I have a couple of questions and will try to save a tiny bit of bandwidth by combining them.
1. Has anyone had NVIZ correctly running on a version of WinGRASS? If so, can this version be uploaded to the WinGRASS binary site?
2. I heard a couple weeks back that GRASS 5.3 and 5.7 had been successfully compiled for Cygwin. Do they seem completely functional? Is there any place I and/or my students can download these binaries to try?
Thanks very much for your help.
Michael Barton
______________________________
Michael Barton, Professor & Curator
Department of Anthropology
Arizona State University
Tempe, AZ 85287-2402
USA
voice: 480-965-6262; fax: 480-965-7671
Michael Barton wrote:
I have a couple of questions and will try to save a tiny bit of bandwidth by combining them.
1. Has anyone had NVIZ correctly running on a version of WinGRASS? If so, can this version be uploaded to the WinGRASS binary site?
I have never gotten it to work, however Glynn said he was successful.
2. I heard a couple weeks back that GRASS 5.3 and 5.7 had been successfully compiled for Cygwin. Do they seem completely functional? Is there any place I and/or my students can download these binaries to try?
I have 5.0.3 and 5.7 distributions at:
www.greenwoodmap.com/grass
As I said above, I have not had any luck with NVIZ. I have not tested
5.7 very much, but I think any testing you did would be useful to the
developers. If you are interested, I would be happy to put a 5.3
distribution together.
Thanks very much for your help.
Michael Barton
--
Richard Greenwood
www.greenwoodmap.com
Michael Barton wrote:
I have a couple of questions and will try to save a tiny bit of
bandwidth by combining them.
1. Has anyone had NVIZ correctly running on a version of WinGRASS?
Yes. Well, I haven't tested it extensively, but the basics definitely
work.
If so, can this version be uploaded to the WinGRASS binary site?
I've uploaded my NVWISH2.2 binary (from 5.3) to grass.itc.it/incoming,
although I can't guarantee that it will work on any given system (at
least Tcl/Tk is linked statically, so that's one package which
shouldn't cause any problems).
--
Glynn Clements <glynn.clements@virgin.net>
Glynn Clements wrote:
I've uploaded my NVWISH2.2 binary (from 5.3) to grass.itc.it/incoming,
although I can't guarantee that it will work on any given system (at
least Tcl/Tk is linked statically, so that's one package which
shouldn't cause any problems).
I would be very interested in trying your binary on my Cygwin installations, which I have been using to build WinGRASS. I am suspicious that I have been over zealous with updating my system.
However I can not read from grass.itc.it/incoming. Is it write-only? Is there an alternative method by which I could fetch your binary? Or would you be able to email or ftp it to me?
Regards,
Rich
--
Richard Greenwood
www.greenwoodmap.com
Richard Greenwood wrote:
> I've uploaded my NVWISH2.2 binary (from 5.3) to grass.itc.it/incoming,
> although I can't guarantee that it will work on any given system (at
> least Tcl/Tk is linked statically, so that's one package which
> shouldn't cause any problems).
I would be very interested in trying your binary on my Cygwin
installations, which I have been using to build WinGRASS. I am
suspicious that I have been over zealous with updating my system.
However I can not read from grass.itc.it/incoming. Is it write-only?
I believe so.
Is there an alternative method by which I could fetch your binary?
I presume that it will be moved to a publicly-accessible location in
due course.
The usual setup is that uploaded files require human intervention
before they become available for download (in order to prevent the
system from being used for trading warez/porn/etc). Hopefully that
will happen sooner rather than later.
Or would you be able to email or ftp it to me?
If it doesn't become available on grass.itc.it tomorrow, I'll email
you a copy (even stripped and bzip'd, it's around 700kb, as Tcl/Tk is
statically linked).
--
Glynn Clements <glynn.clements@virgin.net>
Glynn Clements wrote:
If it doesn't become available on grass.itc.it tomorrow, I'll email
you a copy (even stripped and bzip'd, it's around 700kb, as Tcl/Tk is
statically linked).
Thanks Glynn. The size should not be a problem for my incoming SMTP. I do not know whether my ability, or inability, to run your binary would tell us much, but I would still be interested in trying. A month or so ago you listed the versions of Cygwin, XFree, gcc, etc. that you had used to successfully build and execute NVIZ, all of which were older than I had. I "downgraded" gcc and a couple other things, but I was still not not able to execute any version of NVIZ that I had compiled. I was not able to find older versions of XFree. Cygwin is a hell of a lot easier to "upgrade" than to "downgraded", and I really do not know if my problems are version related or elsewhere.
Thanks again,
Rich
--
Richard Greenwood
www.greenwoodmap.com
On Tue, Mar 02, 2004 at 08:47:23PM +0000, Glynn Clements wrote:
Michael Barton wrote:
> I have a couple of questions and will try to save a tiny bit of
> bandwidth by combining them.
>
> 1. Has anyone had NVIZ correctly running on a version of WinGRASS?
Yes. Well, I haven't tested it extensively, but the basics definitely
work.
> If so, can this version be uploaded to the WinGRASS binary site?
I've uploaded my NVWISH2.2 binary (from 5.3) to grass.itc.it/incoming,
although I can't guarantee that it will work on any given system (at
least Tcl/Tk is linked statically, so that's one package which
shouldn't cause any problems).
Were to move that file in the grass.itc.it directory tree?
Markus
On Tue, Mar 02, 2004 at 07:52:46PM -0700, Richard Greenwood wrote:
Glynn Clements wrote:
>I've uploaded my NVWISH2.2 binary (from 5.3) to grass.itc.it/incoming,
>although I can't guarantee that it will work on any given system (at
>least Tcl/Tk is linked statically, so that's one package which
>shouldn't cause any problems).
I would be very interested in trying your binary on my Cygwin
installations, which I have been using to build WinGRASS. I am
suspicious that I have been over zealous with updating my system.
However I can not read from grass.itc.it/incoming. Is it write-only?
Right. Otherwise we would find tons of bad things in that directory.
Is there an alternative method by which I could fetch your binary? Or would
you be able to email or ftp it to me?
Once I know were to put it it will be available on grass.itc.it (and the
mirrors).
Markus
On Wed, 3 Mar 2004, Markus Neteler wrote:
On Tue, Mar 02, 2004 at 08:47:23PM +0000, Glynn Clements wrote:
>
> Michael Barton wrote:
>
> > I have a couple of questions and will try to save a tiny bit of
> > bandwidth by combining them.
> >
> > 1. Has anyone had NVIZ correctly running on a version of WinGRASS?
>
> Yes. Well, I haven't tested it extensively, but the basics definitely
> work.
>
> > If so, can this version be uploaded to the WinGRASS binary site?
>
> I've uploaded my NVWISH2.2 binary (from 5.3) to grass.itc.it/incoming,
> although I can't guarantee that it will work on any given system (at
> least Tcl/Tk is linked statically, so that's one package which
> shouldn't cause any problems).
Were to move that file in the grass.itc.it directory tree?
I would say put it in /grass5/binary/windows_cygnus/wingrass_xserver/
I also remember that Glynn said it was working on Windows 98. The GRASS
windows instructions say however that you must install it on an NTFS
filing system so that means NT or 2000 or newer will be what most people
tried it on. Actually I use it on a FAT32 partition---there are no major
problems but might be obscure ones.
Paul
On Wed, Mar 03, 2004 at 02:44:44PM +0000, Paul Kelly wrote:
On Wed, 3 Mar 2004, Markus Neteler wrote:
> On Tue, Mar 02, 2004 at 08:47:23PM +0000, Glynn Clements wrote:
> >
> > Michael Barton wrote:
> >
> > > I have a couple of questions and will try to save a tiny bit of
> > > bandwidth by combining them.
> > >
> > > 1. Has anyone had NVIZ correctly running on a version of WinGRASS?
> >
> > Yes. Well, I haven't tested it extensively, but the basics definitely
> > work.
> >
> > > If so, can this version be uploaded to the WinGRASS binary site?
> >
> > I've uploaded my NVWISH2.2 binary (from 5.3) to grass.itc.it/incoming,
> > although I can't guarantee that it will work on any given system (at
> > least Tcl/Tk is linked statically, so that's one package which
> > shouldn't cause any problems).
>
> Were to move that file in the grass.itc.it directory tree?
>
I would say put it in /grass5/binary/windows_cygnus/wingrass_xserver/
Done.
Markus
Markus Neteler wrote:
> > If so, can this version be uploaded to the WinGRASS binary site?
>
> I've uploaded my NVWISH2.2 binary (from 5.3) to grass.itc.it/incoming,
> although I can't guarantee that it will work on any given system (at
> least Tcl/Tk is linked statically, so that's one package which
> shouldn't cause any problems).
Were to move that file in the grass.itc.it directory tree?
Well, it's a Cygwin binary which requires X, so probably:
grass5/binary/windows_cygnus/wingrass_xserver/
--
Glynn Clements <glynn.clements@virgin.net>
Markus Neteler wrote:
Were to move that file in the grass.itc.it directory tree?
I would say put it in /grass5/binary/windows_cygnus/wingrass_xserver/
Done.
I am looking in:
http://grass.ibiblio.org/grass5/binary/windows_cygnus/wingrass_xserver/
but I do not see it.
--
Richard Greenwood
www.greenwoodmap.com
Richard Greenwood wrote:
>>>Were to move that file in the grass.itc.it directory tree?
>>
>>I would say put it in /grass5/binary/windows_cygnus/wingrass_xserver/
>
> Done.
I am looking in:
http://grass.ibiblio.org/grass5/binary/windows_cygnus/wingrass_xserver/
but I do not see it.
It's on the main GRASS site:
http://grass.itc.it/grass5/binary/windows_cygnus/wingrass_xserver/NVWISH2.2.bz2
[grass.ibiblio.org is a mirror; grass.itc.it is the master site.]
--
Glynn Clements <glynn.clements@virgin.net>
On Wed, Mar 03, 2004 at 08:08:00PM -0700, Richard Greenwood wrote:
Markus Neteler wrote:
>>>Were to move that file in the grass.itc.it directory tree?
>>
>>I would say put it in /grass5/binary/windows_cygnus/wingrass_xserver/
>
>Done.
I am looking in:
http://grass.ibiblio.org/grass5/binary/windows_cygnus/wingrass_xserver/
but I do not see it.
Rich,
keep in mind that mirroring takes up to 24h for the tier-1 mirrors.
Today it's there.
Cheers
Markus
Hello list members,
Richard and me are currently testing some recent compilation
of 5.7 under Cygwin. A strange error occurs:
On Mon, Mar 29, 2004 at 11:37:54AM -0700, Richard Greenwood wrote:
cp /cygdrive/e/projects/grass/grass57_cvs/dist.i686-pc-cygwin/driver/db/dbf /cygdrive/e/projects/grass/grass57_cvs/dist.i686-pc-cygwin/driver/db/shp
cp: `/cygdrive/e/projects/grass/grass57_cvs/dist.i686-pc-cygwin/driver/db/dbf' and `/cygdrive/e/projects/grass/grass57_cvs/dist.i686-pc-cygwin/driver/db/shp' are the same file
make[3]: ***
[/cygdrive/e/projects/grass/grass57_cvs/dist.i686-pc-cygwin/driver/db/dbf] Error 1
make[3]: Leaving directory `/cygdrive/e/projects/grass/grass57_cvs/db/drivers/dbf'
make[2]: *** [subdirs] Error 1
make[2]: Leaving directory `/cygdrive/e/projects/grass/grass57_cvs/db/drivers'
make[1]: *** [subdirs] Error 1
make[1]: Leaving directory `/cygdrive/e/projects/grass/grass57_cvs/db'
make: *** [default] Error 1
I'm a bit clueless since it should simply copy a file to a new filename:
tail -3 ~/grass57/db/drivers/dbf/Makefile
$(CC) $(LDFLAGS) -o $@ $(ARCH_OBJS) $(LIBES) $(MATHLIB) $(XDRLIB)
cp $(DRIVER) $(SHPDRIVER)
Any ideas?
Markus
Markus Neteler wrote:
Richard and me are currently testing some recent compilation
of 5.7 under Cygwin. A strange error occurs:
cp:
`/cygdrive/e/projects/grass/grass57_cvs/dist.i686-pc-cygwin/driver/db/dbf'
and
`/cygdrive/e/projects/grass/grass57_cvs/dist.i686-pc-cygwin/driver/db/shp'
are the same file
Is this error correct? I.e. is .../db/shp a link (hard or symbolic) to
.../db/dbf?
I'm a bit clueless since it should simply copy a file to a new filename:
tail -3 ~/grass57/db/drivers/dbf/Makefile
$(CC) $(LDFLAGS) -o $@ $(ARCH_OBJS) $(LIBES) $(MATHLIB) $(XDRLIB)
cp $(DRIVER) $(SHPDRIVER)
Any ideas?
According to the error message, the new filename refers to the same
file as the old filename, so you're asking "cp" to copy a file onto
itself, which it refuses to do.
More generally, Makefiles should normally use $(INSTALL) rather than
"cp". The "install" program (or autoconf's install-sh script on
systems which don't have an install program) will handle this
situation[1] somewhat better than "cp", as it replaces an existing
file whereas "cp" will attempt to overwrite it ("cp" behaves similarly
to "cat srcfile > dstfile").
[1] It will also handle the situation where the destination file is
read-only, or where the destination is the executable for some
process.
--
Glynn Clements <glynn.clements@virgin.net>
Hi again,
I got GRASS 5.7 sort of working under Cygwin
(see my new install docs at:
http://grass.itc.it/grass51/binary/mswindows_cygwin/
-> cygwin_grass.html
)
but found a small problem. While d.mon is able to handle
the XDRIVER.exe file (note the .exe extension under Cygwin),
the DBMI tools are unable to handle the DBF driver (named dbf.exe
in $GISBASE/driver/db/dbf.exe ). When I rename the file to
'dbf' only it works fine.
Any ideas what to change to allow for 'dbf.exe' as well?
Markus
On Tue, Mar 30, 2004 at 04:22:38PM +0100, Glynn Clements wrote:
Markus Neteler wrote:
> Richard and me are currently testing some recent compilation
> of 5.7 under Cygwin. A strange error occurs:
> cp:
> `/cygdrive/e/projects/grass/grass57_cvs/dist.i686-pc-cygwin/driver/db/dbf'
> and
> `/cygdrive/e/projects/grass/grass57_cvs/dist.i686-pc-cygwin/driver/db/shp'
> are the same file
Is this error correct? I.e. is .../db/shp a link (hard or symbolic) to
.../db/dbf?
Actually I don't understand the error. It's copying the file,
why does it fail? The command in the Makefile is 'cp'. Quite strange,
but I cannot check from here.
> I'm a bit clueless since it should simply copy a file to a new filename:
>
> tail -3 ~/grass57/db/drivers/dbf/Makefile
> $(CC) $(LDFLAGS) -o $@ $(ARCH_OBJS) $(LIBES) $(MATHLIB) $(XDRLIB)
> cp $(DRIVER) $(SHPDRIVER)
>
> Any ideas?
According to the error message, the new filename refers to the same
file as the old filename, so you're asking "cp" to copy a file onto
itself, which it refuses to do.
More generally, Makefiles should normally use $(INSTALL) rather than
"cp". The "install" program (or autoconf's install-sh script on
systems which don't have an install program) will handle this
situation[1] somewhat better than "cp", as it replaces an existing
file whereas "cp" will attempt to overwrite it ("cp" behaves similarly
to "cat srcfile > dstfile").
Ah, so quite a few Makefiles (5.3 as well?) will need such changes.
[1] It will also handle the situation where the destination file is
read-only, or where the destination is the executable for some
process.
Sounds good. Thanks for the insights.
Markus
On Tuesday 30 March 2004 18:04, you wrote:
Hi again,
I got GRASS 5.7 sort of working under Cygwin
(see my new install docs at:
http://grass.itc.it/grass51/binary/mswindows_cygwin/
-> cygwin_grass.html
)
but found a small problem. While d.mon is able to handle
the XDRIVER.exe file (note the .exe extension under Cygwin),
the DBMI tools are unable to handle the DBF driver (named dbf.exe
in $GISBASE/driver/db/dbf.exe ). When I rename the file to
'dbf' only it works fine.
Any ideas what to change to allow for 'dbf.exe' as well?
mon.start is probably using this execl:
#ifdef __W98__
#include <process.h>
#define execl(fullpath,path,name,bg,link,nul) \
do { \
spawnl(_P_DETACH,fullpath,path,name,"-",link,nul); \
return 0; \
} while (0)
#endif /* __W98__ */
-> G_execl() ?
Radim
Markus Neteler wrote:
I got GRASS 5.7 sort of working under Cygwin
(see my new install docs at:
http://grass.itc.it/grass51/binary/mswindows_cygwin/
-> cygwin_grass.html
)
but found a small problem. While d.mon is able to handle
the XDRIVER.exe file (note the .exe extension under Cygwin),
the DBMI tools are unable to handle the DBF driver (named dbf.exe
in $GISBASE/driver/db/dbf.exe ). When I rename the file to
'dbf' only it works fine.
Any ideas what to change to allow for 'dbf.exe' as well?
AFAICT, the problem is in the "dbmscap" code (in
lib/db/dbmi_base/dbmscap.c), which scans the $GISBASE/driver/db
directory to generate the list of drivers. If the driver executable is
called "dbf.exe", then the driver name (for DB_DRIVER, driver= etc)
will be "dbf.exe", not "dbf".
--
Glynn Clements <glynn.clements@virgin.net>