skt at stat.psu.edu (Steven K Thompson) writes:
> I have on occasion had the problem of missing title bars on X11
> graphics windows when using R, and would like to know what others have
> found in terms of the occurance, source, or solution of this problem.
> In searching for information on this I found only the brief thread
> from last November which I have copied below. Anyone who has
> experienced this knows it brings a very unwelcome interruption to work
> as one struggles to move or close the obstructing graphics windows
> having no frames.
>
> I first experienced this problem with a Mac OSX system in which I was
> using R under X11. My R program called C functions and the problem
> only occured when I was using graphics with fairly intensive
> computations going on at the same time.
>
> Recently, I've encountered this problem with a Linux system (Ubuntu)
> using Gnome 2.8 with its default window manager Metacity. On the
> advice from the thread below, I changed window manager (to icewm) and
> that eliminated the problem, except that the default Metacity seems
> otherwise to work more efficiently and faster with Gnome.
>
> I should note that I have been using Debian 3.0 (with gnome/icewm) for
> almost three years with never any problem like this.
>
> Thanks in advance for any information anyone may have on this issue.
>
> Steve
I did a bit of poking around a month or so ago, and found that the
example programs from an Xlib tutorial (I believe it was from
http://tronche.com/gui/x/xlib-tutorial/xlib-tutorial.tar.gz) did the
same thing. So this has much more to do with Metacity than it does
with R. The mystery is of course that all the non-raw-Xlib
applications seem to get by quite nicely, even xterms, so there is
probably a trick but I was unable to find it.
One should probably relay this information to the Gnome/Metacity guys
and get them either to acknowledge the bug or divulge the workaround.
(BTW, alt-F7 and alt-F8 allow you to move and resize the frame even
without the title bar, and maximize followed by unmaximize does put
the decoration right).
> From: Luke Tierney <luke_at_stat.uiowa.edu>
> Date: Fri 12 Nov 2004 - 01:16:00 EST
>
> On Thu, 11 Nov 2004, Martyn Plummer wrote:
>
> > On Wed, 2004-11-10 at 09:19 +0100, Martyn Plummer wrote:
> > > On Tue, 2004-11-09 at 19:47, Jonathan Baron wrote:
> > > > On 11/09/04 20:37, Jari Oksanen wrote:
> > > > >
> > > > >On 9 Nov 2004, at 19:44, Jonathan Baron wrote:
> > > > >
> > > > >> The RPM for Fedora Core 2 seems to work just fine
on Core 3.
> > > > >>
> > > > >> (The graphics window got smaller, but I'm sure
there is a
> > setting
> > > > >> for that.)
> > > > >>
> > > > >That would be good news. I really don't know how the
graphics
> > window
> > > > >became so big at some stage. (MacOS X is just cute here:
tiny,
> > sharp,
> > > > >fast graphics window.)
> > > >
> > > > I have the opposite problem, a 1680x1050 display.
> > > >
> > > > >Has the options()printcmd reappeared, so that
dev.print() works
> > without
> > > > >changing default options?
> > > >
> > > > I can't imagine how this would change. This is the same
"old"
> > > > RPM, not a new one. The option is there, and I don't
think it
> > > > ever disappeared. I can't test it. This is my laptop,
which is
> > > > not set up to print anything.
> > >
> > > My mistake. The default print command is determined at configure
> > time.
> > > But the RedHat RPMS are built in a sandbox that has only the
> > minimal
> > > configuration needed to build R. This doesn't include the lpr
> > package so
> > > the default print command is null. I will fix this in the next
RPM
> > > release, but right now I am upgrading to FC3.
> >
> > An RPM for Fedora Core 3 should be on a CRAN mirror near you by the
> > weekend. This fixes the printcmd bug.
> >
> > The X11() window is the right size for me, but it doesn't have a
> > title
> > bar, which is a nuisance.
> >
> > Martyn
> >
>
> I've seen this with X11 a little on FC2 but much more with rgl (maybe
> 50% of the time with rgl vs 5-10% for X11()). Ross Ihaka tried
> replacing the default metacity window manager with sawfish and found
> that the problem seemed to go away. Ross also found a bug report for
> metacity,
>
> https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=126571
>
> that might be related. It might also be something we are not doing
> quite right in opening a window that happens to bite metacity more
> than other wms. If anyone has the time and energy to pursue this
> please do.
>
> Best,
>
> luke
--
O__ ---- Peter Dalgaard Blegdamsvej 3
c/ /'_ --- Dept. of Biostatistics 2200 Cph. N
(*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907