-- Originally posted by (AT sourceforge.net): rtautenh -- -- This ticket was imported from http://sourceforge.net/p/rkward/bugs/129 on 2017-05-30 15:26:51 +0100 -- plots do not work any more for me in rkward on Ubuntu 10.04.4 e.g. an initial plot(1,1) creates the window and the grid, but does not show the dot. any plot command after that only creates an empty window but does not show anything. ---Session Info--- -- Frontend -- RKWard version: 0.6.1z+0.6.2+devel1 KDE version (runtime): 4.4.92 (KDE 4.4.92 (KDE 4.5 RC2)) KDE version (compile time): 4.4.92 (KDE 4.4.92 (KDE 4.5 RC2)) Qt build key: x86_64 linux g++-4 full-config Backend version (as known to the frontend): 2.15.1 Debug message file (this may contain relevant diagnostic output in case of trouble): /tmp/rkward.frontend.nS3063 -- Backend -- Debug message file (this may contain relevant diagnostic output in case of trouble): /tmp/rkward.rbackend.TJ3081 R version (compile time): 2.15.1 (2012-06-22 r59600) R runtime session info: R version 2.15.1 (2012-06-22) Platform: x86_64-unknown-linux-gnu (64-bit) locale: [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8 LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8 [7] LC_PAPER=en_US.UTF-8 LC_NAME=en_US.UTF-8 LC_ADDRESS=en_US.UTF-8 LC_TELEPHONE=en_US.UTF-8 LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=en_US.UTF-8 attached base packages: [1] stats graphics grDevices utils datasets methods base other attached packages: [1] rkward_0.6.1 loaded via a namespace (and not attached): [1] tools_2.15.1
Hi Ralf! (I tried to follow-up by mail, but this did not seem to work. If a double-post appears, you'll know, what happened.) Is this on the most recent svn revision (svn info)? Compiled from source? I know some revisions had similar problems, but I am not aware of a current problem. My testing is mostly with R 3.0.x and KDE >= 4.9, though. Please try this: - RMB click on one of the tool bars (e.g. the right hand one), check the "RKWard Debug Messages" window. Activate it. - Run some simple plot commands in the console until the problem appears. Copy any output from the RKWard Debug Messages window. - Change Configure RKWard->Debug->Debug level to "4", and repeat. Also: - What happens, if you follow up plot(1,1) with a) title("test") b) locator(1) Most likely you can work around this problem by changing the device settings at Configure RKWard->Onscreen Graphics, but of course it would be best, if we can track this down. Regards Thomas
Hi Ralf, On Monday 15 July 2013 17:23:33 Ralf Tautenhahn wrote: > plots do not work any more for me in rkward on Ubuntu 10.04.4 > > e.g. an initial plot(1,1) creates the window and the grid, but does not show > the dot. > > any plot command after that only creates an empty window but does not show > anything. is this on the most recent svn revision (svn info)? Compiled from source? I know some revisions had similar problems, but I am not aware of a current problem. My testing is mostly with R 3.0.x and KDE >= 4.9, though. Please try this: - RMB click on one of the tool bars (e.g. the right hand one), check the "RKWard Debug Messages" window. Activate it. - Run some simple plot commands in the console until the problem appears. Copy any output from the Debug Messages window. - Change Configure RKWard->Debug->Debug level to "4", and repeat. Also: - What happens, if you follow up plot(1,1) with a) title("test") b) locator(1) Most likely you can work around this problem by changing the device settings at Configure RKWard->Onscreen Graphics, but of course it would be nice to track this down. Regards Thomas
-- Originally posted by (AT sourceforge.net): rtautenh -- Hi Thomas, yes, I compiled the most recent version from source. URL: svn://svn.code.sf.net/p/rkward/code/trunk/rkward Repository Root: svn://svn.code.sf.net/p/rkward/code Repository UUID: 7936002b-cd1d-0410-9ce3-d7ebf93b5162 Revision: 4759 Node Kind: directory Schedule: normal Last Changed Author: m-eik Last Changed Rev: 4759 Last Changed Date: 2013-07-05 07:14:16 -0700 (Fri, 05 Jul 2013) There are no debugging message with debug level 2. With debug level 4 I get: DEBUG running command: plot(1,1) DEBUG Transmitting request of length 52 DEBUG Transmitting request of length 42 DEBUG Received transmission of type 7, devnum 2, size 6 DEBUG Received transmission of type 11, devnum 2, size 34 DEBUG Received transmission of type 11, devnum 2, size 34 DEBUG Received transmission of type 1, devnum 2, size 46 DEBUG Received transmission of type 11, devnum 2, size 34 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 18, devnum 2, size 55 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 6, devnum 2, size 95 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 6, devnum 2, size 95 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 6, devnum 2, size 95 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 6, devnum 2, size 95 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 6, devnum 2, size 95 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 2, devnum 2, size 60 DEBUG Received transmission of type 18, devnum 2, size 55 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 6, devnum 2, size 95 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 6, devnum 2, size 95 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 6, devnum 2, size 95 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 6, devnum 2, size 95 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 18, devnum 2, size 59 DEBUG Received transmission of type 6, devnum 2, size 95 DEBUG Received transmission of type 4, devnum 2, size 112 DEBUG Received transmission of type 11, devnum 2, size 34 DEBUG Received transmission of type 18, devnum 2, size 55 DEBUG Received transmission of type 6, devnum 2, size 91 DEBUG Received transmission of type 18, devnum 2, size 55 DEBUG Transmitting request of length 42 DEBUG Received transmission of type 6, devnum 2, size 91 DEBUG running command: locator(1) DEBUG Transmitting request of length 54 DEBUG Transmitting request of length 42 DEBUG Received transmission of type 20, devnum 2, size 2 - I changed setting to Embed Original Device and No Integration. That did not work. - However, changing it to Platform default device does work. Kind Regards, Ralf PS: Maybe I should upgrade to R3.0.1 ...
Hi! Thanks for your testing so far. What happened (with the "RKWard native" / RK() device), when running locator(1) after plot(1,1)? Did this cause the plot to appear? Could you try the following small modification in the sources? - edit rkward/rbackend/rkwarddevice/rkgraphicsdevice.h - add '//' (comment mark) at the start of lines 27 (#ifdef Q_WS_MAC) and 33 (#endif). (Or remove these two lines). - recompile - try again (making sure you really test the "RKWard native" / RK() device, of course) Thanks!
-- Originally posted by (AT sourceforge.net): rtautenh -- Hi Thomas, using locator(1) after plot(1,1) makes no difference. The plot does not show up (but I do see the location on the console). when I comment out lines 27 and 33 in rkgraphicsdevice.h and try to compile I get this error (but it compiles fine when I put them back in): [ 78%] Built target rkward.rbackend /vol/R/rkward/trunk-new/rkward/rkward/rbackend/rkwarddevice/rkgraphicsdevice.cpp: In member function ‘void RKGraphicsDevice::clear(const QColor&)’: /vol/R/rkward/trunk-new/rkward/rkward/rbackend/rkwarddevice/rkgraphicsdevice.cpp:126: error: no matching function for call to ‘QImage::fill(const QColor&)’ /usr/include/qt4/QtGui/qimage.h:212: note: candidates are: void QImage::fill(uint) /vol/R/rkward/trunk-new/rkward/rkward/rbackend/rkwarddevice/rkgraphicsdevice.cpp:127: error: no matching function for call to ‘QImage::fill(QColor)’ /usr/include/qt4/QtGui/qimage.h:212: note: candidates are: void QImage::fill(uint) make[2]: *** [rkward/rbackend/rkwarddevice/CMakeFiles/rkgraphicsdevice.frontend.dir/rkgraphicsdevice.o] Error 1 Thanks! Ralf
> when I comment out lines 27 and 33 in rkgraphicsdevice.h and try to compile I get this error (but it compiles fine when I put them back in): Ok, comment out the lines again, then replace the offending lines 126 and 127 in rkgraphicsdevice.cpp with if (col.isValid ()) area.fill (col.rgb()); else area.fill (qRgb (255, 255, 255)); then try once more. Thanks! Thomas
-- Originally posted by (AT sourceforge.net): rtautenh -- Hi Thomas, perfect! rkward's native device now works with this patch! Thanks, Ralf
Hi! > perfect! rkward's native device now works with this patch! Glad to hear that. That leaves the bug (in the unpatched version) somewhere between Qt and the X server. Could you provide version numbers for these two for future reference? (Background: This patch causes all rendering to be done inside the rkward process, rather than in the X-server (which might utilize hardware acceleration). My impression - and rudimentary timing - is that this does not make _too_ much of a difference for R plots, and so I guess I'll simply enable it for all X11-compiles for now. However, in case we run into performance problems at some point of time, it will be useful to know better, when using the X server for rendering is reliable, and when it is not.) Regards Thomas
-- Originally posted by (AT sourceforge.net): rtautenh -- libqt is 4:4.7.0~beta2 libx11 is 2:1.3.2 xserver-xorg is 1:7.5 Thanks again for fixing! Ralf
-- Originally posted by (AT sourceforge.net): ziogelis77 -- I wonder if I perhaps have the same bug: if I run hist(x), only an empty device appears. but if I resize the window just a bit, the plot does appear. If after that I run hist(y) (or any other plot command), the new plot does not get displayed till I resize the window again. I am using Version 0.6.1 on KDE 4.10.5, Kubuntu 13.04, from the repository deb http://ppa.launchpad.net/rkward-devel/rkward-stable/ubuntu raring main Please advise me if I should open another bug or just wait for this fix to reach the repos.
Hi Donatas! This is almost definitely a different bug. Ralf was having trouble with the RK()-device, which is a new feature introduced after 0.6.1 (only in development versions, so far). > Please advise me if I should open another bug or just wait for this fix to reach the repos. Go ahead, and open a new report. If you have any idea, when the problem has started, please include that bit of info. However, I'll be very short on time the upcoming weeks, so I won't promise anything. You may want to give the current development snapshot a try (https://launchpad.net/~rkward-devel/+archive/rkward-dailys), where the RK()-device is the default device, already. Regards Thomas
- **status**: open --> closed-fixed
Original issue fixed, long ago. Secondary issue as reported by Donatas not reproduced, since. Closing. Donatas: Do let us know, if the issue still persists.