* Non-OpenGL drawing is not supported anymore
* Disable wiimote support
* I have tested on Fedora 22 with xf86-video-nv
Changelog:
SuperTuxkart 0.8.1
~~~~~~~~~~~~~~~~~~
* New track STK Enterprise
* Updated track The old mines
* Updated Lighthouse track
* Updated Zen Garden track
* New Soccer mode
* New Egg Hunt mode
* New karts Xue and Sara
* Updated Beastie kart
* Added Tutorial
* Added new Supertux difficulty
* New bubblegum shield weapon
* New Speeodmeter and nitro meter
* Add ability to filter addons
* Updated nitro models
* Add ability to save and resume Grand Prix
* Improve skid marks and nitro effects
* Wiimote support
* Split mediaindo into devel/libzen, libmediainfo, and mediainfo.
Changelog:
Version 0.7.65, 2013-11-20
--------------
+ MXF: forcing detection of MPEG Video in case EssenceCompression is not present but MPEG2VideoDescriptor is present
+ GXF: detection of some captions and time codes event if they are not present at the beginning of the file (testing middle of the file)
+ DASH MPD: basic support
+ HDS F4M (Flash Media Manifest): basic support
+ DCP AssetMap (AM), PackageList (PKL) and CompositionPlaylist (CPL): basic support
+ IMF AssetMap (AM), PackageList (PKL) and CompositionPlaylist (CPL): basic support
+ Mac dylib: looking for the dylib in @executable_path and CFBundleCopyExecutableURL dir
+ AAC: option for instantaneous bitrate in fast detect mode (MediaInfoLib only)
+ FTP (custom builds only): support of UTF-8 file names
+ Colour description: colour_description_present added, better separation between bitstream values and container values
+ MPEG-4: RLE, color space and bit depth
+ Law rating: support of CEA-608 XDS Content Advisory in MPEG-PS, MPEG-Ts, LXF, GXF
+ MPEG-4/MOV: Bug found in one file, sample size is 16 with a 32-bit CodecID ("fl32"), correcting the output of MediaInfo
x #B775, AVI: AVI can use negative height for raw to signal that it's coded top-down, not bottom-up
x #B780, MPEG-TS: crash with some files having PAT/PMT change between begin and end of the file
x #B782, PBCore 1.2: some fields were not in the right order
x #B784, some humain readable strings were not removed when the corresponding field is removed
x #B787, MPEG-4/QuickTime: Erratic appereance of Bitrate Mode
x #B798: setlocale() remove from DLL
x #B785, DVCPRO HD: streams can be 8 or 10 bit, removing hard coded value from DV parser (MXF header value is used instead when applicable)
x MPEG-4: wrong demux of some E-AC-3 streams
x AAC: detection of HE-AACv2 was missing if the library is configured with fast detection
x MPEG Video: wrong computing of duration of raw stream in case of drop frame time code
x Automation, StreamKind type was set to integer, it is text
x MPEG-4: was reading lot of useless bytes from disk when the raw stream format is not known
x AVI: crash with some malformed text streams
x Reference/playlist files were not supported from FTP (custom builds only)
x MPEG-4/MOV: ScanOrder was using "stored" value instead of "displayed" value
x MXF: Detection of Dolby E was not working in some cases (regression in 0.7.62)
x MPEG-4/MOV: freeze with some files having mono 32-bit PCM
Version 0.7.64, 2013-07-05
--------------
+ New canonical URL of the website: http://MediaArea.net/MediaInfo
+ Portuguese, Basque and Russian translations updated
+ E-AC-3: support of streams having substreams (e.g. more than 6 channels)
+ JPEG: detection of YUVA, RGB, RGBA and YCCK color spaces
+ MPEG Audio: detection of Id3v1 inside an MPEG Audio frame
+ Matroska: support of HEVC/H.265 (based on specifications draft from DivX inc)
+ Canopus: detection of scan mode, scan order, pixel aspect ratio
+ MD5 generation option (work in progress)
+ Id3v2: reading of all Id3v2 blocks (no more only the first one) at the beginning of the file
+ MPEG-4: support of few additional iTunes tags, thanks to Kurtnoise
+ AVI: detection of Ut Video, thanks to Kurtnoise
+ MXF: detection of Dolby E even if EssenceCompression is not SMPTE 337
+ AIFF: detection of Dolby E
+ AIFF: detection of not aligned Dolby E
+ ARIB B24/B37: Caption_conversion_type display (HD, SD, Mobile)
+ MPEG-TS: displaying CEA-708 service and CEA-608 presence if the corresponding ATSC descriptor is present
+ MPEG-TS: quicker analysis in the case of quick parsing option
+ #F412, Matroska: Handling of files having no DocType
x #B761, MPEG-TS/ARIB: crash with some streams, thanks to maki
x #B765, Matroska: was parsing the complete file if a stream indicated in the header is not present
x #B763, MXF: detection of incoherency of channel count between bitstream and wrapper
x #B762, Matroska: detecting 23.976 frame rate (instead of 23.962 fps due to imprecise timestamp)
x #B759, QuickTime: detection of time codes having tcmi not embedded in tmcd box
x #B766, RMVB: a/v delay is incorrect, currently disabling it
x #B768, MPEG-4, crash with some MPEG-4 files
x #B769, MPEG-4, crash with some MPEG-4 files
x #B764, AVC: wrong parsing of some streams having pic_scaling_matrix
x AVC, crash with some AVC streams
x AAC, infinite loop with some AAC streams
x FLV, infinite loop with some FLV files
x Matroska: crash with some malformed files
x MOV: crash/long parsing with some malformed files
x AC-3: crashes and freezes fix
x Java 64-bit: Count_Get(StreamKind) was always returning 0
x Python 64-bit: Count_Get(StreamKind) was always returning 0
x DTS: some streams were not detected
x DTS: some memory leaks with 14-bit streams
x SMPTE ST 302: memory leaks
x SMPTE ST 337: memory leaks
x Total failure if MEDIAINFO_REFERENCES_NO #define was used
x QuickTime: Time code name is in "Title" field
x MPEG-4/QuickTime: handling of weird "negative" frame duration is stts
Version 0.7.63, 2013-05-13
--------------
+ Switched to a BSD-2-Clause license
+ LXF: AFD (from ancillary stream) support
+ Detection of some side car files and showing them as a single package (optional)
+ MOV: more channel positions information
+ TTML: detection
+ SAMI: detection
+ ID3: updated list of genres, thanks to Mats
+ .so: search a local copy of the library before trying default path
+ AVI: Support of Adobe PARf (Aspect Ratio) chunk
+ VC-3: Scan order
+ #P65, Flac/Ogg DISCTOTAL metadata support, thanks to Kurtnoise
+ #P67, MOV, add Hap Video to the database, thanks to Kurtnoise
+ #P67, Matroska, add VP9 to the database, thanks to Kurtnoise
+ #P68, FLV, add HEVC detection in the FLV parser
+ #P66, MOV, tref/chap handling, with chapters information, thanks to Kurtnoise
+ #P72, MOV, HEVC/H265 detection and basic support, thanks to Kurtnoise
+ #F382/P75, PMP format detection
+ MPEG-TS, HEVC/H265 support
+ 16:10 DAR known value
x #B742, MPEG-4/MOV, crash with some truncated/invalid files
x #B746, MPEG-4/MOV: crash with files having moof atom and no moov atom
x #B747, Inconsistent hang with buffer API
x #B757, MediaInfoDLL.py MediaInfoList was not working, thanks to Elbenfreund
x #B740, XML: dt:dt is replaced by dt (for binary data)
x MXF: Dolby E streams starting only at the second or third frame were not well detected
x MPEG-4/MOV: reducing analysis time for I-frame only video streams
x CEA-708: weird behavior phantom streams are displayed) in case of buggy CEA-708
x AVI: crash with audio delay and AvgBytesPerSec of 0
x Wrong demuxing of the caption stream in the case of AVC streams without frame_mbs_only_flag but having 2 fields in one slarsing was very slow with a sequence of files.
x MXF: Pre-charge duration was not read, time code of the first frame was wrong if "Origin" is not 0
x FLV: file was sometimes (e.g. most of the file padded with zeroes) fully parsed
x VC-3: Using values from specifications (based on compression ID) instead of SBD/SST/SPL/APLF from bitstream
x #738, mediainfo-qt doesn't build
* Convert to build from source to configure the localtion of configuration.xml.
Prebuild binary requires /var/opengrok directory. This package requires
the directory inside ${PREFIX}/share/opengrok only.
* Update HOMEPAGE
Changelog:
0.12-rc5:
differentiate between raw and download #578
add H A D links to directory listing #573
fix dtags.eftar file descriptor leak #535
do not unnecessarily create 2 File objects when returning eftar reader #539
0.12-rc4:
highlight line number with anchor (#702)
timeout for command execution (#429, #487)
fix for definitions search displays comment lines (#294)
0.12-rc3:
Lucene 4.6.0
Mercurial rename regression fix (#683)
let sendToConfigHost body use the parameter "env" to keep align with its signature and other methods
0.12-rc2:
bug #672 need double-quotes to specify path
bug #625 xref links containing \u don't seem to be working
follow-up fixes for Mercurial renamed files handling (fixes#666)
Support for Surround SCM history
lucene 4.5 (requires full reindex from 0.12-rc1)
generate history of files renamed in Mercurial repo separately (fixes#22)
0.12-rc1:
JDK7 + tomcat7 tested, JDK6 unsupported!
lucene 4.4 -> A LOT faster engine, no spellindex needed, regexp search supported (see help)
pl/sql, scala, uuencode languages/analyzers support
php analyzers improved
monotone, mercurial SCM support improved
AccuRev SCM added to supported SCMs
new config options in OpenGrok script
new search by analyzer/language
Changelog for <0.12 is not available.
Changelog:
JFlex 1.4.3 (Jan 31, 2009) fixes all known bugs of 1.4.2
Fixed bug #2018299 (lookahead syntax error).
Fixed bug #2010261 (min_int in Java example scanner).
Fixed bug #2007221 (zzEOFDone not reset in yyreset(Reader)).
Fixed bug #1999301 (%type and %int at the same time should produce error message).
JFlex 1.4.2 (May 27, 2008) fixes all known bugs of 1.4.1
Implemented feature request #1212181: Now supports generics syntax for %type, %extends, etc.
Implemented feature request #1762500: Provided %ctorarg option to add arguments to constructor.
Fixed bug #1464525 (Reader.read might return 0).
Fixed bug #1968897 (Ambiguous error message in macro expansion).
Fixed bug #1832973 (Syntax error in input may cause NullPointerException).
Fixed bug #1629920 (Need to defend against path blanks in jflex bash script).
Fixed bug #1540228 (EOF actions may be ignored for same lex state).
Fixed bug #1498726 (syntax error in generated ZZ_CMAP).
Fixed bug #1282840 (lookahead and "|" actions).
Fixed bug #1164852 (yytext() longer than expected with lookahead).
Fixed bug #1065521 (OS/2 Java 1.1.8 Issues).
Fixed bug #421495 (dangerous lookahead check may fail).
HTTP request/response parser for Python compatible with Python 2.x (>=2.6),
Python 3 and Pypy. If possible a C parser based on http-parser from Ryan Dahl
will be used.
NetBSD make library. Bump version (to 20131207, which was when I wrote
this.) Fixes the build of multimedia/bsdav.
Note: this is not a merge from NetBSD HEAD, but a minimal
reimplementation of the feature to get the job done.
These make templates really need to be merged with or replaced by the
latest versions from NetBSD HEAD. Anyone embarking on such a merge
should probably revert this change (which is only to bsd.prog.mk)
before doing so.
No objection after 3+ weeks on tech-pkg.
Version 4.12
============
Volker Quetschke (vq@openoffice.org)
#i83540# Add parallel building mode, see the MULTI PROCESSING section in
the dmake man page, for native W32 dmake (MSVC 6, and newer and MinGW)
versions.
Version 4.11
============
Yuri Dario (ydario@openoffice.org)
#i81855, CWS os2port01# Port dmake to OS/2.
Volker Quetschke (vq@openoffice.org)
#i81252# Add a new control macro SHELLCMDQUOTE that can be used to add
additional characters before and after the command string if the recipe is
executed using a shell. This macro is used to avoid problems with quoting
when using cmd.exe or command.com.
Volker Quetschke (vq@openoffice.org)
#i81296, i64572# Improve dependency handling of prerequisites.
Volker Quetschke (vq@openoffice.org)
#i80687, i66751# Allow the use of .PHONY for %-targets.
Volker Quetschke (vq@openoffice.org)
#i80010, i80009# Fix handling of quoted filenames for :1 :b :d :e :f
and :i macro expansions.
Volker Quetschke (vq@openoffice.org)
#i78776# Add a new function macro $(normpath[,para] data) and a new macro
extension to normalise the given data or macro content. The normalization
is done element-wise (whitespace separated) on the given data. Quotes can
be used to normalize filenames that contain whitespace characters.
On cygwin the result honors the setting of .WINPATH to determine the output
format. If the optional parameter para is given for the function macro its
expanded value is used to override the .WINPATH setting for the output.
Version 4.10 (skipped)
======================
This number was skipped to avoid confusion with Version 4.1, patch level 0
which identified itself as "Version 4.10, PL 0".
Version 4.9
===========
Volker Quetschke (vq@openoffice.org)
#i74700# Add a new special macro OOODMAKEMODE that is used to toggle
OOo build specific behavior. If OOODMAKEMODE is set (i.e. it begins with y)
a possible leading ./ of a path will no longer be removed during target
normalization.
Version 4.8
===========
Volker Quetschke (vq@openoffice.org)
#i73499# Add new .WINPATH attribute to generate Windows style paths (with
regular slashes) instead of the default cygwin style (POSIX) paths for
dmake's dynamic macros.
This attribute is specific for cygwin dmake executables and non-cygwin
environments ignore this attribute. This feature can be used to create DOS
style path parameters for native W32 programs from dynamic macros.
The windows style paths use regular slashes ('/') instead of the usual
windows backslash ('\') as directory separator to avoid quoting problems
(It still is a cygwin dmake feature!) and cygwin as well as native windows
programs should have no problems using this (c:/foo/bar) path
representation.
The affected macros are $@, $*, $>, $?, $<, $&, $^ and $(PWD), $(MAKEDIR)
and $(TMD).
Examples:
Content of $(PWD) without .WINPATH (default): /cygdrive/c/temp
Content of $(PWD) with .WINPATH set: c:/temp
Version 4.7
===========
Volker Quetschke (vq@openoffice.org)
#i71422# Add @@ as a new recipe line property. This new property is a
stronger version of the already existing @ recipe property. The recipe
line and the output (stdout and stderr) of the executed recipe are suppressed
and *NOT* shown on the terminal.
Note: Previously the second @ sign was ignored, that means that using this
new feature doesn't break backwards compatibility but the command output
is naturally not suppressed in older versions.
Volker Quetschke (vq@openoffice.org)
#i70210# Dmake now works correctly with empty dynamic prerequisites.
Volker Quetschke (vq@openoffice.org)
#i70168# Improve dmakes handling of literal $ characters in target and
prerequisite filenames. This was not working before and is working now
but the use of literal $ characters in filenames is still strongly
discouraged. See the man page for details.
Volker Quetschke (vq@openoffice.org)
#i69742# Normalize the pathnames of targets and prerequisites to the shortest
(or most natural, see below for the cygwin case) representation.
Now "./foo", "bar/../foo" and foo are recognized as the same file.
For cygwin this allows the use of target and prerequisite names with DOS
like pathnames with drive letters (with / and \ as possible path seperators)
and cygwin POSIX pathnames. As the internal format uses cygwin's natural
POSIX representation runtime macros will result in POSIX paths.
Version 4.6
===========
Volker Quetschke (vq@openoffice.org)
#i67589# Implement the internal dmake commands noop and echo. These commands
were already existing for the native Windows version but only partially
documented.
Volker Quetschke (vq@openoffice.org)
#i67166# Issue a warning when a previously unset macro (i.e. assumed to be
empty) is used and later set to a different value.
Volker Quetschke (vq@openoffice.org)
#i66567# Issue a warning when .SUFFIXES is used. This special target had no
special meaning (anymore?) and was always (at least since version 4.2)
ignored.
Volker Quetschke (vq@openoffice.org)
#i67025# Fix a bug that was introduced in dmake 4.5 and led dmake to falsely
issue an error if a %-target was defined with the .SILENT attribute and dmake
was started with the -vr (included in -v) verbose flag.
Version 4.5
===========
Volker Quetschke (vq@openoffice.org)
#i66385# Fix the handling of .<suffix> targets. Always treat these targets
as meta targets regardless of the AUGMAKE (-A) status. Previously this was
enabled when the AUGMAKE was *not* set to 'yes' or the -A switch was given.
#i66083# We now stop with an error if AUGMAKE meta targets and regular
targets are used in one rule definition. As the syntax of these target
definitions is slightly different this hints to an error in the makefile,
most probably a misspelled attribute or special target.
#i48087#, #i64869# The parsing of %-target definitions was fixed and
enhanced. Prior to dmake 4.5 only one %-target per target-definition worked
reliably, now this works well, but to help projects that still use older
versions a warning was added.
It's not very nice to issue a warning for something that works now but
didn't before. This was only done to aid development and will probably
be removed in future versions.
There was, and still is, that for a bug in dmake for %-targets with more
than one (not indirect) prerequisite only the first prerequisite is used.
This will be fixed in future versions but until this happens we issue a
warning that prerequisites are ignored.
#i65122# The :| rule operator was handled differently from what the man
page said. The man page was adapted to reflect reality. Changing dmakes
behavior to match the man page would have introduced a lot of compatibility
issues.
Version 4.4
===========
2006/02/03 Volker Quetschke (vq@openoffice.org)
#i61170# Fix the TMD macro. Up to now the TMD macro appended a directory
separator if it ended in "..", but most of the time it was not set correctly
at all and/or without a trailing slash.
Note: If your application relies on TMD please check the dmake manual page
for its definition and fix your makefiles if needed.
2006/02/01 Volker Quetschke (vq@openoffice.org)
#i60948# Add -m option family to generate timing information for targets
and/or recipes.
2005/09/19 Volker Quetschke (vq@openoffice.org)
#i53148# Fix $(shell ...) macro picking up the output of earlier
recipe lines in parallel builds.
#i54766# Ensure that all previous recipe lines to a recipe
line containing a $(shell ...) macro have finished before executing
the macro in parallel builds.
2005/06/01 Volker Quetschke (vq@openoffice.org)
#i39248# Add a testsuite to test some dmake features (only for autotools
builds). Start the tests with "make check".
2005/04/21 Volker Quetschke (vq@openoffice.org)
#i47851# Increase the version number to 4.4-cvs and slightly rearrange the
output of "dmake -V".
Remove the hard-coded (in autotools builds) OS macro. It had the wrong value
to be usable with the startup/*.mk and is not mentioned in the man page.
Clean up configure.in and friends.
2005/04/20 Volker Quetschke (vq@openoffice.org)
#i43310# Fix percent (%) rules to recognize prerequisites that are defined
with the multiple target rule operator (::).
This implicitly fixes the similar problem for .<suffix>.<suffix> targets.
2005/04/16 Volker Quetschke (vq@openoffice.org)
#i46987# Fix the documentation of the $(mktmp ...) function macro. This is
not necessarily a new feature as the macro works this way since dmake was
imported into the OpenOffice.org cvs but mentioning it here might help users
of dmake upgrading from very old versions.
Add the "map escape codes" $(...:m) macro expansion that allows to expand
escape codes in macros. It can also be used to reproduce the former (pre
OpenOffice.org cvs) $(mktmp ...) behavior.
2005/04/10 Volker Quetschke (vq@openoffice.org)
#i43252# Introduce the macro ABSMAKECMD that's value for a native Windows
dmake executable (compiled with MS Visual C++ or MinGW) is the absolute
filename and NULL for other operating systems or dmake build with other
compilers.
If DMAKEROOT is not changed and only for native Windows dmake versions the
value of DMAKEROOT now defaults to "$(ABSMAKECMD:d)startup".
This allows to use the startup files from the startup directory from the
same location where the dmake.exe is located.
2005/04/08 Steve Hay (shay@openoffice.org)
#i43241# Fix build problems with VC++ (in non-OOo environment) and fix
the dmake makefiles so that dmake can rebuild itself (VC++ only).
2005/03/14 Volker Quetschke (vq@openoffice.org)
#i44959# Fix problems when a continued line is followed by an empty
or comment line.
2005/03/13 Volker Quetschke (vq@openoffice.org)
#i37053# Fix function macro problems with continued lines.
2004/11/23 Volker Quetschke (vq@openoffice.org)
#i37491# Predefine dmake macro containing one space (SPACECHAR).
2004/11/21 Volker Quetschke (vq@openoffice.org)
#i36027# Improve function macro parameter parsing.
Version 4.3
===========
2004/10/07 Volker Quetschke (vq@openoffice.org)
#i34746# Fix timestamp of phony targets with prerequisites.
2004/07/11 Volker Quetschke (vq@openoffice.org)
#i31255# Add verbose recipe echoing option (-vr).
2004/07/06 Volker Quetschke (vq@openoffice.org)
#i30779# Fix to allow dmake to echo recipe lines when inference is used
to create files that are to be read by .INCLUDE.
2004/04/01 Michael Meeks (mmeeks@openoffice.org)
#i27319#, #i27981# Don't handle environment variables that begin with a
'!', they are used by Cygwin internally and confuse dmake.
2004/01/09 Dan Williams (fa@openoffice.org)
#i17992# Allows the nesting of expressions and a much more complex syntax
for dmake. It is now possible to nest expressions with () and any
combination of || and &&.
Example:
.IF (("$(OS)"=="MACOSX"&&"$(COM)"=="GCC")||"$(OS)"=="LINUX") && "$(GUIBASE)"=="unx"
...
.ENDIF
2003/10/19 Ken Foskey (waratah@openoffice.org)
#i20504# Correct processing of .LIBRARY targets.
2003/10/08 Chris Halls (haggai@openoffice.org)
#i7800# Fix tempnam security problem.
2003/03/08 Volker Quetschke (vq@openoffice.org)
#i10910# Use the build-in tempnam function for MS VC6.
2003/03/05 Volker Quetschke (vq@openoffice.org)
#i11861# Switch to newer autotools autoconf 2.5x and automake 1.5 and
add MinGW and MSVC to the configure buildable targets.
2002/10/11 Ken Foskey (waratah@openoffice.org)
Fix the "lost a child" problem. (Part II)
2002/10/04 Ken Foskey (waratah@openoffice.org)
Rework the configure process.
2002/04/10 Hans-Joachim Lankenau (hjs@openoffice.org)
Don't search INCLUDEDIRS if the filename to be included starts at fs-root.
2001/05/29 Patrick Luby (pluby@openoffice.org)
Added forceful exit to avoid infinite loop that sometimes occurs when
there is a fatal error. Lost a child problem. (Part I)
Version 4.2
===========
(Note, the dmake -V output was not changed and still said
"Version 4.10, PL 0". This probably should have been 4.1)
2001/04/20 Martin Hollmichel (mh@openoffice.org)
First autoconfiscation of dmake.
2001/02/13 Hans-Joachim Lankenau (hjs@openoffice.org)
Fix for leaving tmpfiles when building group targets.
2000/09/22 Jens-Heiner Rechtien (hr@openoffice.org)
Initial import of dmake into OpenOffice.org cvs
before 2000/09/22 ?
Add smaller/greater arithmetic for .IF conditions.
1999/03/18 Hans-Joachim Lankenau (hjs@openoffice.org)
Implement boolean expression "or".
kamel.derouiche@gmail.com
The greenlet package is a spin-off of Stackless, a version of CPython that
supports micro-threads called "tasklets". Tasklets run pseudo-concurrently
(typically in a single or a few OS-level threads) and are synchronized with data
exchanges on "channels".
A "greenlet", on the other hand, is a still more primitive notion of
micro-thread with no implicit scheduling; coroutines, in other words. This is
useful when you want to control exactly when your code runs. You can build
custom scheduled micro-threads on top of greenlet; however, it seems that
greenlets are useful on their own as a way to make advanced control flow
structures. For example, we can recreate generators; the difference with
Python's own generators is that our generators can call nested functions and the
nested functions can yield values too. Additionally, you don't need a "yield"
keyword. See the example in tests/test_generator.py.
Greenlets are provided as a C extension module for the regular unmodified
interpreter.
Greenlets are lightweight coroutines for in-process concurrent programming.
