All,
To be clear on Bill's point(s) below -- there are no automated build
scripts (at least that I am aware of) in the WSJT-X repository; that is
why I created JTSDK in the first place.
I believe, though I'd have to verify it, in JTSDK v2.0.19 and above,
clean, rcfg and separate *are disabled*
On 15/04/2016 00:00, Richard Bown wrote:
> So wouldn't it make sense to put in the automated build scripts a make clean
> before configure ???
> or at least before make.
Hi Richard,
the whole point of build tools like make and tools like CMake that can
generate Makefiles is to do minimal
OK Bill, noted
but if you are running the jtsdk scripts, gui, there is no, well if there is I
didn't see it,
anyway to do a make clean.
If I'm manually building something I always do a make clean , but with jtsdk
once the selection is
made it just runs through until finished or falls over during
Hi Richard,
See my last post [1] and the release notes [2]
[1] https://sourceforge.net/p/wsjt/mailman/message/35014724/
[2]
https://sourceforge.net/projects/jtsdk/files/linux/2.0.0/documentation/release-notes/jtsdk-nix-2.0.19.html/download
On 04/14/2016 05:00 PM, Richard Bown wrote:
> OK
Hi Richard,
See below.
>
> What does jtsdk look for when there is already an earlier version of wsjtx
> built ?
> maybe something which will give more problems in the future Greg.
There are various JTSDK v2.0.20 options that affect rebuilds after a
large System Wide update as you
On 14/04/2016 23:30, Richard Bown wrote:
> What does jtsdk look for when there is already an earlier version of wsjtx
> built ?
> maybe something which will give more problems in the future Greg.
Hi Richard,
The underlying CMake build script is designed to detect source file
changes and
Hi Greg
I've just nuked /jtsdk/wsjtx/*
and built wsjtx, without errors about fortran/ccx errors
or the multiple errors like :-
/usr/include/arm-linux-gnueabihf/qt5/QtGui/qopenglfunctions.h: In member
function 'void QOpenGLFunctions::glVertexAttrib4fv(GLuint, const GLfloat*)':
On 14/04/2016 22:53, Richard Bown wrote:
> Verifying Fortran/CXX Compiler Compatibility failed with the following output:
>
> thats new today:(
> pity it does give any output in the log file
Hi Richard,
the compiler tests are logged, there should be more output that what you
listed. Can you run
Hi Richard,
See below.
On 04/14/2016 03:53 PM, Richard Bown wrote:
> OK Greg
> qtmultmedia5-dev back in , its now deciding to fail:_
> Verifying Fortran/CXX Compiler Compatibility failed with the following output:
Bill may have to assist on this one, as I am not sure sure how CMake
checks
OK Greg
qtmultmedia5-dev back in , its now deciding to fail:_
Verifying Fortran/CXX Compiler Compatibility failed with the following output:
thats new today :(
pity it does give any output in the log file
gfortran 5 is loaded as well as gfortran4
qt5-qmake is the only qmake installed
On Thu,
Hi Richard
Bill will be able to answer better than I, but the lib(s) you listed
below look to be runtime libs. Purging qt5* more that likely nuked the
dev libs, which may be why the build fails early now.
>From my Debian JTSDK Control file, the following are the build deps:
libqt5opengl5-dev
Hi
I've done a apt-get purge qt5*
then reinstalled qt5-default, 5.4.2,
running jtsdk now fails very early
-- hamlib_STATIC_LIBRARIES: /home/odroid/jtsdk/hamlib3/lib/libhamlib.a;m
CMake Error at CMakeLists.txt:743 (find_package):
By not providing "FindQt5Multimedia.cmake" in CMAKE_MODULE_PATH
On 14/04/2016 20:13, Richard Bown wrote:
> very simple reason,
> odroids use the Mali GPU and hence Opengl ES, its been passed back to
> Hardkernel, as the
> OpenGL ES compatibility issues are theirs & ubuntu's
Hi Richard,
do you know if the problem is missing functions in the OpenGL headers or
Hi Bill very simple reason,
odroids use the Mali GPU and hence Opengl ES, its been passed back to
Hardkernel, as the
OpenGL ES compatibility issues are theirs & ubuntu's, and absolutely no
interest, mail read but not
even acknowledged.
So no interest, probably too busy tied up with the Ubuntu
Hi Richard,
By WSPR, I was referring to WSJT-X >> Mode WSPR, sorry for the confusion.
Alot of WSPR users are converting over to WSJT-X >> Mode WSPR for a
number of reasons, primarily due to the much improved decoder from Steve
and Joe.
Jon Ove (LA3JJ) (and many other WSPR users) do allot of
On 14/04/2016 18:49, Richard Bown wrote:
> I suspect that with QT5 and OpenGL ES out of sync that this is the end of use
> of Odroid SBCs
> with WSJTX
Hi Richard,
I do not see how you come to that conclusion. From your previous message
I gather that there is an issue on armhf Ubuntu with
wspr and hamlib build with jtsdk, wsjtx no longer builds.
I dont have any interest in wspr, but I do use JT4
The only pre built binary is 1.1 , very old .
As for ubuntu 16.04 , dont hold your breath, its a disaster on the 64 bit C2.
I cant see any of the Odroid s/w moving to Ubuntu 16 until they
I'm running r6616 on an RPi V3. Decodes real well on JT65, JT9, and WSPR-2. It
is connected to an OmniaSDR rig. Currently having an intermittent issue with
transmitting, but Bill is working on it with me.
73, Jim / W6JHB
From: KI7MT
To: WSJT software development
Hi Richard,
I've not built WSJT-X on my RPI2 for a few weeks, so I'm nut sure about
r6370, but the many of the folks on WSPRNet use these small SBC devices
for WSPR. It may be worth posting there for further assistance.
Link: http://wsprnet.org
As a side note, Ubuntu 16.04 is approaching, maybe
Hi
Is anyone using either a Odroid C1,C2 or XU4 running WSJTX built since V6370
on any other distribution apart from Ubuntu 15.03 ???
It no longer builds on Ubuntu 15 , has anyone had any success with Fedora, Arch
, Debian
ect. ???
I suspect that with QT5 and OpenGL ES out of sync that this
On 14/04/2016 18:01, Jim Bennett wrote:
> pi@raspberrypi:~/~hamlib-3/build $ rigctl -m2517 F 14076000 f
> 14075999
Hi Jim,
the one above, the second command seems wrong. Not sure why yet.
The extra diagnostic will only show when you add -v to the command
options to get the verbose trace
OK - got it done. Didn't see anything returned other than the frequency that
was requested, though. I ran the command on an even frequency and an odd one.
Also ran the plain "rigctl" command and you can compare the output result, here:
pi@raspberrypi:~/~hamlib-3/build $ tests/rigctl -m2517 F
22 matches
Mail list logo