Conversation
if build.with? "qt-dir" | ||
args << "-DCMAKE_PREFIX_PATH:PATH=#{ENV["HOMEBREW_QTGSTREAMER_QT"]}" | ||
else | ||
args << "-DCMAKE_PREFIX_PATH:PATH=#{Formula[build.with? "qt5" ? "qt5" : "qt"].prefix}" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
this line is failing; needs some parentheses
36895cc
to
59667b2
Compare
Ok, I have incorporated all suggestions. I also tried building against 'qt', but that failed, so I made qt-gstreamer depend on qt5. Thank you for your feedback. |
depends_on "gst-plugins-base" | ||
depends_on "boost" | ||
depends_on "doxygen" => :optional | ||
# does not build against 'qt', because moc trips over boost stuff. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Remove this comment, thanks.
Builds only against qt5, because moc trips over boost stuff when building against qt. Signed-off-by: Hannes Weisbach <hannes.weisbach@mailbox.tu-dresden.de>
59667b2
to
1a3f400
Compare
@BrewTestBot test this please |
There's still a lot of build failure here, even against
|
It works for me locally, so I don't know what I can do to fix that. It seems some Qt macros are not defined and stuff from libc++ cannot be found (wtf?). I will try to replicate the failures with --super-env. |
I’ve spent an ungodly number of hours trying to get qt-gstreamer to build on Mac OS (with no success) and now I bumped into this attempt at creating a homebrew formula for it. I see this pull is having the same errors I’m seeing in my own attempt (not using homebrew):
@hannesweisbach What is your environment and how did you get it to build locally? Would you mind posting the output of your For my project, having a homebrew formula for this would help tremendously. Thanks! |
@dogmaphobic: Yes it does build locally for me (with brew and without brew). However, I have not yet tried building with brew in superenv. I'm currently somewhat time constrained (plus qt-gstreamer works for me). I'll try to find time later today or over the weekend. |
@dogmaphobic: here is my build: http://pastebin.com/NMgqa8M9 |
@hannesweisbach, nah... Thanks for posting your build but there is no apparent significant difference. There is a minor difference with the cmake version, but when I started working on this, I was using the same version you used. The only other difference is the sequence of the build due to the different number of CPU cores. If there is a difference, it's in the environment, which is not reflected in either output. All visible build options are identical. This is quite baffling. I will continue looking into this. I have to. My current project depends on this to work in order to support Mac OS. Thanks again |
@dogmaphobic I've looked into the superenv-stuff and realised, that it is used by default. So now I'm at a loss. Does building qt-gstreamer from git work for you? |
No, I have the same exact errors with either build type. Not using brew at all (local gstreamer, qt, etc) or using all brew, I get the same errors. One of them I can understand. It's a conflict related to using exoteric c++11 features and how they differ between gcc and clang. The other is a total mystery for now. cmake fails to export some defines. QTGLIB_EXPORT in particular. |
QTGLIB_EXPORT is not supposed to be exported by CMake, it's defined in src/QGlib/global.h in qt-gstreamer. It depends on QtGLib_EXPORTS or Qt5GLib_EXPORTS, but it should be #defined either way. |
Sorry, I copied/pasted the wrong version (all uppercase), but the problem remains :) It's supposed to be defined in |
@dogmaphobic: I see. src/QGlib/CMakeFiles/Qt5GLib.dir/flags.make for me contains the CXX_DEFINES:
And for you I don't see how that would cause
So either way, That said, I couldn't figure out for the live of me where |
Any update on this? |
I have not heard back from the qt-gstreamer people. I cannot reproduce these build-errors locally. |
Let's try once more otherwise we might just need to pass on this 😭 |
@BrewTestBot test this please |
1 similar comment
@BrewTestBot test this please |
Still very much refusing to build on the CI so I guess we'll have to pass for now 😢. If you hear back from upstream at all & they can shed some light on the situation we'd be happy to review a fresh PR. Thanks for the submission ❤️ |
Signed-off-by: Hannes Weisbach hannes.weisbach@mailbox.tu-dresden.de