Hi Alan and Arjen
My X problems were just because I updated Cygwin with X running so
some dlls couldn't upgrade until reboot - but I had too much stuff
open on my laptop to reboot just then.

If it helps I have the following in my .bashrc to ensure X11 is
running whenever I start a terminal

#start and connect to X
export DISPLAY=:0.0
XPROC=$(ps aux | egrep /usr/bin/XWin)
if [[ -z "$XPROC" ]]; then
cygstart xwin -multiwindow
fi
unset XPROC

Looking at it I must have copied it from somewhere as it is definitely
not my style of scripting, but I'm not sure where I got it from.
Anyhow it works well for me.

Phil


On 8 June 2015 at 19:20, Alan W. Irwin <ir...@beluga.phys.uvic.ca> wrote:
> On 2015-06-08 08:55-0000 Arjen Markus wrote:
>
>> See the new results. They seem almost perfect. Only the Ada problem is left 
>> at a superficial examination of the reports.
>
> Thanks very much for doing these two tests (with and without Ada).  I
> agree these results are the best Cygwin results (outside your DISPLAY
> issues where I think I have found a solution below) we have ever seen.
> For example, because you are now using lua-5.2, you are getting a
> clean Lua run-time result, e.g.,
>
> lua
>    Missing examples            :
>    Differing graphical output  :
>    Missing stdout              :
>    Differing stdout            :
>
> And you similarly have clean results for c++, f95, python, and tcl.  To see 
> all of these
> voluminous diff results for every configuration, I suggest you try
>
> grep -B1 -A3 "Missing examples" */output_tree/*.out |less
>
> If you run that command, you will notice that Python and Lua results are 
> missing for
> the static case, but that is by design; those bindings only work
> for the shared library case.
>
> With regard to the on-going Ada problems, thanks to your first report
> I have now tracked the issue down to an incomplete link command for
> the Ada library, i.e., my CMake Ada language support is incomplete on
> Cygwin.  I am going to take a quick look at that now, but if it looks
> like the fix is going to be complicated, I will likely put off working
> on it until after this forthcoming release has been made.
>
> I also am just in the process of revising Tcl/Tk linking to eliminate
> redundant linking of libtclmatrix, libplplottcltk, and libplplot when
> ENABLE_DYNDRIVERS=OFF.  (For that case all the libtclmatrix and
> libplplottcltk code is already in libplplot so the first two libraries should
> not even be built, and all links should be made solely to libplplot
> rather than those first two libraries.) I think this change is going
> to sort out some memory management issues we have historically had
> with interactive testing of Tcl/Tk on both Cygwin and Linux in the
> past for the ENABLE_DYNDRIVERS=OFF case.  We worked around those issues
> by dropping such testing, but I am also going to reinstate that testing
> to make sure this elimination of redundant linking cures all these
> problems.
>
> Meanwhile, so that you can start interactive testing again
> you have to sort out the DISPLAY issues that are
> currently prohibiting your interactive tests of Tcl and Tk.  For
> example, in the present results for cmake.out you have the following
> error/warning set of messages:
>
> -- TK_WISH = /bin/wish
> Application initialization failed: no display name and no $DISPLAY 
> environment variable
> Error in startup script: couldn't load file "/usr/bin/tk85.dll": No such file 
> or directory
>      while executing
> "load /usr/bin/tk85.dll Tk"
>      ("package ifneeded Tk 8.5.18" script)
>      invoked from within
> "package require Tk"
>      invoked from within
> "puts -nonewline [package require Tk]"
>      (file 
> "/cygdrive/d/plplot-svn/comprehensive_test_disposeable/shared/build_tree/CheckTK_VERSION.tcl"
>  line 1)
> -- Looking for Tk version with wish - not found
> -- WARNING: setting ENABLE_tk to OFF
>
> I did a google search for the terms
>
> <cygwin "Application initialization failed: no display name">
>
> and came up with
> <http://stackoverflow.com/questions/9393462/cannot-launch-git-gui-using-cygwin-on-windows>
>
> The most popular answer implies that since 2012 or so Cygwin made an
> important Tcl/Tk and X change so a different procedure must be used to
> get X working properly.  That answer and its discussion gives the
> exact details about what has to be done. I hope those instructions
> work for you (and also Phil who is also having trouble getting X to
> work on modern Cygwin).  If those instructions work for you could you
> please put them on our Wiki so all our Cygwin users can get access to
> X?
>
> 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
> __________________________
>
> ------------------------------------------------------------------------------
> _______________________________________________
> Plplot-devel mailing list
> Plplot-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/plplot-devel

------------------------------------------------------------------------------
_______________________________________________
Plplot-devel mailing list
Plplot-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/plplot-devel

Reply via email to