Re: [Plplot-general] [msys2, qt] no rule to make target moc_compilation.cpp

[Thomas G. <t_g...@gm...>]

Hi Alan,

thanks for your help. Unfortunately, still the same problem. In short:

`make plplotqt_autogen` complains that there is no rule to make the
target.(I'm assuming you meant plplotqt_autogen, at least that's the one
that creates moc_compilation.cpp on my linux)

The log files are attached. The terminal session is as follows

    thomas@valhalla MINGW64 /c/Users/thomas/Downloads/plplot-5.12.0
    $ rm -rf build && mkdir build && cd build
    $ cmake .. -G 'MSYS Makefiles' -DCMAKE_VERBOSE_MAKEFILE=ON >& cmake.out
    $ make VERBOSE=1 >& make.out
    $ make VERBOSE=1 plplotqt_automoc >& plplotqt_automoc.out
    $ make VERBOSE=1 plplotqt_autogen >& plplotqt_autogen.out
    $ make VERBOSE=1 plplotqt >& plplotqt.out

Same errors when using `Unix Makefiles`.

Best, Thomas


PS:

+ freshly extracted plplot 5.12.0
+ newly installed + upgraded MSYS2 x86_64
+ mingw-w64-x86_64-{cmake,qt4,base-devel,gcc,gcc-fortran}


Alan W. Irwin wrote on 05/18/2017 10:45 PM:
> On 2017-05-18 08:41+0200 Thomas Gläßle wrote:
>
>> Hi Alan,
>>
>> thanks for your response.
>>
>> Unfortunately, the default configuration fails with the same error (see
>> attached files).
>
> Hi Thomas:
>
> Thanks for your bug report with all the useful file results to help me
> diagnose the issue.  I don't see any obvious Qt-related issues
> with your cmake.out result, and it appears to me from your make.out
> file there is some Qt issue with dependencies in the automoc generated
> files. But that is a puzzling result since there is no such issue on
> the Linux platform.
>
> For example, I just did the following test on that platform starting
> from a freshly configured (with CMake-3.8.1 to match your result as
> closely as possible) build tree.
>
> software@raven> make VERBOSE=1 plplotqt >& plplotqt.out
> software@raven> grep Built plplotqt.out
> [  0%] Built target plplotqt_automoc
> [ 12%] Built target plhershey-unicode-gen
> [ 12%] Built target plhershey-unicode.h_built
> [ 12%] Built target csirocsa
> [ 25%] Built target csironn
> [ 25%] Built target deltaT-gen
> [ 37%] Built target deltaT.h_built
> [ 37%] Built target tai-utc-gen
> [ 37%] Built target tai-utc.h_built
> [ 50%] Built target qsastime
> [100%] Built target plplot
> [100%] Built target plplotqt
>
> Notice that the plplotqt_automoc target is automatically built before
> the plplotqt
> target, but there is no sign of that happening in your own make.out
> file.
>
> Just to confirm that conclusion, would you follow the
> above test there starting from a freshly configured build tree?
> And if you confirm that plplotqt_automoc is not built in your case
> (which should lead to the same error you found before), could
> you try imposing the build of the plplotqt_automoc target beforehand,
> i.e.,
>
> make VERBOSE=1 plplotqt_automoc >& plplotqt_automoc
>
> before
>
> make VERBOSE=1 plplotqt >& plplotqt.out
>
> to see if that works around that dependency issue?
>
> Of course, it is difficult to debug such issues at second hand from a
> different platform so I am hoping PLplot developers here with access
> to the platform will try the above test for themselves to (a) confirm
> the issue, and (b) dig deeper into why it exists just on that
> platform.
>
> Alan
> __________________________
> Alan W. Irwin
>
> Astronomical research affiliation with Department of Physics and
> Astronomy,
> University of Victoria (astrowww.phys.uvic.ca).
>
> Programming affiliations with the FreeEOS equation-of-state
> implementation for stellar interiors (freeeos.sf.net); the Time
> Ephemerides project (timeephem.sf.net); PLplot scientific plotting
> software package (plplot.sf.net); the libLASi project
> (unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
> and the Linux Brochure Project (lbproject.sf.net).
> __________________________
>
> Linux-powered Science
> __________________________
>