R help - Jun 1998 - R-beta: netscape/R interaction

Using netscape 4.03 (or 2.x or 3.x)/ R-0.62.1 with SPARC Solaris 2.4:
when netscape is running for help, graphics in R quit with "Color
allocation
error". Closing netscape removes the problem. netscape does not interfere
with
other X11 imaging programs including ghostscript and ImageMagick except that it
increases the contrast in my mailer windows slightly.

(netscape does block launching of Solaris imaging and help tools. I have
e-mailed netscape through every generation about this problem but with no
response. our systems people tell me that this kind of problem with netscape
exists on all of our SPARCstations and with some software on our MS-Windows
machines leading to their recommendation of a different browser.I have tried a
different browser (w3) but that alternative is significantly slower.)

Any suggestions?

Gordon

 

Gordon M. Harrington		Mail:	3720 Village Place, #6308
Professor Emeritus			Waterloo, IA 50702-5848
University of Northern Iowa 	Phone:	319-291-8535
gordon.harrington at uni.edu	Fax:	319-291-8324

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help 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-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

Our solution to the famous  "Color allocation error"
(not only appearing in R;  
 S-plus e.g., uses black & white color scheme when it can't allocate the
 desired colors)

is to *ALWAYS* use
	netscape -install
	-----------------
(installs its own color map); this might be ugly but is working always,
contrary to	  =	netscape -noinstall
		  =which remedies the problem also in some circumstances, but not in all.

[[The ``best'' solution would be to switch to an environment with
  16 or 24 bit colors  ((eg. Linux / Intel)) ....]]

Martin

PS: I agree that it would be nicer if R dealt more gracefully with this
    problem. Ideally, the x11() driver would try to find ``close colors in
    the given color map'' as soon as it can't allocate new colors.

    Not that I knew anything on how this could be programmed....



-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help 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-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

gordon.harrington at uni.edu writes:
> Using netscape 4.03 (or 2.x or 3.x)/ R-0.62.1 with SPARC Solaris 2.4:
> when netscape is running for help, graphics in R quit with "Color
allocation
> error". Closing netscape removes the problem. netscape does not
interfere with
> other X11 imaging programs including ghostscript and ImageMagick except
that it
> increases the contrast in my mailer windows slightly.
> 
> (netscape does block launching of Solaris imaging and help tools. I have
> e-mailed netscape through every generation about this problem but with no
> response. our systems people tell me that this kind of problem with
netscape
> exists on all of our SPARCstations and with some software on our MS-Windows
> machines leading to their recommendation of a different browser.I have
tried a
> different browser (w3) but that alternative is significantly slower.)
You are running out of colors in your color map.  A typical X server
with 8 bits per pixel only has 256 different colors available at any
time.  Both Netscape and R will use up several of these colors.

Do you know how many bitplanes you can make available for your X
server?  If you can change from 8 bits per pixel to 16 bits per pixel,
that is by far the simplest thing to do.  At 16 bits per pixel you can
run "truecolor" with a 565 weighting (5 bits for red, 6 bits for
green, and 5 bits for blue) and you don't have to mess with color
maps.

If not, try using emacs and ESS (Emacs Speaks Statistics), the emacs
mode for running R (and S-PLUS and xlispstat and SAS and ...).  The
help facilities are integrated into ESS so you can display help pages
without having to go through a browser.

ESS is available from
		 http://franz.stat.wisc.edu/pub/ESS/
-- 
Douglas Bates                            bates at stat.wisc.edu
Statistics Department                    608/262-2598
University of Wisconsin - Madison        http://www.stat.wisc.edu/~bates/
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help 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-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

Hi,
> Using netscape 4.03 (or 2.x or 3.x)/ R-0.62.1 with SPARC Solaris 2.4:
> when netscape is running for help, graphics in R quit with "Color
allocation
> error". 
I'm not familiar with this specific configuration, but often the problem is:

Netscape allocates a lot of colors by default which then can't be used by
other programs. So after Netscape has "occupied" most of your
colormap,
other programs often cannot find enough colors to be able to work with. This
is not a specific Netscape problem, although netscape could behave somewhat
better.
> Any suggestions?
4, in the order I would try them:

1. Try to find a 16bit colordepth monitor :-)
2. Start R first, then Netscape (Netscape *can* behave itself...)
3. Start Netscape using the command line option "-install". It then
installs
   a private colormap. This will make your screen "flash" each time
you
   switch windows, but probably will solve your problem.
4. Start Netscape using the command line option "-mono". Away are your
   problems... and your colors in Netscape. :-)

Good luck,
  Hedderik.

---
http://www.swi.psy.uva.nl/usr/rijn or ICQ: 13506689

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help 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-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._