On Dec 29 2010, 6:49 pm, emil <emil.widm...@gmail.com> wrote: > On 29 Dez., 17:33, kcrisman <kcris...@gmail.com> wrote: > > > > > The main problem is that Puppy Linux or whatever must have compiled > > Sage without the right headers available. It's entirely mysterious to > > us exactly why R requires certain developer tools installed to give > > you X11. But it doesn't, and so in the spkg-install we check for as > > many of these things as we know about. However, the Puppy > > installation probably doesn't have them all, so when it compiled R, it > > did so without support for this. Here is part of it (without some > > comments): > > > if [ -f /usr/include/X11/Xwindows.h ]; then > > XSUPPORT=yes > > else > > if [ "x`uname`" = xSunOS ] && [ -f /usr/X11/lib/libXv.so ] ; then > > XSUPPORT=yes > > else > > XSUPPORT=no > > fi > > fi > > > On Mac we instead enable Aqua, because X11 support is very difficult > > to implement for the general case (it depends heavily on version of OS > > X). > > > So one solution would be to compile Sage from scratch (which is easy); > > another is to download a binary built on a machine which had these > > things. We really wish there was a good solution to this, but > > repeated requests on the R help list did not lead to further > > elucidation. > > > Also, not all R packages behave nicely in this way - I've had trouble > > installing ones that assumed access to certain things like OpenGL as > > well. But you are right that big ones like R Commander should be ok. > > > I'm not sure whether this will help, but it should help understand > > what the problem is, anyway. > > > - kcrisman > > Ok, thanks for the pointers! > > In fact, as I said, I have just used the usual ubuntu binaries from > the sage server to make > the live cd. I did this because I figured this could be a possibility > to reduce the amount of work to have a working > live CD considerably (given the fast release cycle of sage). And, as I > said, it passed sage -testall without errors (exception: cython throws > on error because there is no gcc installed in the base distribution). > Also, as mentioned, the current ubuntu vmware image has exactly the > same problem (i.e. plotting in R doesn't work, r.capabilities() give > the same values). > > Also I have every header installed in my installation. There is a well > tested development package, and I have even installed all the kernel > headers . So I doubt it is really a problem of missing header files in > the base OS. The same packages are used for all compiling purposes. > > I will compile a version of R and try to get it working, then I will > compile sage from scratch an use it instead of the ubuntu binaries. > Lets see where "the dog is burried" (german proverb). > > So is it right to say that plotting from R, and installation of most R- > packages should work in a working sage install? > Just to be sure it's no ghost I'm hunting ... > emil
So R-graphics in sage don't work by default, this is probably exactly http://trac.sagemath.org/sage_trac/ticket/8868 It is not working in the Ubuntu 32 bit binaries, the ubuntu 64 bit binaries and the current VM image. I built R from source on my machine, and I have jpeg (TRUE ) png (TRUEE) tiff (TRUE ) tcltk (TRUE) X11 (TRUE) aqua (FALSE) http/ftp (TRUE ) sockets (TRUE ) libxml (TRUE ) fifo (TRUE) cledit (TRUE ) iconv (TRUE) NLS (TRUE) profmem (FALSE ) cairo (TRUE) - i.e. all but aqua and profmem are TRUE So if I want to recompile sage from source (with FAT_BINARIES=TRUE), what compile options do I need to get those things working? Or could anyone point me to the right place inside the sage install scripts? I also think that there should be an error or at least a warning about failing R plotting when doing sage -testall. Happy new year to all, emil -- To post to this group, send email to sage-support@googlegroups.com To unsubscribe from this group, send email to sage-support+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-support URL: http://www.sagemath.org