[Rd] X11(colortype="true") does not give a truecolor visual(PR#1387)

[E.Velds@balpol.tudelft.nl]

ripley@stats.ox.ac.uk wrote:
> 
> On Sat, 16 Mar 2002 Edwin@balpol.tudelft.nl wrote:
> 
> > Full_Name: Edwin
> > Version: 1.4.1
> > OS: SunOS/Solaris
> > Submission from: (NULL) (130.161.1.115)
> >
> >
> > When selecting an X11 visual on Solaris 2.7 with
> >
> > X11(colortype="true")
> >
> > a PseudoColor visual is actually created. (verified with xwininfo)
> 
> Note that the help page does *not* say that this `selects a visual'. It
> says it selects a `color model'.  One can use a model which differs from
> the X11 visual class, but the default class and depth are always used.
> 
> This seems to be a misunderstanding of the `colortype' argument: there is
> no argument to select a visual class.

Not as much a misunderstanding, but more considered as closely related.

> 
> > The system in question (like many UNIX workstations) has both Pseudo and
> > truecolor visual types available, however the default type points to an 8 plane
> > PseudoColor visual.
> 
> I believe that is unusual.  Sun Xservers allow you to set the default
> visual on the command line: has the default been overridden?  Ours with
> multiple visuals are selecting TrueColor, but it does say the default is
> hardware-dependent.

It *can* be set as default, but on all of the Suns and most of the SGIs
I work with, it isn't. And I don't have the ability to change that
unfortunately. And there's an added problem that the truecolor visuals
in 2D are very slow compared to the pseudocolor, which makes it
undesireable to have the truecolor as default.

> 
> > Some applications look for the truecolor visual (netscape, acroread) and use it
> > when available. For others it needs to be explicitely requested (VNC viewer). R
> > seems to do neither and just goes with the default, even if a truecolor type is
> > requested.
> 
> As it does not say it will do anything else, this is not strictly a bug.
> It needs someone with the problem to contribute an enhancement, perhaps
> additional arguments to select the visual class and depth?
> 

I wondered about whether to file this as a bug and I reasoned that when
selecting a colour mode which is incompatible with the created visual,
there is something wrong. It does seem logical that the driver would try
to get the best suited visual for the colour mode and that there would
be a warning if the colour mode can't be supported.

In any case, it would be a nice enhancement for graphics support.

Edwin

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._