This port is no longer based on the Qt3 version of dvbcut plus a large
patch to make it build on Qt4 with the Qt3 emulation layer. A fork that
uses native Qt4 calls is used instead.
The generated files moc_progressstatusbar.cpp and moc_progresswindow.cpp
trigger an internal error in clang, if compiled with optimization. This
bug has been accepted as 18098 in the llvm bug tracker. To work around
this problem, the above mentioned two files are compiled with -O0 appended
to the compiler flages. The post-configure target in the ports Makefile is
to be removed, when a fixed version of clang is distributed in all FreeBSD
releases supported by the ports infrastructure.
This setting has not had any effect since r1.13 to bsd.cmake.mk, as it
was only useful when we supported FreeBSD < 7.
Approved by: avilla (mentor), portmgr (miwi)
Feature safe: yes
Qt4 support is based on the patch developed by Raik Bieniek (raik.bieniek
at googlemail dot com) as found in the SVN repository and has been cleaned
up for this port by me.
The Qt4 version of dvbcut.cpp is built without optimisation (the GUI is not
functional, else).
The affected ports are the ones with gettext as a run-dependency
according to ports/INDEX-7 (5007 of them) and the ones with USE_GETTEXT
in Makefile (29 of them).
PR: ports/124340
Submitted by: edwin@
Approved by: portmgr (pav)
sources, not the FreeBSD port), again. It is hard to catch up with all
the API changes ffmpeg sees every few weeks.
A similar patch will probably be required in other ports that depend on
ffmpeg and that use the same method to access encoders/decoders.
This port remains broken until the ffmpeg port installs the new header
(allcodecs.h), which seems to have been forgotten by the ffmpeg developers
(and until the ffmpeg port has added that file to its PLIST).
binary operator '>?' that is accepted by gcc-3.x but not by gcc-4.2.
I did not find the definition of this operator and do not know whether
it is an extension found in g++-3.x, but I assume that it is a maximum
value operator (a >? b) <==> max(a, b) and this patch implements this
operation explicitly instead of via the (unknown in g++-4.2) operator.
Submitted by: pointyhat via kris
function from the latest ffmpeg library, which has been declared
deprecated for some time). That function has been re-enabled in
the FreeBSD port of ffmpeg and the port does build again ...
Approved by: portmgr (pav)
function img_convert() which is a special case of the new sws_scale().
Since I need to prepare and test a patch to have avframe.ccp use the
new function, the port will remain broken for a few days ...
This version is announced as equivalent to SVN rev 25.
All patches are reverted to previous versions (i.e., most are removed).
This upgrade should also fix build errors that were caused by a recent
upgrade of to the ffmpeg version in the FreeBSD ports tree. The ffmpeg
developers publicly state, that they only want to support use of the
library from C sources (c99, specifically) and that it is planned to
drop C++ compatibility. This will affect dvbcut, which is writte in
C++. At that time, the port may want to use the old version of ffmpeg
that is distributed with dvbcut (but which may miss FreeBSD specific
patches).
by Rick C. Petty. This is achieved by patches that update all the affected
files (ignoring the "import" sub-directory, which is not used for the port)
to match the versions in SVN.
These patches do also include a fix for the Scons incompatibility that caused
the port to be broken for some time.
Thanks go to Rick C. Petty for bringing the SVN version to my attention and
his effort to get the port fixed.
