plplot-general — PLplot user discussion group

Re: [Plplot-general] Trying to build examples

[Alan W. I. <ir...@be...>]

On 2015-01-21 23:09+0100 Arnaud Darmont wrote:

> Alan,
>
> thank you for the detailed answer. Cmake is what i'm trying to use.
>
> As i understand this won't be easy and i fear that there will be deployment 
> difficulties on customer systems. My goal would be to distribute an 
> executable of the final software along with the source so that they can just 
> run it without any installation. Will in the end PLplot allow for this or is 
> it the wrong library?

I think your first priority should be to get the PLplot build to work
on either Cygwin or MinGW-w64/MSYS2.  Both of those platforms should
provide a full suite of external libraries that PLplot depends upon to
provide complete functionality.  Those external libraries include
libqhull and shapelib (both needed by the core PLplot library for full
functionality), libharu (needed for the pdf device driver), the
wxwidgets library (needed by the PLplot wxwidgets device driver), Qt5
(needed for the qt device driver), and the Pango/Cairo subset of GTK+
(needed for the cairo device driver).  Note, that qt and cairo are
currently our best device drivers, and one of our developers is
working hard to bring our wxwidgets device driver up to that same
level of quality.  I would avoid MinGW/MSYS because it supplies few
libraries and therefore your PLplot build there will generate an
extremely lite version of PLplot that doesn't have the capabilities
(e.g., none of the wxwidgets, qt, or cairo device drivers) provided by
the equivalent PLplot build on those other two Windows platforms.

If you like the PLplot capabilities you find on either Cygwin or
MinGW-w64/MSYS2, then I believe the answer to the deployment issue you
brought up above is to build your software package against a static
version of PLplot (which should be trivial to arrange) which in turn
has been built against static versions of the external libraries
(non-trivial to arrange). If you can do that, it virtually solves your
deployment issues since you don't have to worry about version
incompatibilities of ABI incompatibility issues with shared external
libraries your users may or may not have installed, i.e., no dll-hell
issues.

My understanding is that CMake does have a method to force linking
with static external libraries if they are available rather than
external shared libraries (its default preference).  But I have never
actually tried that myself, and you would likely have to ask on the
CMake list or google for how to do it.

In sum, my advice to solve the deployment issue is to abandon
MinGW/MSYS (due to lack of external libraries) and try either/both
MinGW-w64/MSYS2 or else Cygwin. If satisfied with one of those verify
it has static versions of the necessary external libraries that you
can install.  Then figure out how to convince CMake to preferentially
use static libraries for linking, and build a static version of PLplot
against those external libraries.  Then statically link your software with
that static PLplot for your deployment.

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
__________________________

Re: [Plplot-general] Trying to build examples

[Arnaud D. <ada...@ap...>]

Alan,

thank you for the detailed answer. Cmake is what i'm trying to use.

As i understand this won't be easy and i fear that there will be 
deployment difficulties on customer systems. My goal would be to 
distribute an executable of the final software along with the source so 
that they can just run it without any installation. Will in the end 
PLplot allow for this or is it the wrong library?

Thank you.




On 1/21/2015 8:56 PM, Alan W. Irwin wrote:
> On 2015-01-21 18:26+0100 Arnaud Darmont wrote:
>
>> Hi all,
>>
>> i'm just new to PLplot and today i successfully managed to install
>> MinGW, CMake and all the related stuff and then i compiled the DLL and A
>> files. It was not very easy but it seems that i got all files in the 
>> end.
>>
>> In a next step i tried to compile an example. I moved all .h files, .a
>> files, .dll files and the example's .c file into a test directory and
>> successfully compiled using "gcc -otest.exe lib...dll.a etc (the list of
>> all .a files)".
>>
>> When i run the file i get the error message that it couldn't find the
>> driver's directory.
>>
>> Note that i have built PLplot without the examples as i need to build my
>> own example separately in either Embarcadero C++ builder, devCpp or 
>> MinGW.
>>
>> I don't think i have any build issues, just need to find out what to do
>> with the drivers, where to locate them and so on. I have tried to move
>> the driver's directory into my projects directory without success. I
>> actually don't think that those files are the drivers. Should the
>> drivers be built separately? if yes, what is the cmake command to build
>> the drivers? Where will "make install" locate the drivers after i make
>> them with gcc. Why are they not compiler together with PLplot?
>>
>> I have found a discussion about the same issue on the mailing list but
>> the discussion ends without providing an answer.
>
> As you have discovered, attempting to collect everything you need can
> be error-prone so instead use the cmake option
> -DCMAKE_INSTALL_PREFIX=<wherever you want the install prefix to be>,
> and after cmake is run, then run the command "make install" to collect
> all necessary PLplot files (including libraries, drivers, etc.) in an
> organized way in the install tree with the prefix location you have
> specified.  Then change directories to
> $prefix/share/plplot5$version/examples (where $prefix is the above
> install prefix and $version refers to whatever version of plplot you
> have installed.  In that location you will find two build systems for
> the installed examples which should give you some ideas about the best
> way to build your own applications against the PLplot libraries.
>
> 1. The traditional build system for the installed examples is based on
> make + pkg-config, and you can find out more about it by looking into
> the Makefile files in that installed examples directory and
> subdirectories.  This build approach may or may not be suitable for
> your needs since it all depends on whether you have access to
> pkg-config which on Windows platforms is available for Cygwin, and the
> combination of MinGW-w64 + MSYS2, but not for MinGW alone (a quite
> different and much older project than MinGW-w64) or for the
> combination of MinGW + MSYS (where MSYS is a quite different and older
> project than MSYS2).  To find out more about the MSYS2 possibility
> (which automatically includes MinGW-w64), take a look at
> http://sourceforge.net/p/msys2/wiki/Home/.
>
> 2. The CMake-based build system for the installed examples is
> implemented by the CMakeLists.txt file you will see in the installed
> examples directory and files it refers to there. The CMake-based
> approach is the one I recommend for you if you have some basic
> knowledge of CMake logic (or are willing to spend a day or so learning
> such basic knowledge) since it should work on any platform.
>
> To see this CMake-based build system in action you should (important!
> since this leaves the installed examples tree in a desireable pristine
> condition) create an intially empty build directory, change to that
> directory, (important!) put $install/bin on your PATH environment
> variable, then execute
>
> cmake $install/share/plplot5$version/examples
>
> to configure the build of the installed examples.  Afterwards, you can
> build all examples by running, e.g.,
>
> make VERBOSE=1
>
> and you can also run some of those built examples using, e.g.,
>
> c/x00c -dev psc -o test.psc
>
> to confirm, for example, that the drivers are being found properly.
>
> At this stage, you have two further choices.  You could simply cut
> and paste from the make VERBOSE=1 results to figure out how to
> build your own code against PLplot or you could adapt the existing
> CMake-based build system for the installed examples (mostly by getting
> rid of large parts of it that are not relevant to your needs) to
> configure building your own examples against PLplot.
>
> 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
> __________________________
>
>

-- 
APHESA
Arnaud Darmont
CEO

Direct Tel: +32 (0)4 365 06 80
Mobile: +32 (0)472 643 620
General Tel: +32 (0)4 366 18 70
Fax: +32 (0)4 366 08 10

Rue de Lorcé, 39
B-4920 HARZE
BELGIUM

www.aphesa.com

Twitter: @Aphesa

Re: [Plplot-general] Trying to build examples

[Alan W. I. <ir...@be...>]

On 2015-01-21 18:26+0100 Arnaud Darmont wrote:

> Hi all,
>
> i'm just new to PLplot and today i successfully managed to install
> MinGW, CMake and all the related stuff and then i compiled the DLL and A
> files. It was not very easy but it seems that i got all files in the end.
>
> In a next step i tried to compile an example. I moved all .h files, .a
> files, .dll files and the example's .c file into a test directory and
> successfully compiled using "gcc -otest.exe lib...dll.a etc (the list of
> all .a files)".
>
> When i run the file i get the error message that it couldn't find the
> driver's directory.
>
> Note that i have built PLplot without the examples as i need to build my
> own example separately in either Embarcadero C++ builder, devCpp or MinGW.
>
> I don't think i have any build issues, just need to find out what to do
> with the drivers, where to locate them and so on. I have tried to move
> the driver's directory into my projects directory without success. I
> actually don't think that those files are the drivers. Should the
> drivers be built separately? if yes, what is the cmake command to build
> the drivers? Where will "make install" locate the drivers after i make
> them with gcc. Why are they not compiler together with PLplot?
>
> I have found a discussion about the same issue on the mailing list but
> the discussion ends without providing an answer.

As you have discovered, attempting to collect everything you need can
be error-prone so instead use the cmake option
-DCMAKE_INSTALL_PREFIX=<wherever you want the install prefix to be>,
and after cmake is run, then run the command "make install" to collect
all necessary PLplot files (including libraries, drivers, etc.) in an
organized way in the install tree with the prefix location you have
specified.  Then change directories to
$prefix/share/plplot5$version/examples (where $prefix is the above
install prefix and $version refers to whatever version of plplot you
have installed.  In that location you will find two build systems for
the installed examples which should give you some ideas about the best
way to build your own applications against the PLplot libraries.

1. The traditional build system for the installed examples is based on
make + pkg-config, and you can find out more about it by looking into
the Makefile files in that installed examples directory and
subdirectories.  This build approach may or may not be suitable for
your needs since it all depends on whether you have access to
pkg-config which on Windows platforms is available for Cygwin, and the
combination of MinGW-w64 + MSYS2, but not for MinGW alone (a quite
different and much older project than MinGW-w64) or for the
combination of MinGW + MSYS (where MSYS is a quite different and older
project than MSYS2).  To find out more about the MSYS2 possibility
(which automatically includes MinGW-w64), take a look at
http://sourceforge.net/p/msys2/wiki/Home/.

2. The CMake-based build system for the installed examples is
implemented by the CMakeLists.txt file you will see in the installed
examples directory and files it refers to there. The CMake-based
approach is the one I recommend for you if you have some basic
knowledge of CMake logic (or are willing to spend a day or so learning
such basic knowledge) since it should work on any platform.

To see this CMake-based build system in action you should (important!
since this leaves the installed examples tree in a desireable pristine
condition) create an intially empty build directory, change to that
directory, (important!) put $install/bin on your PATH environment
variable, then execute

cmake $install/share/plplot5$version/examples

to configure the build of the installed examples.  Afterwards, you can
build all examples by running, e.g.,

make VERBOSE=1

and you can also run some of those built examples using, e.g.,

c/x00c -dev psc -o test.psc

to confirm, for example, that the drivers are being found properly.

At this stage, you have two further choices.  You could simply cut
and paste from the make VERBOSE=1 results to figure out how to
build your own code against PLplot or you could adapt the existing
CMake-based build system for the installed examples (mostly by getting
rid of large parts of it that are not relevant to your needs) to
configure building your own examples against PLplot.

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-general] Trying to build examples

[Arnaud D. <ada...@ap...>]

Hi all,

i'm just new to PLplot and today i successfully managed to install 
MinGW, CMake and all the related stuff and then i compiled the DLL and A 
files. It was not very easy but it seems that i got all files in the end.

In a next step i tried to compile an example. I moved all .h files, .a 
files, .dll files and the example's .c file into a test directory and 
successfully compiled using "gcc -otest.exe lib...dll.a etc (the list of 
all .a files)".

When i run the file i get the error message that it couldn't find the 
driver's directory.

Note that i have built PLplot without the examples as i need to build my 
own example separately in either Embarcadero C++ builder, devCpp or MinGW.

I don't think i have any build issues, just need to find out what to do 
with the drivers, where to locate them and so on. I have tried to move 
the driver's directory into my projects directory without success. I 
actually don't think that those files are the drivers. Should the 
drivers be built separately? if yes, what is the cmake command to build 
the drivers? Where will "make install" locate the drivers after i make 
them with gcc. Why are they not compiler together with PLplot?

I have found a discussion about the same issue on the mailing list but 
the discussion ends without providing an answer.

Thanks for your help.

Arnaud.

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[Arjen M. <Arj...@de...>]

Hi John,



I was triggered by similar behaviour under MinGW, with the only difference the programmatic choice for a device. I tried this myself via one of the standard examples, but that worked without the problems you reported. A relief :). That means only MinGW is special.



Regards,



Arjen



From: John Baumgardner [mailto:jrb...@co...]
Sent: Tuesday, January 20, 2015 7:30 PM
To: Arjen Markus; John Baumgardner
Cc: plp...@li...; Alan W. Irwin
Subject: RE: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

Hi Arjen,

   I am sure that your assessment as to the two possible causes for the failure is correct.  Given that I deleted all the plplot files in /usr/local except those I loaded into /usr/local/plplot, it is now close to impossible for me to reproduce the conditions responsible for the failure.  I had been trying to get the plplot-5.9.10 version to work, and I suspect my application was accessing some of those files which I had failed to remove or overwrite.  In any case, it is a great relief to have things working now.

   Just to let you know, most of the graphics software I am using here I wrote in the early 1980's, almost 35 years ago, while at Los Alamos National Laboratory, where I utilized what  was called the Common Graphics System (CGS).  These graphics allow me to visualize results from a large 3D spherical geometry finite element program called TERRA used for modeling the dynamics of planetary mantles. Plplot has enabled me, with little additional effort, to continue to utilize these CGS-based graphics after CGS was retired about 15 years ago.  I want to express my earnest thanks to you for helping maintain and improve Plplot.

John

At 07:25 AM 1/20/2015, Arjen Markus wrote:


Hi John,



I am glad you were able to solve it. I am not entirely sure which step in the device selection process failed (there are two: get the device information and load the library that implements it). One of the two must have been working on an incomplete directory and by properly installing all the stuff it has been solved. Might still be worth looking into at some point J.



Regards,



Arjen




From: John Baumgardner [ mailto:jrb...@co...]
Sent: Tuesday, January 20, 2015 4:20 PM
To: Arjen Markus; John Baumgardner
Cc: plp...@li...<mailto:plp...@li...>
Subject: RE: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

Hi Arjen,

  I took the step of placing the bin, include, and lib folders in a directory named /usr/local/plplot and I removed the plplot files everywhere else in /usr/local.  I think it was this cleanup that fixed the problem.  The program now runs beautifully with the line 'call plsdev(wingcc)'.  Thanks so much for your help!

Kind regards,
John


At 12:06 AM 1/20/2015, Arjen Markus wrote:


Hi John,



I have seen this type of behaviour under MinGW. I still have to figure why this is happening there.



What happens if you leave out the call to plsdev and instead use the command-line option "-dev wingcc"? I assume you use the routine plparseopts like in all examples - that does the trick for me.



Regards,



Arjen



From: John Baumgardner [ mailto:jrb...@co...]
Sent: Monday, January 19, 2015 6:53 PM
To: Arjen Markus; Alan W. Irwin; John Baumgardner
Cc: plp...@li...<mailto:plp...@li...>
Subject: RE: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

Hi Alan and Arjen,

   Thanks for that piece of information about the f95 implementation not requiring the array dimensions.  With that awareness, I was able quickly to modify the f77 code to compile correctly.

   However, I am now having difficulties when I attempt to execute.  My executable is unable to find the device drivers.  Here is the message I get:

*** PLPLOT ERROR, ABORTING OPERATION ***
plInitDispatchTable: Could not open drivers directory, aborting operation
Requested device wingcc not available

Plotting Options:

Enter device number or keyword:

  At the top of my main program, as implied by this output, I have the line call plsdev('wingcc').

   I copied the plplot dll and drivers directories into the directory where I am trying to execute, hoping that might resolve the problem.  But it did not.  Any suggestions as to what I am missing?

Thanks,
John

At 01:44 AM 1/19/2015, Arjen Markus wrote:


Hi John,



> -----Original Message-----
> From: Alan W. Irwin [ <mailto:ir...@be...> mailto:ir...@be...]
> Sent: Sunday, January 18, 2015 2:49 AM
> To: John Baumgardner
> Cc: plp...@li...<mailto:plp...@li...>
> Subject: Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10
> libraries
>
> However, the good news is I found what I needed (circle.f and how you built it) to
> figure out this problem which turns out not to be specific to Cygwin.
>
> > $ compile.sh
> > circle.f:26.72:
> >
> >      call plline(n101,xx,yy)
> >
> > 1
> > Error: There is no specific subroutine for the generic 'plline' at (1)
>

This is a typical message relating to the available interfaces for plline. The thing is that with Fortran 95 and beyond you can define a single interface name to be applied to multiple implementations. The compiler looks at the actual argument list and decides to use the implementation that matches that list (by number of arguments and the type, kind and rank of each). If it can not find any matching implementation, it will respond with a message like the above.

For the Fortran 95 bindings we have used the features provided by that standard as much as possible. One is that arrays ?know? their size. So the routine plline queries the arrays xx and yy for their size and there is no need to pass the size explicitly. That method is actually rather error-prone.

If you want to pass a section of the array only, you can use:

Call plline( xx(1:10), yy(1:10) )

For instance.

> I verified that issue.  To solve it, I simply used the redacted form of
> subroutine/function call with the redundant dimension information dropped, i.e.,
>
> call plline(xx,yy)
>
> Note, it is quite a while since we have used the f77 interface so my original advice to
> you was incomplete about converting over to f95.  I should have also added to the
> "use plplot" advice that since we dropped f77, our f95 capabilities and API have
> evolved and, for example, we are now taking advantage of certain Fortran 95
> capabilities like knowing the redundant array dimensions.  So you have to use the
> redacted form (like above with redundant dimension information dropped) of all calls.
>
> When in doubt about exactly what the redacted form is, look at files in examples/f95/
> for working examples of all the calls to the PLplot Fortran binding API.
>

Regards,



Arjen
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[John B. <jrb...@co...>]

Hi Arjen,

    I am sure that your assessment as to the two 
possible causes for the failure is 
correct.  Given that I deleted all the plplot 
files in /usr/local except those I loaded into 
/usr/local/plplot, it is now close to impossible 
for me to reproduce the conditions responsible 
for the failure.  I had been trying to get the 
plplot-5.9.10 version to work, and I suspect my 
application was accessing some of those files 
which I had failed to remove or overwrite.  In 
any case, it is a great relief to have things working now.

    Just to let you know, most of the graphics 
software I am using here I wrote in the early 
1980's, almost 35 years ago, while at Los Alamos 
National Laboratory, where I utilized what  was 
called the Common Graphics System (CGS).  These 
graphics allow me to visualize results from a 
large 3D spherical geometry finite element 
program called TERRA used for modeling the 
dynamics of planetary mantles. Plplot has enabled 
me, with little additional effort, to continue to 
utilize these CGS-based graphics after CGS was 
retired about 15 years ago.  I want to express my 
earnest thanks to you for helping maintain and improve Plplot.

John

At 07:25 AM 1/20/2015, Arjen Markus wrote:

>Hi John,
>
>
>
>I am glad you were able to solve it. I am not 
>entirely sure which step in the device selection 
>process failed (there are two: get the device 
>information and load the library that implements 
>it). One of the two must have been working on an 
>incomplete directory and by properly installing 
>all the stuff it has been solved. Might still be 
>worth looking into at some point J.
>
>
>
>Regards,
>
>
>
>Arjen
>
>
>
>
>From: John Baumgardner [mailto:jrb...@co...]
>Sent: Tuesday, January 20, 2015 4:20 PM
>To: Arjen Markus; John Baumgardner
>Cc: plp...@li...
>Subject: RE: [Plplot-general] plwidth fails to 
>link under Cygwin against plplot5.9.10 libraries
>
>Hi Arjen,
>
>   I took the step of placing the bin, include, 
> and lib folders in a directory named 
> /usr/local/plplot and I removed the plplot 
> files everywhere else in /usr/local.  I think 
> it was this cleanup that fixed the 
> problem.  The program now runs beautifully with 
> the line 'call plsdev(wingcc)'.  Thanks so much for your help!
>
>Kind regards,
>John
>
>
>At 12:06 AM 1/20/2015, Arjen Markus wrote:
>
>
>Hi John,
>
>
>
>I have seen this type of behaviour under MinGW. 
>I still have to figure why this is happening there.
>
>
>
>What happens if you leave out the call to plsdev 
>and instead use the command-line option “–dev 
>wingcc”? I assume you use the routine 
>plparseopts like in all examples – that does the trick for me.
>
>
>
>Regards,
>
>
>
>Arjen
>
>
>
>From: John Baumgardner [ mailto:jrb...@co...]
>Sent: Monday, January 19, 2015 6:53 PM
>To: Arjen Markus; Alan W. Irwin; John Baumgardner
>Cc: 
><mailto:plp...@li...>plp...@li...
>Subject: RE: [Plplot-general] plwidth fails to 
>link under Cygwin against plplot5.9.10 libraries
>
>Hi Alan and Arjen,
>
>    Thanks for that piece of information about 
> the f95 implementation not requiring the array 
> dimensions.  With that awareness, I was able 
> quickly to modify the f77 code to compile correctly.
>
>    However, I am now having difficulties when I 
> attempt to execute.  My executable is unable to 
> find the device drivers.  Here is the message I get:
>
>*** PLPLOT ERROR, ABORTING OPERATION ***
>plInitDispatchTable: Could not open drivers directory, aborting operation
>Requested device wingcc not available
>
>Plotting Options:
>
>Enter device number or keyword:
>
>   At the top of my main program, as implied by 
> this output, I have the line call plsdev('wingcc').
>
>    I copied the plplot dll and drivers 
> directories into the directory where I am 
> trying to execute, hoping that might resolve 
> the problem.  But it did not.  Any suggestions as to what I am missing?
>
>Thanks,
>John
>
>At 01:44 AM 1/19/2015, Arjen Markus wrote:
>
>
>Hi John,
>
>
>
> > -----Original Message-----
> > From: Alan W. Irwin 
> [<mailto:ir...@be...> 
> <mailto:ir...@be...>mailto:ir...@be...]
> > Sent: Sunday, January 18, 2015 2:49 AM
> > To: John Baumgardner
> > Cc: 
> <mailto:plp...@li...>plp...@li...
> > Subject: Re: [Plplot-general] plwidth fails 
> to link under Cygwin against plplot5.9.10
> > libraries
> >
> > However, the good news is I found what I 
> needed (circle.f and how you built it) to
> > figure out this problem which turns out not to be specific to Cygwin.
> >
> > > $ compile.sh
> > > circle.f:26.72:
> > >
> > >      call plline(n101,xx,yy)
> > >
> > > 1
> > > Error: There is no specific subroutine for the generic 'plline' at (1)
> >
>
>This is a typical message relating to the 
>available interfaces for plline. The thing is 
>that with Fortran 95 and beyond you can define a 
>single interface name to be applied to multiple 
>implementations. The compiler looks at the 
>actual argument list and decides to use the 
>implementation that matches that list (by number 
>of arguments and the type, kind and rank of 
>each). If it can not find any matching 
>implementation, it will respond with a message like the above.
>
>For the Fortran 95 bindings we have used the 
>features provided by that standard as much as 
>possible. One is that arrays ?know? their size. 
>So the routine plline queries the arrays xx and 
>yy for their size and there is no need to pass 
>the size explicitly. That method is actually rather error-prone.
>
>If you want to pass a section of the array only, you can use:
>
>Call plline( xx(1:10), yy(1:10) )
>
>For instance.
>
> > I verified that issue.  To solve it, I simply used the redacted form of
> > subroutine/function call with the redundant 
> dimension information dropped, i.e.,
> >
> > call plline(xx,yy)
> >
> > Note, it is quite a while since we have used 
> the f77 interface so my original advice to
> > you was incomplete about converting over to 
> f95.  I should have also added to the
> > "use plplot" advice that since we dropped 
> f77, our f95 capabilities and API have
> > evolved and, for example, we are now taking advantage of certain Fortran 95
> > capabilities like knowing the redundant array 
> dimensions.  So you have to use the
> > redacted form (like above with redundant 
> dimension information dropped) of all calls.
> >
> > When in doubt about exactly what the redacted 
> form is, look at files in examples/f95/
> > for working examples of all the calls to the PLplot Fortran binding API.
> >
>
>Regards,
>
>
>
>Arjen
>DISCLAIMER: This message is intended exclusively 
>for the addressee(s) and may contain 
>confidential and privileged information. If you 
>are not the intended recipient please notify the 
>sender immediately and destroy this message. 
>Unauthorized use, disclosure or copying of this 
>message is strictly prohibited. The foundation 
>'Stichting Deltares', which has its seat at 
>Delft, The Netherlands, Commercial Registration 
>Number 41146461, is not liable in any way 
>whatsoever for consequences and/or damages 
>resulting from the improper, incomplete and 
>untimely dispatch, receipt and/or content of this e-mail.
>DISCLAIMER: This message is intended exclusively 
>for the addressee(s) and may contain 
>confidential and privileged information. If you 
>are not the intended recipient please notify the 
>sender immediately and destroy this message. 
>Unauthorized use, disclosure or copying of this 
>message is strictly prohibited. The foundation 
>'Stichting Deltares', which has its seat at 
>Delft, The Netherlands, Commercial Registration 
>Number 41146461, is not liable in any way 
>whatsoever for consequences and/or damages 
>resulting from the improper, incomplete and 
>untimely dispatch, receipt and/or content of this e-mail.
>DISCLAIMER: This message is intended exclusively 
>for the addressee(s) and may contain 
>confidential and privileged information. If you 
>are not the intended recipient please notify the 
>sender immediately and destroy this message. 
>Unauthorized use, disclosure or copying of this 
>message is strictly prohibited. The foundation 
>'Stichting Deltares', which has its seat at 
>Delft, The Netherlands, Commercial Registration 
>Number 41146461, is not liable in any way 
>whatsoever for consequences and/or damages 
>resulting from the improper, incomplete and 
>untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[Arjen M. <Arj...@de...>]

Hi John,



I am glad you were able to solve it. I am not entirely sure which step in the device selection process failed (there are two: get the device information and load the library that implements it). One of the two must have been working on an incomplete directory and by properly installing all the stuff it has been solved. Might still be worth looking into at some point :).



Regards,



Arjen




From: John Baumgardner [mailto:jrb...@co...]
Sent: Tuesday, January 20, 2015 4:20 PM
To: Arjen Markus; John Baumgardner
Cc: plp...@li...
Subject: RE: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

Hi Arjen,

  I took the step of placing the bin, include, and lib folders in a directory named /usr/local/plplot and I removed the plplot files everywhere else in /usr/local.  I think it was this cleanup that fixed the problem.  The program now runs beautifully with the line 'call plsdev(wingcc)'.  Thanks so much for your help!

Kind regards,
John


At 12:06 AM 1/20/2015, Arjen Markus wrote:


Hi John,



I have seen this type of behaviour under MinGW. I still have to figure why this is happening there.



What happens if you leave out the call to plsdev and instead use the command-line option "-dev wingcc"? I assume you use the routine plparseopts like in all examples - that does the trick for me.



Regards,



Arjen



From: John Baumgardner [ mailto:jrb...@co...]
Sent: Monday, January 19, 2015 6:53 PM
To: Arjen Markus; Alan W. Irwin; John Baumgardner
Cc: plp...@li...<mailto:plp...@li...>
Subject: RE: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

Hi Alan and Arjen,

   Thanks for that piece of information about the f95 implementation not requiring the array dimensions.  With that awareness, I was able quickly to modify the f77 code to compile correctly.

   However, I am now having difficulties when I attempt to execute.  My executable is unable to find the device drivers.  Here is the message I get:

*** PLPLOT ERROR, ABORTING OPERATION ***
plInitDispatchTable: Could not open drivers directory, aborting operation
Requested device wingcc not available

Plotting Options:

Enter device number or keyword:

  At the top of my main program, as implied by this output, I have the line call plsdev('wingcc').

   I copied the plplot dll and drivers directories into the directory where I am trying to execute, hoping that might resolve the problem.  But it did not.  Any suggestions as to what I am missing?

Thanks,
John

At 01:44 AM 1/19/2015, Arjen Markus wrote:


Hi John,



> -----Original Message-----
> From: Alan W. Irwin [ <mailto:ir...@be...> mailto:ir...@be...]
> Sent: Sunday, January 18, 2015 2:49 AM
> To: John Baumgardner
> Cc: plp...@li...<mailto:plp...@li...>
> Subject: Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10
> libraries
>
> However, the good news is I found what I needed (circle.f and how you built it) to
> figure out this problem which turns out not to be specific to Cygwin.
>
> > $ compile.sh
> > circle.f:26.72:
> >
> >      call plline(n101,xx,yy)
> >
> > 1
> > Error: There is no specific subroutine for the generic 'plline' at (1)
>

This is a typical message relating to the available interfaces for plline. The thing is that with Fortran 95 and beyond you can define a single interface name to be applied to multiple implementations. The compiler looks at the actual argument list and decides to use the implementation that matches that list (by number of arguments and the type, kind and rank of each). If it can not find any matching implementation, it will respond with a message like the above.

For the Fortran 95 bindings we have used the features provided by that standard as much as possible. One is that arrays ?know? their size. So the routine plline queries the arrays xx and yy for their size and there is no need to pass the size explicitly. That method is actually rather error-prone.

If you want to pass a section of the array only, you can use:

Call plline( xx(1:10), yy(1:10) )

For instance.

> I verified that issue.  To solve it, I simply used the redacted form of
> subroutine/function call with the redundant dimension information dropped, i.e.,
>
> call plline(xx,yy)
>
> Note, it is quite a while since we have used the f77 interface so my original advice to
> you was incomplete about converting over to f95.  I should have also added to the
> "use plplot" advice that since we dropped f77, our f95 capabilities and API have
> evolved and, for example, we are now taking advantage of certain Fortran 95
> capabilities like knowing the redundant array dimensions.  So you have to use the
> redacted form (like above with redundant dimension information dropped) of all calls.
>
> When in doubt about exactly what the redacted form is, look at files in examples/f95/
> for working examples of all the calls to the PLplot Fortran binding API.
>

Regards,



Arjen
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[John B. <jrb...@co...>]

Hi Arjen,

   I took the step of placing the bin, include, 
and lib folders in a directory named 
/usr/local/plplot and I removed the plplot files 
everywhere else in /usr/local.  I think it was 
this cleanup that fixed the problem.  The program 
now runs beautifully with the line 'call 
plsdev(wingcc)'.  Thanks so much for your help!

Kind regards,
John


At 12:06 AM 1/20/2015, Arjen Markus wrote:

>Hi John,
>
>
>
>I have seen this type of behaviour under MinGW. 
>I still have to figure why this is happening there.
>
>
>
>What happens if you leave out the call to plsdev 
>and instead use the command-line option “–dev 
>wingcc”? I assume you use the routine 
>plparseopts like in all examples – that does the trick for me.
>
>
>
>Regards,
>
>
>
>Arjen
>
>
>
>From: John Baumgardner [mailto:jrb...@co...]
>Sent: Monday, January 19, 2015 6:53 PM
>To: Arjen Markus; Alan W. Irwin; John Baumgardner
>Cc: plp...@li...
>Subject: RE: [Plplot-general] plwidth fails to 
>link under Cygwin against plplot5.9.10 libraries
>
>Hi Alan and Arjen,
>
>    Thanks for that piece of information about 
> the f95 implementation not requiring the array 
> dimensions.  With that awareness, I was able 
> quickly to modify the f77 code to compile correctly.
>
>    However, I am now having difficulties when I 
> attempt to execute.  My executable is unable to 
> find the device drivers.  Here is the message I get:
>
>*** PLPLOT ERROR, ABORTING OPERATION ***
>plInitDispatchTable: Could not open drivers directory, aborting operation
>Requested device wingcc not available
>
>Plotting Options:
>
>Enter device number or keyword:
>
>   At the top of my main program, as implied by 
> this output, I have the line call plsdev('wingcc').
>
>    I copied the plplot dll and drivers 
> directories into the directory where I am 
> trying to execute, hoping that might resolve 
> the problem.  But it did not.  Any suggestions as to what I am missing?
>
>Thanks,
>John
>
>At 01:44 AM 1/19/2015, Arjen Markus wrote:
>
>
>Hi John,
>
>
>
> > -----Original Message-----
> > From: Alan W. Irwin [ mailto:ir...@be...]
> > Sent: Sunday, January 18, 2015 2:49 AM
> > To: John Baumgardner
> > Cc: 
> <mailto:plp...@li...>plp...@li...
> > Subject: Re: [Plplot-general] plwidth fails 
> to link under Cygwin against plplot5.9.10
> > libraries
> >
> > However, the good news is I found what I 
> needed (circle.f and how you built it) to
> > figure out this problem which turns out not to be specific to Cygwin.
> >
> > > $ compile.sh
> > > circle.f:26.72:
> > >
> > >      call plline(n101,xx,yy)
> > >
> > > 1
> > > Error: There is no specific subroutine for the generic 'plline' at (1)
> >
>
>This is a typical message relating to the 
>available interfaces for plline. The thing is 
>that with Fortran 95 and beyond you can define a 
>single interface name to be applied to multiple 
>implementations. The compiler looks at the 
>actual argument list and decides to use the 
>implementation that matches that list (by number 
>of arguments and the type, kind and rank of 
>each). If it can not find any matching 
>implementation, it will respond with a message like the above.
>
>For the Fortran 95 bindings we have used the 
>features provided by that standard as much as 
>possible. One is that arrays ?know? their size. 
>So the routine plline queries the arrays xx and 
>yy for their size and there is no need to pass 
>the size explicitly. That method is actually rather error-prone.
>
>If you want to pass a section of the array only, you can use:
>
>Call plline( xx(1:10), yy(1:10) )
>
>For instance.
>
> > I verified that issue.  To solve it, I simply used the redacted form of
> > subroutine/function call with the redundant 
> dimension information dropped, i.e.,
> >
> > call plline(xx,yy)
> >
> > Note, it is quite a while since we have used 
> the f77 interface so my original advice to
> > you was incomplete about converting over to 
> f95.  I should have also added to the
> > "use plplot" advice that since we dropped 
> f77, our f95 capabilities and API have
> > evolved and, for example, we are now taking advantage of certain Fortran 95
> > capabilities like knowing the redundant array 
> dimensions.  So you have to use the
> > redacted form (like above with redundant 
> dimension information dropped) of all calls.
> >
> > When in doubt about exactly what the redacted 
> form is, look at files in examples/f95/
> > for working examples of all the calls to the PLplot Fortran binding API.
> >
>
>Regards,
>
>
>
>Arjen
>DISCLAIMER: This message is intended exclusively 
>for the addressee(s) and may contain 
>confidential and privileged information. If you 
>are not the intended recipient please notify the 
>sender immediately and destroy this message. 
>Unauthorized use, disclosure or copying of this 
>message is strictly prohibited. The foundation 
>'Stichting Deltares', which has its seat at 
>Delft, The Netherlands, Commercial Registration 
>Number 41146461, is not liable in any way 
>whatsoever for consequences and/or damages 
>resulting from the improper, incomplete and 
>untimely dispatch, receipt and/or content of this e-mail.
>DISCLAIMER: This message is intended exclusively 
>for the addressee(s) and may contain 
>confidential and privileged information. If you 
>are not the intended recipient please notify the 
>sender immediately and destroy this message. 
>Unauthorized use, disclosure or copying of this 
>message is strictly prohibited. The foundation 
>'Stichting Deltares', which has its seat at 
>Delft, The Netherlands, Commercial Registration 
>Number 41146461, is not liable in any way 
>whatsoever for consequences and/or damages 
>resulting from the improper, incomplete and 
>untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[Arjen M. <Arj...@de...>]

Hi John,



I have seen this type of behaviour under MinGW. I still have to figure why this is happening there.



What happens if you leave out the call to plsdev and instead use the command-line option "-dev wingcc"? I assume you use the routine plparseopts like in all examples - that does the trick for me.



Regards,



Arjen



From: John Baumgardner [mailto:jrb...@co...]
Sent: Monday, January 19, 2015 6:53 PM
To: Arjen Markus; Alan W. Irwin; John Baumgardner
Cc: plp...@li...
Subject: RE: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

Hi Alan and Arjen,

   Thanks for that piece of information about the f95 implementation not requiring the array dimensions.  With that awareness, I was able quickly to modify the f77 code to compile correctly.

   However, I am now having difficulties when I attempt to execute.  My executable is unable to find the device drivers.  Here is the message I get:

*** PLPLOT ERROR, ABORTING OPERATION ***
plInitDispatchTable: Could not open drivers directory, aborting operation
Requested device wingcc not available

Plotting Options:

Enter device number or keyword:

  At the top of my main program, as implied by this output, I have the line call plsdev('wingcc').

   I copied the plplot dll and drivers directories into the directory where I am trying to execute, hoping that might resolve the problem.  But it did not.  Any suggestions as to what I am missing?

Thanks,
John

At 01:44 AM 1/19/2015, Arjen Markus wrote:


Hi John,



> -----Original Message-----
> From: Alan W. Irwin [ mailto:ir...@be...]
> Sent: Sunday, January 18, 2015 2:49 AM
> To: John Baumgardner
> Cc: plp...@li...<mailto:plp...@li...>
> Subject: Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10
> libraries
>
> However, the good news is I found what I needed (circle.f and how you built it) to
> figure out this problem which turns out not to be specific to Cygwin.
>
> > $ compile.sh
> > circle.f:26.72:
> >
> >      call plline(n101,xx,yy)
> >
> > 1
> > Error: There is no specific subroutine for the generic 'plline' at (1)
>

This is a typical message relating to the available interfaces for plline. The thing is that with Fortran 95 and beyond you can define a single interface name to be applied to multiple implementations. The compiler looks at the actual argument list and decides to use the implementation that matches that list (by number of arguments and the type, kind and rank of each). If it can not find any matching implementation, it will respond with a message like the above.

For the Fortran 95 bindings we have used the features provided by that standard as much as possible. One is that arrays ?know? their size. So the routine plline queries the arrays xx and yy for their size and there is no need to pass the size explicitly. That method is actually rather error-prone.

If you want to pass a section of the array only, you can use:

Call plline( xx(1:10), yy(1:10) )

For instance.

> I verified that issue.  To solve it, I simply used the redacted form of
> subroutine/function call with the redundant dimension information dropped, i.e.,
>
> call plline(xx,yy)
>
> Note, it is quite a while since we have used the f77 interface so my original advice to
> you was incomplete about converting over to f95.  I should have also added to the
> "use plplot" advice that since we dropped f77, our f95 capabilities and API have
> evolved and, for example, we are now taking advantage of certain Fortran 95
> capabilities like knowing the redundant array dimensions.  So you have to use the
> redacted form (like above with redundant dimension information dropped) of all calls.
>
> When in doubt about exactly what the redacted form is, look at files in examples/f95/
> for working examples of all the calls to the PLplot Fortran binding API.
>

Regards,



Arjen
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.
DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[John B. <jrb...@co...>]

Hi Alan and Arjen,

    Thanks for that piece of information about the f95 implementation 
not requiring the array dimensions.  With that awareness, I was able 
quickly to modify the f77 code to compile correctly.

    However, I am now having difficulties when I attempt to 
execute.  My executable is unable to find the device drivers.  Here 
is the message I get:

*** PLPLOT ERROR, ABORTING OPERATION ***
plInitDispatchTable: Could not open drivers directory, aborting operation
Requested device wingcc not available

Plotting Options:

Enter device number or keyword:

   At the top of my main program, as implied by this output, I have 
the line  call plsdev('wingcc').

    I copied the plplot dll and drivers directories into the 
directory where I am trying to execute, hoping that might resolve the 
problem.  But it did not.  Any suggestions as to what I am missing?

Thanks,
John

At 01:44 AM 1/19/2015, Arjen Markus wrote:

>Hi John,
>
>
>
> > -----Original Message-----
> > From: Alan W. Irwin 
> [<mailto:ir...@be...>mailto:ir...@be...]
> > Sent: Sunday, January 18, 2015 2:49 AM
> > To: John Baumgardner
> > Cc: plp...@li...
> > Subject: Re: [Plplot-general] plwidth fails to link under Cygwin 
> against plplot5.9.10
> > libraries
> >
> > However, the good news is I found what I needed (circle.f and how 
> you built it) to
> > figure out this problem which turns out not to be specific to Cygwin.
> >
> > > $ compile.sh
> > > circle.f:26.72:
> > >
> > >      call plline(n101,xx,yy)
> > >
> > > 1
> > > Error: There is no specific subroutine for the generic 'plline' at (1)
> >
>
>This is a typical message relating to the available interfaces for 
>plline. The thing is that with Fortran 95 and beyond you can define 
>a single interface name to be applied to multiple implementations. 
>The compiler looks at the actual argument list and decides to use 
>the implementation that matches that list (by number of arguments 
>and the type, kind and rank of each). If it can not find any 
>matching implementation, it will respond with a message like the above.
>
>For the Fortran 95 bindings we have used the features provided by 
>that standard as much as possible. One is that arrays "know" their 
>size. So the routine plline queries the arrays xx and yy for their 
>size and there is no need to pass the size explicitly. That method 
>is actually rather error-prone.
>
>If you want to pass a section of the array only, you can use:
>
>Call plline( xx(1:10), yy(1:10) )
>
>For instance.
>
> > I verified that issue.  To solve it, I simply used the redacted form of
> > subroutine/function call with the redundant dimension information 
> dropped, i.e.,
> >
> > call plline(xx,yy)
> >
> > Note, it is quite a while since we have used the f77 interface so 
> my original advice to
> > you was incomplete about converting over to f95.  I should have 
> also added to the
> > "use plplot" advice that since we dropped f77, our f95 
> capabilities and API have
> > evolved and, for example, we are now taking advantage of certain Fortran 95
> > capabilities like knowing the redundant array dimensions.  So you 
> have to use the
> > redacted form (like above with redundant dimension information 
> dropped) of all calls.
> >
> > When in doubt about exactly what the redacted form is, look at 
> files in examples/f95/
> > for working examples of all the calls to the PLplot Fortran binding API.
> >
>
>Regards,
>
>
>
>Arjen
>DISCLAIMER: This message is intended exclusively for the 
>addressee(s) and may contain confidential and privileged 
>information. If you are not the intended recipient please notify the 
>sender immediately and destroy this message. Unauthorized use, 
>disclosure or copying of this message is strictly prohibited. The 
>foundation 'Stichting Deltares', which has its seat at Delft, The 
>Netherlands, Commercial Registration Number 41146461, is not liable 
>in any way whatsoever for consequences and/or damages resulting from 
>the improper, incomplete and untimely dispatch, receipt and/or 
>content of this e-mail.

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[Arjen M. <Arj...@de...>]

Hi John,



> -----Original Message-----
> From: Alan W. Irwin [mailto:ir...@be...]
> Sent: Sunday, January 18, 2015 2:49 AM
> To: John Baumgardner
> Cc: plp...@li...
> Subject: Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10
> libraries
>
> However, the good news is I found what I needed (circle.f and how you built it) to
> figure out this problem which turns out not to be specific to Cygwin.
>
> > $ compile.sh
> > circle.f:26.72:
> >
> >      call plline(n101,xx,yy)
> >
> > 1
> > Error: There is no specific subroutine for the generic 'plline' at (1)
>


This is a typical message relating to the available interfaces for plline. The thing is that with Fortran 95 and beyond you can define a single interface name to be applied to multiple implementations. The compiler looks at the actual argument list and decides to use the implementation that matches that list (by number of arguments and the type, kind and rank of each). If it can not find any matching implementation, it will respond with a message like the above.

For the Fortran 95 bindings we have used the features provided by that standard as much as possible. One is that arrays "know" their size. So the routine plline queries the arrays xx and yy for their size and there is no need to pass the size explicitly. That method is actually rather error-prone.

If you want to pass a section of the array only, you can use:

Call plline( xx(1:10), yy(1:10) )

For instance.

> I verified that issue.  To solve it, I simply used the redacted form of
> subroutine/function call with the redundant dimension information dropped, i.e.,
>
> call plline(xx,yy)
>
> Note, it is quite a while since we have used the f77 interface so my original advice to
> you was incomplete about converting over to f95.  I should have also added to the
> "use plplot" advice that since we dropped f77, our f95 capabilities and API have
> evolved and, for example, we are now taking advantage of certain Fortran 95
> capabilities like knowing the redundant array dimensions.  So you have to use the
> redacted form (like above with redundant dimension information dropped) of all calls.
>
> When in doubt about exactly what the redacted form is, look at files in examples/f95/
> for working examples of all the calls to the PLplot Fortran binding API.
>


Regards,



Arjen

DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[Alan W. I. <ir...@be...>]

On 2015-01-17 16:46-0800 John Baumgardner wrote:

> Hi Alan,
>
>   I followed your recommendation and created a simple Fortran example that 
> illustrates the problem on Cygwin.  I created a tarball of the entire 
> plplot-5.10.0 directory that contains everything that cmake, make, and make 
> install generated in a subdirectory named build_dir, including the cmake.out, 
> make.out, and make_install.out files.  The shell script cmake.sh I used to 
> invoke cmake is in the main 5.10.0 directory.  Also in build_dir there is a 
> directory named 'test' which contains the Fortran example code named circle.f 
> that illustrates the problem.  There is also a .sh script named compile.sh 
> that I used to attempt to compile circle.f.  In addition I copied all the 
> contents of /bindings/f95 into this directory.

Hi John:

It's a judgement call, but 20MB is a fairly large attachment to be
sent to everyone on this list.  But partly that is my fault; I should
have been more specific about my request.

However, the good news is I found what I needed (circle.f and how you
built it) to figure out this problem which turns out not to be
specific to Cygwin.

> $ compile.sh
> circle.f:26.72:
>
>      call plline(n101,xx,yy)
>                                                                        1
> Error: There is no specific subroutine for the generic 'plline' at (1)

I verified that issue.  To solve it, I simply used the redacted form
of subroutine/function call with the redundant dimension information
dropped, i.e.,

call plline(xx,yy)

Note, it is quite a while since we have used the f77 interface so my
original advice to you was incomplete about converting over to f95.  I
should have also added to the "use plplot" advice that since we
dropped f77, our f95 capabilities and API have evolved and, for
example, we are now taking advantage of certain Fortran 95
capabilities like knowing the redundant array dimensions.  So you have
to use the redacted form (like above with redundant dimension
information dropped) of all calls.

When in doubt about exactly what the redacted form is, look at files
in examples/f95/ for working examples of all the calls to the PLplot
Fortran binding API.

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
__________________________

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[John B. <jrb...@co...>]

Hi Alan,

   I am re-sending this email with a smaller version of the tarfile, 
since the sourceforge server could not accept the 27 MB size of the 
original one.  This tarfile contains only portions of what is in 
build_dir.  Specifically, in build_dir I omitted all the bindings 
except for f95 and the share folder.  Hopefully what I am sending is 
adequate to sleuth out the issue.

Best regards,
John


Hi Alan,

    I followed your recommendation and created a simple Fortran 
example that illustrates the problem on Cygwin.  I created a tarball 
of the entire plplot-5.10.0 directory that contains everything that 
cmake, make, and make install generated in a subdirectory named 
build_dir, including the cmake.out, make.out, and make_install.out 
files.  The shell script cmake.sh I used to invoke cmake is in the 
main 5.10.0 directory.  Also in build_dir there is a directory named 
'test' which contains the Fortran example code named circle.f that 
illustrates the problem.  There is also a .sh script named compile.sh 
that I used to attempt to compile circle.f.  In addition I copied all 
the contents of /bindings/f95 into this directory.  Here is what I 
get when I attempt to compile this simple routine:

$ compile.sh
circle.f:26.72:

       call plline(n101,xx,yy)
                                                                         1
Error: There is no specific subroutine for the generic 'plline' at (1)

   My guess is that the problem is not something esoteric, but rather 
something very simple.  I suspect it may have to do with something 
related to the way I am trying to compile, perhaps something like 
missing some files not in /bindings/F95.

    I appreciate any help your team can provide--
John

At 11:47 AM 1/17/2015, Alan W. Irwin wrote:
>On 2015-01-17 11:01-0800 John Baumgardner wrote:
>
>>Hi Alan,
>>
>>Thanks for your quick response!  I had earlier tried the 5.10.0 
>>release with a version of my Fortran code where I had inserted 'use 
>>plplot' everywhere a plplot routine is called, but had encountered 
>>problems I could not resolve. That is why I went back to the 5.9.10 
>>release.  When I commented out all the calls to plwidth at least 
>>the code compiled and could be executed.  Here are the sorts of 
>>errors I get with the 5.10.0 release:
>>
>>$ make sphplt
>>gfortran -fdefault-real-8 -O3 -Iinclude -Llib  -c -o graphx.o graphx.f
>>graphx.f:106.72:
>>
>>      call plline(n101,xx,yy)
>>                                                                        1
>>Error: There is no specific subroutine for the generic 'plline' at (1)
>>graphx.f:2079.72:
>>
>>         call plfill(4,x,y)
>>                                                                        1
>>Error: There is no specific subroutine for the generic 'plfill' at (1)
>>graphx.f:172.72:
>>
>>                  call plfill3(3,xr(1,1),xr(1,2),xr(1,3))
>>                                                                        1
>>Error: There is no specific subroutine for the generic 'plfill3' at (1)
>>
>>I copied the lib and include directories created in 
>>plplot-5.10.0/build_dir to the filespace where I am seeking to 
>>compile.  I also copied all the files in the directory bindings/f95 
>>to the local filespace.  I suspect I am still missing some crucial 
>>files.  Can you identify what is wrong here?
>>
>>Thanks so much for your assistance!
>
>Hi John:
>
>Could you make a self-contained (i.e., including your Makefile) but
>simplest possible example that demonstrates the problem on Cygwin, and
>post it as a tarball attachment to this list? I don't have any Cygwin
>experience myself, but one our our developers, in fact, uses that
>platform to develop our Fortran binding so given a convenient example
>to work with, he should be able to figure out what is going wrong
>pretty easily.  Also, I will attempt to build and run your example on
>Linux as well in case there is something generic that needs fixing
>with your example.
>
>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
>__________________________

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[Alan W. I. <ir...@be...>]

On 2015-01-17 11:01-0800 John Baumgardner wrote:

> Hi Alan,
>
> Thanks for your quick response!  I had earlier tried the 5.10.0 release with 
> a version of my Fortran code where I had inserted 'use plplot' everywhere a 
> plplot routine is called, but had encountered problems I could not resolve. 
> That is why I went back to the 5.9.10 release.  When I commented out all the 
> calls to plwidth at least the code compiled and could be executed.  Here are 
> the sorts of errors I get with the 5.10.0 release:
>
> $ make sphplt
> gfortran -fdefault-real-8 -O3 -Iinclude -Llib  -c -o graphx.o graphx.f
> graphx.f:106.72:
>
>      call plline(n101,xx,yy)
>                                                                        1
> Error: There is no specific subroutine for the generic 'plline' at (1)
> graphx.f:2079.72:
>
>         call plfill(4,x,y)
>                                                                        1
> Error: There is no specific subroutine for the generic 'plfill' at (1)
> graphx.f:172.72:
>
>                  call plfill3(3,xr(1,1),xr(1,2),xr(1,3))
>                                                                        1
> Error: There is no specific subroutine for the generic 'plfill3' at (1)
>
> I copied the lib and include directories created in plplot-5.10.0/build_dir 
> to the filespace where I am seeking to compile.  I also copied all the files 
> in the directory bindings/f95 to the local filespace.  I suspect I am still 
> missing some crucial files.  Can you identify what is wrong here?
>
> Thanks so much for your assistance!

Hi John:

Could you make a self-contained (i.e., including your Makefile) but
simplest possible example that demonstrates the problem on Cygwin, and
post it as a tarball attachment to this list? I don't have any Cygwin
experience myself, but one our our developers, in fact, uses that
platform to develop our Fortran binding so given a convenient example
to work with, he should be able to figure out what is going wrong
pretty easily.  Also, I will attempt to build and run your example on
Linux as well in case there is something generic that needs fixing
with your example.

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
__________________________

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[John B. <jrb...@co...>]

Hi Alan,

Thanks for your quick response!  I had earlier tried the 5.10.0 
release with a version of my Fortran code where I had inserted 'use 
plplot' everywhere a plplot routine is called, but had encountered 
problems I could not resolve.  That is why I went back to the 5.9.10 
release.  When I commented out all the calls to plwidth at least the 
code compiled and could be executed.  Here are the sorts of errors I 
get with the 5.10.0 release:

$ make sphplt
gfortran -fdefault-real-8 -O3 -Iinclude -Llib  -c -o graphx.o graphx.f
graphx.f:106.72:

       call plline(n101,xx,yy)
                                                                         1
Error: There is no specific subroutine for the generic 'plline' at (1)
graphx.f:2079.72:

          call plfill(4,x,y)
                                                                         1
Error: There is no specific subroutine for the generic 'plfill' at (1)
graphx.f:172.72:

                   call plfill3(3,xr(1,1),xr(1,2),xr(1,3))
                                                                         1
Error: There is no specific subroutine for the generic 'plfill3' at (1)

I copied the lib and include directories created in 
plplot-5.10.0/build_dir to the filespace where I am seeking to 
compile.  I also copied all the files in the directory bindings/f95 
to the local filespace.  I suspect I am still missing some crucial 
files.  Can you identify what is wrong here?

Thanks so much for your assistance!
John

At 12:02 AM 1/17/2015, Alan W. Irwin wrote:
>On 2015-01-16 17:45-0800 John Baumgardner wrote:
>
>>For some strange reason when I try to link my Fortran graphics 
>>application under Cygwin against the plplot 5.9.10 libraries, 
>>routine plwidth fails to link, despite the fact that many other 
>>plplot procedures link and work flawlessly.  Can anyone provide a 
>>suggestion as to how I might resolve this?
>
>Hi John:
>
>I noticed that you were using the f77 binding of PLplot which was
>deprecated for release 5.9.10, and dropped for 5.9.11 for the reasons
>we mentioned in those release announcements.  Instead, you should be
>using the f95 bindings of the latest released version of PLplot which
>is 5.10.0.
>
>The support for Fortran 77 in our f95 bindings is actually pretty
>good since Fortran 95 generally supports most of Fortran 77. For
>example, your Fortran 77 code that worked (more or less) with the f77
>binding normally should require very few changes (usually just the
>addition of a "use plplot" statement) to work with our latest f95
>binding. So please try that possibility with our latest release
>(5.10.0), and let us know how it goes.
>
>By the way, we are rewriting much of our f95 binding right now to make
>those bindings much simpler for us to maintain, and also to make it a
>lot easier to use by our users. For example, that rewritten binding
>should provide a version of all PLplot calls with single precision
>arguments and also double precision arguments which will be selected
>automatically by the f95 binding (via function overloading) depending
>on whether all the real arguments of a call to a PLplot Fortran
>routine are all single precision or all double precision.  That
>updated version is not yet working correctly for us, but when it does
>we will commit it to our git master branch to give Fortran users like
>you a chance to test it thoroughly on all platforms before we
>officially release it. So please keep that future Fortran testing
>opportunity in mind (which we will announce here when the time comes).
>
>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
>__________________________

Re: [Plplot-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[Alan W. I. <ir...@be...>]

On 2015-01-16 17:45-0800 John Baumgardner wrote:

> For some strange reason when I try to link my Fortran graphics application 
> under Cygwin against the plplot 5.9.10 libraries, routine plwidth fails to 
> link, despite the fact that many other plplot procedures link and work 
> flawlessly.  Can anyone provide a suggestion as to how I might resolve this?

Hi John:

I noticed that you were using the f77 binding of PLplot which was
deprecated for release 5.9.10, and dropped for 5.9.11 for the reasons
we mentioned in those release announcements.  Instead, you should be
using the f95 bindings of the latest released version of PLplot which
is 5.10.0.

The support for Fortran 77 in our f95 bindings is actually pretty
good since Fortran 95 generally supports most of Fortran 77. For
example, your Fortran 77 code that worked (more or less) with the f77
binding normally should require very few changes (usually just the
addition of a "use plplot" statement) to work with our latest f95
binding. So please try that possibility with our latest release
(5.10.0), and let us know how it goes.

By the way, we are rewriting much of our f95 binding right now to make
those bindings much simpler for us to maintain, and also to make it a
lot easier to use by our users. For example, that rewritten binding
should provide a version of all PLplot calls with single precision
arguments and also double precision arguments which will be selected
automatically by the f95 binding (via function overloading) depending
on whether all the real arguments of a call to a PLplot Fortran
routine are all single precision or all double precision.  That
updated version is not yet working correctly for us, but when it does
we will commit it to our git master branch to give Fortran users like
you a chance to test it thoroughly on all platforms before we
officially release it. So please keep that future Fortran testing
opportunity in mind (which we will announce here when the time comes).

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-general] plwidth fails to link under Cygwin against plplot5.9.10 libraries

[John B. <jrb...@co...>]

For some strange reason when I try to link my Fortran graphics 
application under Cygwin against the plplot 5.9.10 libraries, routine 
plwidth fails to link, despite the fact that many other plplot 
procedures link and work flawlessly.  Can anyone provide a suggestion 
as to how I might resolve this?  Thanks,  John Baumgardner


gfortran -fdefault-real-8 -O3 -I/usr/local/include 
-L/usr/local/lib  -fdefault-real-8 -O3 
-L~/plplot-5.9.10/build_dir/lib  -o sphplt sphplt.o graphx.o \
-lplplotd -lplplotf77d -lplplotf77cd
graphx.o:graphx.f:(.text+0xe832): undefined reference to `plwidth_'
graphx.o:graphx.f:(.text+0xe832): relocation truncated to fit: 
R_X86_64_PC32 against undefined symbol `plwidth_'
collect2: error: ld returned 1 exit status
makefile:33: recipe for target 'sphplt' failed
make: *** [sphplt] Error 1

Re: [Plplot-general] Plotting from multithreaded environment

[Arjen M. <Arj...@de...>]

Hi Sergei,

> -----Original Message-----
> From: Sergei Naumov [mailto:vo...@ra...]
> Sent: Tuesday, January 13, 2015 5:45 PM
> To: Arjen Markus
> Cc: plp...@li...
> Subject: RE: RE: RE: RE: RE: [Plplot-general] Plotting from multithreaded
> environment
>
>
>
> > I do not know anything about the GTK multithreading capabilities, so what I would
> like to know is, where are the threads introduced, is that in g_timeout()?
> > But I would think that it initialises a single extra thread. Plplot
> > itself is not threadsafe, that is, you should not use the functions from different
> threads at the same time, but I would imagine that global variables (as maintained by
> the Plplot library) are accessible in any thread. Hence, there should be no problem.
>
> The call simply creates a new thread that executes a specified function every N
> seconds. In my case it is 15. All other variables and structures declared globally in
> my code are accessable from inside. There seem to be nothing unusual about it.
>


Indeed, that seems rather straightforward.

> > What happens if you use the PSC device? That produces a file on disk.
>
> It seems I did sometging wrong... But it produced a file with nothing in it.
>
> [serge@yaro c]$ ./ext
> Enter graphics output file name: oo
>
> *** PLPLOT ERROR, ABORTING OPERATION ***
> plvpor: Invalid limits, aborting operation
>
> *** PLPLOT ERROR, ABORTING OPERATION ***
> plwind: Please set up viewport first, aborting operation
>
> *** PLPLOT ERROR, ABORTING OPERATION ***
> plbox: Please set up window first, aborting operation

The error messages indicate that Plplot was not properly initialized - at least according to the underlying routines. Since it has been initialized, I am beginning to doubt whether the internal structures are indeed accessible. But that is just a hunch and I have no idea how that would come about. Hm, perhaps I can try this in a small C program myself, apart from GTK.

Regards,

Arjen





DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] Plotting from multithreaded environment

[Sergei N. <vo...@ra...>]

> I do not know anything about the GTK multithreading capabilities, so what I would like to know is, where are the threads introduced, is that in g_timeout()? 
> But I would think that it initialises a single extra thread. Plplot itself is not threadsafe, that is, you should not use the functions from different threads 
> at the same time, but I would imagine that global variables (as maintained by the Plplot library) are accessible in any thread. Hence, there should be no problem.

The call simply creates a new thread that executes a specified function every N seconds. In my case it is 15. All other variables and structures declared globally 
in my code are accessable from inside. There seem to be nothing unusual about it.

> What happens if you use the PSC device? That produces a file on disk.

It seems I did sometging wrong... But it produced a file with nothing in it.

[serge@yaro c]$ ./ext
Enter graphics output file name: oo

*** PLPLOT ERROR, ABORTING OPERATION ***
plvpor: Invalid limits, aborting operation

*** PLPLOT ERROR, ABORTING OPERATION ***
plwind: Please set up viewport first, aborting operation

*** PLPLOT ERROR, ABORTING OPERATION ***
plbox: Please set up window first, aborting operation

Re: [Plplot-general] Plotting from multithreaded environment

[Sergei N. <vo...@ra...>]

 13.01.2015, 17:12:27 пользователь Arjen Markus (Arj...@de...) написал:

> What happens if you do not use g_timeout()? I’d say the plot should appear (at least the axes).

Nop... I moved it to the end if init_app() and the axes appeared. Probably because plotting happens in the main gtk loop. But in this case when the time comes to call update_from_modbus() it dumps this to the console:

*** PLPLOT ERROR, ABORTING OPERATION ***
plsym: Please set up window first, aborting operation

The reason seems to be understandable: plend() closes the context of drawing. So it comes to my original question, how is drawing done across threads?

-- Sergei

Re: [Plplot-general] Plotting from multithreaded environment

[Arjen M. <Arj...@de...>]

Hi Sergei,



If you select a driver that writes to a pixmap, yes, then it should be there.



Try the simplest one: PostScript Colour.



Regards,



Arjen




> -----Original Message-----
> From: Sergei Naumov [mailto:vo...@ra...]
> Sent: Tuesday, January 13, 2015 2:14 PM
> To: Plplot General
> Cc: Arjen Markus
> Subject: RE: RE: RE: RE: [Plplot-general] Plotting from multithreaded environment
>
>
>
> > 13.01.2015, 17:09:55 пользователь Sergei Naumov (vo...@ra...)
> написал:
> >
> >
> >
> > 13.01.2015, 16:50:49 пользователь Arjen Markus (Arj...@de...)
> написал:
> >
> > > Hi Sergei,
> > > The purpose of plend() (called in setup_plot_drawable()) is to close off the Plplot
> library, so that it can free all the resources. It should only be called if you are done
> with plotting.
> > > What happens if you move the call to the end of the main program?
> > > (Sorry, can't try this myself as I do not have a GTK environment ready)
> >
> > I kind of understand it. That's why the first thing I did was to move it to the end of
> main(). The screen is black, nothing is drawn.
>
> May be there is something in plend() that also dumps the picture into a PIXMAP and
> I can call it directly?
>
> -- Sergei
>
>
>  =

DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] Plotting from multithreaded environment

[Sergei N. <vo...@ra...>]

> 13.01.2015, 17:09:55 пользователь Sergei Naumov (vo...@ra...) написал:
> 
> 
> 
> 13.01.2015, 16:50:49 пользователь Arjen Markus (Arj...@de...) написал:
> 
> > Hi Sergei,
> > The purpose of plend() (called in setup_plot_drawable()) is to close off the Plplot library, so that it can free all the resources. It should only be called if you are done with plotting. 
> > What happens if you move the call to the end of the main program?
> > (Sorry, can’t try this myself as I do not have a GTK environment ready)
> 
> I kind of understand it. That's why the first thing I did was to move it to the end of main(). The screen is black, nothing is drawn.

May be there is something in plend() that also dumps the picture into a PIXMAP and I can call it directly?

-- Sergei


 

Re: [Plplot-general] Plotting from multithreaded environment

[Arjen M. <Arj...@de...>]

Hi Sergei,



What happens if you do not use g_timeout()? I'd say the plot should appear (at least the axes).

Regards,

Arjen


> -----Original Message-----
> From: Sergei Naumov [mailto:vo...@ra...]
> Sent: Tuesday, January 13, 2015 2:10 PM
> To: Arjen Markus
> Cc: plp...@li...
> Subject: RE: RE: RE: [Plplot-general] Plotting from multithreaded environment
>
>
>
>
>
>
> 13.01.2015, 16:50:49 пользователь Arjen Markus (Arj...@de...)
> написал:
>
> > Hi Sergei,
> > The purpose of plend() (called in setup_plot_drawable()) is to close off the Plplot
> library, so that it can free all the resources. It should only be called if you are done
> with plotting.
> > What happens if you move the call to the end of the main program?
> > (Sorry, can't try this myself as I do not have a GTK environment ready)
>
> I kind of understand it. That's why the first thing I did was to move it to the end of
> main(). The screen is black, nothing is drawn.
>
> -- Sergei
>

DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] Plotting from multithreaded environment

[Arjen M. <Arj...@de...>]

Hi Sergei,



Yes, I got it. Which is why I could reply to it :). I think it is the use of plend() in setup_plot_drawable() that is the culprit. Move it to the end of main().



Regards,



Arjen

> -----Original Message-----
> From: Sergei Naumov [mailto:vo...@ra...]
> Sent: Tuesday, January 13, 2015 2:07 PM
> To: Arjen Markus
> Cc: plp...@li...
> Subject: RE: RE: [Plplot-general] Plotting from multithreaded environment
>
>
> 13.01.2015, 16:01:48 пользователь Arjen Markus (Arj...@de...)
> написал:
>
> > Hi Sergei,
> >
> > Can you give us a (preferrably small) example of this behaviour?
>
> Seems the attachment did not come to the mailing list. Did it come to you?
>
> -- Sergei
>

DISCLAIMER: This message is intended exclusively for the addressee(s) and may contain confidential and privileged information. If you are not the intended recipient please notify the sender immediately and destroy this message. Unauthorized use, disclosure or copying of this message is strictly prohibited. The foundation 'Stichting Deltares', which has its seat at Delft, The Netherlands, Commercial Registration Number 41146461, is not liable in any way whatsoever for consequences and/or damages resulting from the improper, incomplete and untimely dispatch, receipt and/or content of this e-mail.

Re: [Plplot-general] Plotting from multithreaded environment

[Sergei N. <vo...@ra...>]

13.01.2015, 16:01:48 пользователь Arjen Markus (Arj...@de...) написал:

> Hi Sergei,
> 
> Can you give us a (preferrably small) example of this behaviour?

Seems the attachment did not come to the mailing list. Did it come to you?

-- Sergei