[R] x11 and pseudo-color

[Paul Murrell]

Hi


joerg van den hoff wrote:
> for some reason the following message seems not to have reached the list 
>  in the first try, at least I can't find it. my apologies if this is my 
> fault:
> 
> we are running R under Solaris with SunRay Terminals, which are set to 8
> bit color to comply with some other software. In this configuration,
> X11() opens with colortype=true, i.e., it is not recognized that
> actually the display is only 8 bit. This leads to error messages
> (advising to use 'pseudo.cube').
> 
> question 1: is X11() supposed to recognize  the actual color
> capabilities? i.e. is this a bug?


R does try to determine the capabilities of the X11 server.  I think 
what may be happening is that your display is being correctly identified 
as pseudocolor, but R has two strategies for handling pseudocolor, 
"pseudo" and "pseudo.cube" and it uses "pseudo" by default (see ?x11). 
If you use more than 256 colours under this scheme you will run out of 
colours.


> question 2: is there a possibility to query the color capabilities from
> within R in order to being able to open the X11() displays always (for
> true color as well as for 8 bit) with the correct colortype setting from
> within a function?


Not that I know of.

Paul
-- 
Dr Paul Murrell
Department of Statistics
The University of Auckland
Private Bag 92019
Auckland
New Zealand
64 9 3737599 x85392
paul at stat.auckland.ac.nz
http://www.stat.auckland.ac.nz/~paul/