV
- Cygwin build
- win32/contrib Makefile
CVS patchset: 6915
CVS date: 2004/08/28 20:03:42
|
|
CVS patchset: 6838
CVS date: 2004/07/22 14:27:48
|
|
* since plugin catalog assumes a limit on plugin count, we should actually
enforce this limit on plugin loading
* reduce DECODER_MAX from 256 to 128, because PLUGIN_MAX is already 256
and having only decoder plugins is unlikely
CVS patchset: 6738
CVS date: 2004/06/23 19:45:52
|
|
it right this time: use a different pointer for every decoder by putting the
descriptions into a dynamically growing chunk of heap memory
the name of the decoder in the description is necessary, because frontends
should not have to display the config entry key and this is currently the only
string to contain the decoder's name
CVS patchset: 6709
CVS date: 2004/06/19 19:48:42
|
|
pointers of the strings.
CVS patchset: 6693
CVS date: 2004/06/16 15:41:37
|
|
CVS patchset: 6684
CVS date: 2004/06/13 15:40:30
|
|
* use the info pointer before increasing it
* the plugin type is just a unit8_t, so (1<<8) is not a good constant
for PLUGIN_NO_UNLOAD, use (1<<6) instead; I hope 64 plugin types will be
enough...
CVS patchset: 6663
CVS date: 2004/06/09 17:31:08
|