Lawrence Houston grass at boyd.geog.mcgill.ca
Thu Dec 29 12:22:14 EST 1994

On Thu, 29 Dec 1994, Frederic Laffay wrote:

> 	I make new call to the solidarity of users of GRASS4.1. Indeed, the 
> software is now installed and seemed to function on a Linux platform.
> 	However, when we launch d.mon to open a graphic screen, It plants.  I have 
> followed instructions of the file ginstall, and have therefore recopied the 
> file monitor.sample on monitorcap.  The former suitable call, for x0 and 
> others, to files fifo.1a, fifo.1b, etc.  ..  These files fifo are emptied 
> (length 0).
> 	Is this the to that ?  Has one to config, if yes, how?


The ginstall script will have adjusted the absolute path in the monitorcap
file to reflect the actuall location you choose for your GRASS 4.1 Tree. 
If you have overwritten the monitorcap generated by the ginstall script
with sample one, then adjustments to the paths used to access the "pipes" 
will have been lost and you will now be required to edit those entries by
hand (2 absolute paths per XDRIVER entry). 

The FIFOs are "special" files, in particular they are called "pipes". 
They should have lengths of 0 bytes, but will have a 'p' in the first
column of the permissions fields (.ie "prw-rw-rw-").  The ginstall script
should have created them correctly as well and hence NO further
configuration should be needed! 

I have been able to successfully run all of GRASS for Linux Binaries which
Andy has been kind enough to make available.  There may be problems if you
are NOT using a X11R5 based 2.x versions of XFree86 (.ie the newer X11R6
based 3.x versions may cause problems).  One "quirk" with GRASS binaries
running under Linux being a need to run "d.erase" or "d.frame -e" 
following any resizing of an XDRIVER Window, but other than that they have
seemed pretty stable.  Another problem may be GRASS's requirement to have 
a 256 (8-bit) Colour Display? 

Lawrence Houston  -  (houston at boyd.geog.mcgill.ca) 

More information about the grass-user mailing list