Re: [Plplot-devel] [plplot-devel] Re: Problem compiling latest SVN
On Mon, Aug 06, 2007 at 10:17:38AM -0600, Orion Poplawski wrote:
> Andrew Ross wrote:
> > Orion,
> >
> > -I${FREETYPE_INCLUDE_DIR} is not explicitly included in the driver's
> > includes. Each driver is supposed to set the xxx_COMPILE_FLAGS in the
> > driver cmake module cmake/modules/xxx. For the case of gcw this is
> > done in cmake/modules/gcw.cmake using pkg-config. Can you confirm what
> > output you get from the command
> > pkg-config libgnomeprintui-2.2 --cflags
> > For my linux (Debian and Ubuntu) systems this includes
> > -I/usr/include/freetype2 as required. It is possible that this is not
> > always the case?
> >
>
> # pkg-config libgnomeprintui-2.2 --cflags
> -I/usr/include/libgnomeprintui-2.2 -I/usr/include/libgnomeprint-2.2
> -I/usr/include/libgnomecanvas-2.0 -I/usr/include/libart-2.0
> -I/usr/include/glib-2.0 -I/usr/lib64/glib-2.0/include
> -I/usr/include/libxml2 -I/usr/include/pango-1.0 -I/usr/include/gtk-2.0
> -I/usr/lib64/gtk-2.0/include -I/usr/include/atk-1.0 -I/usr/include/cairo
>
> It appears that no public interface for libgnomeprinttui22 references a
> ft2 header, so this is as it should be. If plfreetype.h references
> ft2build.h directly, then any driver that uses plfreetype.h should use
> pkg-config freetype2 directly.
I've committed a svn change to explicitly add in the freetype include
flags / libraries for the gcw driver. Can you test this? It doesn't
break anything for me.
>
> > Which other drivers fail (try disabling gcw and see how far you get)?
> > psttf also uses pkg-config, while gd explicitly includes the required
> > directories.
>
> No others appear to fail.
It appears the other drivers handle the freetype libraries correctly -
that's good.
Andrew
-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
___
Plplot-devel mailing list
Plplot-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/plplot-devel
Re: [Plplot-devel] Cairo Driver Family
On 2007-08-07 00:28-0400 Hazen Babcock wrote: > > On Aug 5, 2007, at 2:56 PM, Alan W. Irwin wrote: > >> On 2007-08-05 12:02-0400 Hazen Babcock wrote: >> >>> >>> On Aug 3, 2007, at 3:54 AM, Alan W. Irwin wrote: >>> On 2007-08-02 23:04-0700 Alan W. Irwin wrote: > Now for some issues I discovered with the pscairo results. > (1) The bounding box is not correct. (I believe that is a known > issue which > you plan to fix.) >>> >>> Is the problem that the bounding box does not match the size of the >>> plot? >> >> Yes. > > Ok, I may have fixed this, at least for linux users. It was pretty simple but > also a bit puzzling. I don't really understand why Preview on OS-X was > mangling the plots. To compensate I was setting the bounding box and then > rotating the plot so that it would appear right side up and inside the > bounding box in Preview. To get the right results with gv on linux I had to > drop the rotation, but I still choose to flip the plot vertically so that it > would appear right side up. Let me know if that was also a mistake. Anyway, > I'll assume that the gv is correctly following the postscript standard and I > will use it for future testing of this driver. The orientation you had before was correct. i.e., the previous pscairo results and also ps and psttf view correctly under gv _only_ when you change its view to landscape. In the future, once you get the PLplot -portrait mode to work for pscairo, then those results (and also ps and psttf with -portrait mode) will view correctly under gv only when you use its (default) portrait mode. 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); PLplot scientific plotting software package (plplot.org); 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 __ - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
Re: [Plplot-devel] FT_Load_Glyph error
Alan W. Irwin wrote: > Hi Hazen: > > I have downloaded plplot-5.7.2.tar.gz, checked its signature, built it, > tested both the build tree and install tree examples. All seems well on both > Debian stable and Ubuntu Dapper except for the same issue recently reported > by Werner for MinGW; for both examples 7 and 23 on Ubuntu -dev psttfc gives > the following error message: > > terminate called after throwing an instance of 'std::runtime_error' >what(): Error returned from FT_Load_Glyph > Aborted I'm seeing these on Fedora. Has anyone gotten any farther on a fix? Andrew Ross posted: The error message comes from lasi, but only because FT_Load_Glyph returns an error. Supressing the fatal error in lasi means that most of example 07 plots successfully. There are just a couple of symbols which produce an error, and then just appear as a blank square in the plot. It does depend slightly which font you use. Removing Arial (the default font used) resulted in slightly fewer problems. The double quote character still caused errors though. So from this I surmise it is not (just) a font problem. It appears to be a freetype problem. Note that the plplot freetype code does not use FT_Load_Glyph, which may be why the problem does not occur in the gd driver. I've search the web and found no explicit reference to this so far. I will try and investigate further. - I'm tempted to just disable this test for now so that I can release a new plplot rpm for Fedora development. Any easy way that I can disable this specific test? I guess ctest -E psttfc should do the trick, although that disables all of the tests, not just number 07. -- Orion Poplawski Technical Manager 303-415-9701 x222 NWRA/CoRA DivisionFAX: 303-415-9702 3380 Mitchell Lane [EMAIL PROTECTED] Boulder, CO 80301 http://www.cora.nwra.com - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
Re: [Plplot-devel] [plplot-devel] Re: Problem compiling latest SVN
Andrew Ross wrote: > > I've committed a svn change to explicitly add in the freetype include > flags / libraries for the gcw driver. Can you test this? It doesn't > break anything for me. Works for me. Thanks! -- Orion Poplawski Technical Manager 303-415-9701 x222 NWRA/CoRA DivisionFAX: 303-415-9702 3380 Mitchell Lane [EMAIL PROTECTED] Boulder, CO 80301 http://www.cora.nwra.com - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
Re: [Plplot-devel] python examples hang on ppc64
On Mon, Aug 06, 2007 at 12:23:56PM -0600, Orion Poplawski wrote:
> The python examples x01 and x09 hang in an infinite loop on Fedora 7 on
> ppc64. The loop is in plline.c in grdashline :
>
>
> (gdb) print plsc->pendn
> $1 = 1
> (gdb) print plsc->space
> $2 = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0}
> (gdb) print plsc->curel
> $3 = 0
> (gdb) print plsc->nms
> $4 = 1
> (gdb) print plsc->mark
> $5 = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0}
>
OK. I think I know what the problem is. Both the mark and the space
arrays are blank, and so the dashed line is undefined. The problem must
be with plstyl and not with plline. plstyl takes two arrays of PLINT as
arguments.
I have recently altered this code since it was not 64-bit clean and did
not work on 64-bit sytems. I managed to get it working on a 64-bit intel
based system, but it clearly doesn't work on a 64-bit PPC. From this I
conjecture that it is an endian issue. PPC linux is big-endian isn't it?
The code you need to check is in bindings/python/plplotcmodule.i.
svn diff -r 7764:7765 bindings/python/plplotcmodule.i
will show you the relevant changes.
Things to check:
What size are int and long on your architecture?
Do you have stdint.h? (Should be a yes)
What is the python SIZEOF_LONG defined as - this is the default python
integer size. Plplot works with a 32-bit int.
myIntArray_ContiguousFromObject is the function doing the casting. This
is based on the floating point equivalent. Does single precision
floating point work on your architecture?
Andrew
-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
___
Plplot-devel mailing list
Plplot-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/plplot-devel
Re: [Plplot-devel] FT_Load_Glyph error
On Tue, Aug 07, 2007 at 10:24:22AM -0600, Orion Poplawski wrote: > > I'm seeing these on Fedora. Has anyone gotten any farther on a fix? > Andrew Ross posted: > > The error message comes from lasi, but only because > FT_Load_Glyph returns an error. Supressing the fatal error in lasi means > that most of example 07 plots successfully. There are just a couple of > symbols which produce an error, and then just appear as a blank square > in the plot. It does depend slightly which font you use. Removing Arial > (the default font used) resulted in slightly fewer problems. The double > quote character still caused errors though. > > So from this I surmise it is not (just) a font problem. It appears to be > a freetype problem. Note that the plplot freetype code does not use > FT_Load_Glyph, which may be why the problem does not occur in the gd > driver. > > I've search the web and found no explicit reference to this so far. I > will try and investigate further. Since this is a lasi / pango issue, it is slightly outside our control. It has been raised on the lasi list and Ed Trager is looking into it. The latest svn version of lasi has a temporary workaround. I don't think there is an easy way of removing an example just for a particular driver. You either need to remove the examples for all test using the c language bindings (by editing test/test_c.sh.in) , or you need to disable all psttf tests as you suggested. Andrew - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
Re: [Plplot-devel] python examples hang on ppc64
Andrew Ross wrote: > > OK. I think I know what the problem is. Both the mark and the space > arrays are blank, and so the dashed line is undefined. The problem must > be with plstyl and not with plline. plstyl takes two arrays of PLINT as > arguments. > > I have recently altered this code since it was not 64-bit clean and did > not work on 64-bit sytems. I managed to get it working on a 64-bit intel > based system, but it clearly doesn't work on a 64-bit PPC. From this I > conjecture that it is an endian issue. PPC linux is big-endian isn't it? > > The code you need to check is in bindings/python/plplotcmodule.i. > svn diff -r 7764:7765 bindings/python/plplotcmodule.i > will show you the relevant changes. > These changes appear to fix the issue. -- Orion Poplawski Technical Manager 303-415-9701 x222 NWRA/CoRA DivisionFAX: 303-415-9702 3380 Mitchell Lane [EMAIL PROTECTED] Boulder, CO 80301 http://www.cora.nwra.com - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
Re: [Plplot-devel] python examples hang on ppc64
On Tue, Aug 07, 2007 at 11:13:16AM -0600, Orion Poplawski wrote: > Andrew Ross wrote: > > > > OK. I think I know what the problem is. Both the mark and the space > > arrays are blank, and so the dashed line is undefined. The problem must > > be with plstyl and not with plline. plstyl takes two arrays of PLINT as > > arguments. > > > > I have recently altered this code since it was not 64-bit clean and did > > not work on 64-bit sytems. I managed to get it working on a 64-bit intel > > based system, but it clearly doesn't work on a 64-bit PPC. From this I > > conjecture that it is an endian issue. PPC linux is big-endian isn't it? > > > > The code you need to check is in bindings/python/plplotcmodule.i. > > svn diff -r 7764:7765 bindings/python/plplotcmodule.i > > will show you the relevant changes. > > > > These changes appear to fix the issue. So to confirm, the current svn version now works for you? We should probably have a check in for this blank line case anyway to avoid user error leading to a hang. Andrew - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
Re: [Plplot-devel] python examples hang on ppc64
Andrew Ross wrote: > On Tue, Aug 07, 2007 at 11:13:16AM -0600, Orion Poplawski wrote: >> Andrew Ross wrote: >>> OK. I think I know what the problem is. Both the mark and the space >>> arrays are blank, and so the dashed line is undefined. The problem must >>> be with plstyl and not with plline. plstyl takes two arrays of PLINT as >>> arguments. >>> >>> I have recently altered this code since it was not 64-bit clean and did >>> not work on 64-bit sytems. I managed to get it working on a 64-bit intel >>> based system, but it clearly doesn't work on a 64-bit PPC. From this I >>> conjecture that it is an endian issue. PPC linux is big-endian isn't it? >>> >>> The code you need to check is in bindings/python/plplotcmodule.i. >>> svn diff -r 7764:7765 bindings/python/plplotcmodule.i >>> will show you the relevant changes. >>> >> These changes appear to fix the issue. > > So to confirm, the current svn version now works for you? Yup. I had been using 5.7.3, but the latest svn works. > > We should probably have a check in for this blank line case anyway to > avoid user error leading to a hang. > Yeah, that would be good too. > Andrew -- Orion Poplawski Technical Manager 303-415-9701 x222 NWRA/CoRA DivisionFAX: 303-415-9702 3380 Mitchell Lane [EMAIL PROTECTED] Boulder, CO 80301 http://www.cora.nwra.com - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
Re: [Plplot-devel] Changes in octave
I haven't yet tried Octave 2.9.x, but as I understand it a major overhaul has gone on in the octave graphics. This is likely to need some investment of time from someone to 1) Check plplot works. 2) Make plplot work in the "correct way" using the new syntax. This might also be a good opportunity to brush up the matlab compatability routines. Personally this would be a great boon to me! Any octave gurus like to volunteer to look into this? If not it will get added to my list of things to do... Andrew On Mon, Aug 06, 2007 at 03:52:55PM -0600, Orion Poplawski wrote: > Looks like with the change from octave 2.9.10 -> 2.9.11, > automatic_replot is gone from octave along with an apparent rewrite of > the graphics system. I'm not sure what all the implications are yet for > plplot, but all references to automatic_replot need to be removed when > running against octave 2.9.11 or later. After I did that though, I'm > still seeing lots of errors running the tests: > > error: max_recursion_limit exceeded > error: evaluating assignment expression near line 34, column 9 > error: evaluating if command near line 31, column 5 > error: evaluating if command near line 29, column 3 > error: called from `gcf' in file `/usr/share/octave/2.9.13/m/plot/gcf.m' > error: evaluating argument list element number 1 > error: evaluating assignment expression near line 30, column 7 > error: evaluating if command near line 29, column 3 > error: called from `gca' in file `/usr/share/octave/2.9.13/m/plot/gca.m' > error: evaluating argument list element number 1 > error: evaluating argument list element number 1 > error: evaluating assignment expression near line 29, column 12 > error: evaluating if command near line 28, column 3 > error: called from `ishold' in file > `/usr/share/octave/2.9.13/m/plot/ishold.m' > error: evaluating assignment expression near line 80, column 26 > error: evaluating if command near line 79, column 3 > error: called from `figure' in file > `/builddir/build/BUILD/plplot-5.7.3/examples/../bindings/octave/PLplot/figure.m' > > and so on. So it appears that some other changes will be needed to be > made as well. > > > > -- > Orion Poplawski > Technical Manager 303-415-9701 x222 > NWRA/CoRA DivisionFAX: 303-415-9702 > 3380 Mitchell Lane [EMAIL PROTECTED] > Boulder, CO 80301 http://www.cora.nwra.com > > - > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > ___ > Plplot-devel mailing list > Plplot-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/plplot-devel > - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
Re: [Plplot-devel] Cairo Driver Family
On Aug 7, 2007, at 12:08 PM, Alan W. Irwin wrote: > On 2007-08-07 00:28-0400 Hazen Babcock wrote: > >> >> On Aug 5, 2007, at 2:56 PM, Alan W. Irwin wrote: >> >> Ok, I may have fixed this, at least for linux users. It was pretty >> simple but >> also a bit puzzling. I don't really understand why Preview on OS-X >> was >> mangling the plots. To compensate I was setting the bounding box >> and then >> rotating the plot so that it would appear right side up and inside >> the >> bounding box in Preview. To get the right results with gv on linux >> I had to >> drop the rotation, but I still choose to flip the plot vertically >> so that it >> would appear right side up. Let me know if that was also a >> mistake. Anyway, >> I'll assume that the gv is correctly following the postscript >> standard and I >> will use it for future testing of this driver. > > The orientation you had before was correct. i.e., the previous > pscairo > results and also ps and psttf view correctly under gv _only_ when > you change > its view to landscape. I've changed the orientation back and the bounding box should now also be correct. The ability to change the plot orientation remains unimplemented. -Hazen - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Plplot-devel mailing list Plplot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/plplot-devel
