[GRASS5] GRASS5.0.0pre1 on cygwin/Windows 2000

Hi,

i just updated my cygwin installation. There is a new cygwin dll that
solves some of the problems that i found using the grass modules
recently.

But recompiling the latest pre release code did not fix the X related
problem.

The X Driver/Monitor starts, a black, empty window displays, but the
display modules (d.mon, d.rast etc.) can not connect to the grass
monitor. The monitor can not be closed with the mouse nor with d.mon
stop..., it even is not killed if the StarNet X Window system is shut
down. Only a direct kill with the Task Manager removes the process of
the XDRIVER. Then a stackdump is left behind.

If someone has ideas what to check, please tell me. It once worked with
this setup.

cu,

Andreas

--
Andreas Lange, 65187 Wiesbaden, Germany, Tel. +49 611 807850
Andreas.Lange@Rhein-Main.de - A.C.Lange@GMX.net

On Thu, May 24, 2001 at 07:51:52PM +0200, Andreas Lange wrote:

Hi,

i just updated my cygwin installation. There is a new cygwin dll that
solves some of the problems that i found using the grass modules
recently.

But recompiling the latest pre release code did not fix the X related
problem.

The X Driver/Monitor starts, a black, empty window displays, but the
display modules (d.mon, d.rast etc.) can not connect to the grass
monitor. The monitor can not be closed with the mouse nor with d.mon
stop..., it even is not killed if the StarNet X Window system is shut
down. Only a direct kill with the Task Manager removes the process of
the XDRIVER. Then a stackdump is left behind.

If someone has ideas what to check, please tell me. It once worked with
this setup.

I don't have any ideas, but I get different behavior (using XFree86)
under CygWin. I just set it up a couple days ago and built GRASS, but
the monitor exits immediately after showing itself (some Cygwin bug with
select() ???). If the monitor is locked like you say, maybe it somehow
is stuck in an uninterruptable system call? Does "ps" give any useful
output (like a "D" state) while this is going on? Given the kludgey
nature of CygWin, I'm as likely to suspect it as anything.

BTW: If you get GRASSLIST you might want to chime in on the discussion
of no datum transformations in v.proj...

--
Eric G. Miller <egm2@jps.net>

Hi all.

> i just updated my cygwin installation. There is a new cygwin dll that
> solves some of the problems that i found using the grass modules
> recently.

Having seen you guys meet this problem, I have not upgraded.

> But recompiling the latest pre release code did not fix the X related
> problem.
>
> The X Driver/Monitor starts, a black, empty window displays, but the
> display modules (d.mon, d.rast etc.) can not connect to the grass
> monitor. The monitor can not be closed with the mouse nor with d.mon
> stop..., it even is not killed if the StarNet X Window system is shut
> down. Only a direct kill with the Task Manager removes the process of
> the XDRIVER. Then a stackdump is left behind.

...

I don't have any ideas, but I get different behavior (using XFree86)
under CygWin. I just set it up a couple days ago and built GRASS, but
the monitor exits immediately after showing itself (some Cygwin bug with
select() ???).

Today I built the prerelease under NT both as-is and also with the Win98
mods from Malcolm.

Under Nutcracker/DEC X-Server, using Cygwin 1.1.8, I have no troubles with
the standard build.

With the W98 modified version I get exactly the behaviour just described by
Eric - the monitor shows and immediately exits.

Regarding my earlier posts about read() and Cygwin binary mounted drives, I
realised that if open() is called with O_BINARY as the second argument, it
should fix the problem. I'll try it out this weekend if I can get near my
computer.

I am trying to get the font building program "splitfonts" to work under
Windows too. The popen call seems to cause problems.

Cheers

Mike Thomas