[winGRASS] Re: winGRASS, cygwin, XFree86

[John Huddleston]

Andreas, (Cc to Wingrass)

I checked out the Grass5 system to a new directory and did the configure.
The head file that is created still has some confusing include and link
arguments.   If you have old X11 and GL directories on your system,
the head file may access older files.

Also, a "feature" of the gcc compiler is to take user include directories
(e.g. -I/usr/local/include) before the gcc directories.  Thus if you have
older header files, it will grab those before the newer system ones.
One of the ways (besides cleaning up your system) to ensure that
you get the system headers is to use "-I/usr/include" before the other
include directories.

For your own sanity, run the "updatedb" command over night and then
use the "locate" command the next day to find libraries and includes
headers that are critical to the X11, GL, PostgreSQL, ODBC, PNG,
and other builds.  Remove the older ones, so that even if you do not
fix the head file (which I do for every build) you will be assured of
getting current files.

Concerning the startxwin.bat file, it needs to be edited for your system
configuration.  Copy it to a new file, and use that file to start the
XFree86 system.

I still have my own Gmakefile for nviz.  The CVS one does not build.

John Huddleston

----- Original Message -----
From: "Andreas Lange" <Andreas.Lange at Rhein-Main.de>
To: <mthomas at gil.com.au>; <jhudd at lamar.colostate.edu>; <mblue at nb.sympatico.ca>; <raghava at media.osaka-cu.ac.jp>; <aprasad at fs.fed.us>;
<neteler at itc.it>; <glynn.clements at virgin.net>; <bernhard at intevation.de>
Sent: Sunday, August 05, 2001 4:47 PM
Subject: winGRASS, cygwin, XFree86


> Hi winGRASS friends,
>
> the problem with the XDRIVER is solved now, many thanks to all who
> helped. configure is working again, too. So from the GRASS side
> everything is fine (i compiled and tested the PNGdriver, too).
>
> But i got today really frustrated with cygwin and XFree86 for cygwin.
>
> All xterms do not return after exiting, so tcltkgrass is screwed up if i
> start a grass monitor, i think there is a problem that the xterm does
> not close/return, so control is not returned to wish/tcltkgrass. The
> problem is already reported on the xfree-cygwin list. This affects xterm
> (and other X programs?), so the StarNet Server does not work too.
>
> The XFree86 cygwin installation procedure does not work. There is a bug
> report on the mailing list, but this generated a similar answer to the
> report of Mike. And the startxwin.bat file provided with the cygwin
> installation bails out doing nothing. So all in all XFree86 does not
> work any more. And from the discussions on the xfree mailing list i
> conclude that it will be a long way until there will be a usable
> installation procedure.
>
> I think it is a must that the windows users can start GRASS by one
> mouse-klick. Otherwise winGRASS will not find many new friends.
>
> I am still working on gdal (hope to create a cygwin version with the
> help of Frank Warmerdam). I had partial success in compiling the
> postgreSQL modules, i'll check in a fix for one module. Still a problem
> with a reference to postgres.h, which does not exist on cygwin.
>
> I'll work on the problems later this week again. All in all i think that
> we make (somewhat) slow progress to a usable winGRASS port, but as we
> depend on the cygwin and XFree86 for cygwin, i can not tell any time
> frame when the port will be finished.
>
> More in a few days,
>
> cu,
>
> Andreas
>
>
> --
> Andreas Lange, 65187 Wiesbaden, Germany, Tel. +49 611 807850
> url: http://mitglied.tripod.de/AndreasLange
> mail: Andreas.Lange_at_Rhein-Main.de - A.C.Lange_at_GMX.net
>