From hamish_nospam at yahoo.com Mon Sep 3 07:56:20 2007 From: hamish_nospam at yahoo.com (Hamish) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] Re: [GRASS-user] grass.bat script for Cygwin GRASS 6.2.2-1 In-Reply-To: <46D7C3A5.8000709@infinito.it> References: <000f01c7e89d$e7453350$24476395@VALUED976296A0> <46D6FF3E.5040101@infinito.it> <20070831154414.1c0ee375.hamish_nospam@yahoo.com> <46D7C3A5.8000709@infinito.it> Message-ID: <20070903235620.949cefcc.hamish_nospam@yahoo.com> Luigi Ponti wrote: > >> Installation instructions at > >> .. > >> Starting GRASS ... > >> Application initialization failed: this isn't a Tk > >> applicationcouldn't connect to display "localhost:0" > >> Error in startup script: dynamic loading is not currently available > >> on this system > >> while executing > >> "load /usr/local/lib/tk8.4/../libtk8.4.a Tk" > > .. Hamish: > > Did you modify the bash line to run startx, launching X at startup? > > bash --login -i -c startx > > Yes, I did. Here is my cygwin_grass.bat file: .. > the next thing I see is the GRASS startup message > > Cleaning up temporary files..... > Starting GRASS ... > > And no X runs. GRASS starts in text mode but then it tries to load a > GUI and off course this fails. No idea as to what is going wrong. The plan is that cygwin starts X11 on startup (-c startx) in the cygwin dos-box; then X11 opens a xterm when it starts; and that xterm launches grass. The '-n "$DISPLAY"' part should protect against it trying to start grass from the cygwin dos-box outside of X11. What does: echo $DISPLAY say from the cygwin dos-box prompt (not the xterm). ? Hamish From lponti at infinito.it Mon Sep 3 14:44:24 2007 From: lponti at infinito.it (Luigi Ponti) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] Re: [GRASS-user] grass.bat script for Cygwin GRASS 6.2.2-1 In-Reply-To: <20070903235620.949cefcc.hamish_nospam@yahoo.com> References: <000f01c7e89d$e7453350$24476395@VALUED976296A0> <46D6FF3E.5040101@infinito.it> <20070831154414.1c0ee375.hamish_nospam@yahoo.com> <46D7C3A5.8000709@infinito.it> <20070903235620.949cefcc.hamish_nospam@yahoo.com> Message-ID: <46DC5608.9070104@infinito.it> Hamish wrote: > Luigi Ponti wrote: > >>>> Installation instructions at >>>> >>>> > .. > >>>> Starting GRASS ... >>>> Application initialization failed: this isn't a Tk >>>> applicationcouldn't connect to display "localhost:0" >>>> Error in startup script: dynamic loading is not currently available >>>> on this system >>>> while executing >>>> "load /usr/local/lib/tk8.4/../libtk8.4.a Tk" >>>> >>> .. >>> > Hamish: > >>> Did you modify the bash line to run startx, launching X at startup? >>> bash --login -i -c startx >>> >> Yes, I did. Here is my cygwin_grass.bat file: >> > .. > >> the next thing I see is the GRASS startup message >> >> Cleaning up temporary files..... >> Starting GRASS ... >> >> And no X runs. GRASS starts in text mode but then it tries to load a >> GUI and off course this fails. >> > > > No idea as to what is going wrong. > > The plan is that cygwin starts X11 on startup (-c startx) in the cygwin > dos-box; then X11 opens a xterm when it starts; and that xterm launches > grass. The '-n "$DISPLAY"' part should protect against it trying to start > grass from the cygwin dos-box outside of X11. What does: > echo $DISPLAY > say from the cygwin dos-box prompt (not the xterm). > > ? > Hamish It says 127.0.0.1:0.0 Nice cue: I had a $DISPLAY set via the winXP control panel that was apparently interfering with the if [ -n "$DISPLAY" ] && [ -n "$GRASS_AUTOSTART" ] ; then switch at the bottom of .bashrc (this is probably why GRASS would start before the XTerm was running, as you said. Now I understand the rationale behind it). As an aside, if I run a cygwin_grass_modified.bat script that is basically grass61.bat (and all related environmental variables that are set therein) except that I deleted the run xterm ... line and added set GRASS_AUTOSTART=1 bash --login -i -c startx what happens is that GRASS runs inside dos-box (or at least this what the shell looks like -- the title bar says "C:\WINDOWS\system32\cmd.exe") but since there is an X Server running, the GUI loads OK and faster than with the regular cygwin_grass.bat file. Apparently, the lower loading time is due to no XTerm starting up in the latter case -- there is only one shell open. Don't know if this can be of any use. Unrelated thing: in the process I also found out that in order for the g.manual script to display cygwin/GRASS HTML help files, the cygwin_root environmental variable needs to be set up in cygwin_grass.bat SET CYGWIN_ROOT=c:\cygwin otherwise it is empty (it used to be set up by grass61.bat, but not by cygwin_grass.bat) when g.manual calls it. (Again, from a previous post, there is a lot of missing HTML help files in the version I downloaded -- 6.2.2-1: is it just a download error?) Thanks and regards, Luigi -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/grass-windows/attachments/20070903/e8cb8a64/attachment.html From tnebitko at can.rogers.com Tue Sep 4 08:21:07 2007 From: tnebitko at can.rogers.com (Ted Nebitko) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] native winGRASS pre-alpha testers needed Message-ID: <000e01c7eeee$12e2f780$9e476395@VALUED976296A0> Is there an index of those modules that are working and those that require further testing? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/grass-windows/attachments/20070904/19517596/attachment.html From fx.prunayre at oieau.fr Tue Sep 4 12:19:13 2007 From: fx.prunayre at oieau.fr (=?iso-8859-1?Q?Fran=E7ois-Xavier_PRUNAYRE?=) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] Starting grass command from cmd.exe Message-ID: <200709041622.l84GMK6G003219@ns1.oieau.fr> Hi list, I?m new to grass on windows and I?m wandering if there?s any way to have grass command running directly from cmd.exe without starting grass. I?ve this up&running on linux setting env variable. I tried the same on windows but can?t get it works. :-( I added : C:\grass\grass-6.3.cvs\bin and C:\grass\grass-6.3.cvs\lib to PATH Defined : GISBASE, GRASS_GUI, LOCATION_NAME and MAPSET variables Any ideas ? Thanks a lot for your help. Francois -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/grass-windows/attachments/20070904/32d06feb/attachment.html From neteler at itc.it Tue Sep 4 16:02:42 2007 From: neteler at itc.it (Markus Neteler) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] Starting grass command from cmd.exe In-Reply-To: <200709041622.l84GMK6G003219@ns1.oieau.fr> References: <200709041622.l84GMK6G003219@ns1.oieau.fr> Message-ID: <12485585.post@talk.nabble.com> In recent CVS versions, a grass63.bat file is generated. It is based on this template: http://freegis.org/cgi-bin/viewcvs.cgi/*checkout*/grass6/lib/init/grass.bat?rev=HEAD Maybe you can pick relevant entries from that file. Markus Fran?ois-Xavier PRUNAYRE wrote: > > Hi list, I?m new to grass on windows and I?m wandering if there?s any way > to > have grass command running directly from cmd.exe without starting grass. > > > > I?ve this up&running on linux setting env variable. > > I tried the same on windows but can?t get it works. :-( > > > > I added : C:\grass\grass-6.3.cvs\bin and C:\grass\grass-6.3.cvs\lib to > PATH > > Defined : GISBASE, GRASS_GUI, LOCATION_NAME and MAPSET variables > > > Any ideas ? > > Thanks a lot for your help. > > > Francois > -- View this message in context: http://www.nabble.com/Starting-grass-command-from-cmd.exe-tf4378808.html#a12485585 Sent from the Grass - Win mailing list archive at Nabble.com. From glynn at gclements.plus.com Tue Sep 4 18:36:12 2007 From: glynn at gclements.plus.com (Glynn Clements) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] Starting grass command from cmd.exe In-Reply-To: <200709041622.l84GMK6G003219@ns1.oieau.fr> References: <200709041622.l84GMK6G003219@ns1.oieau.fr> Message-ID: <18141.56796.864280.775834@cerise.gclements.plus.com> Fran?ois-Xavier PRUNAYRE wrote: > Hi list, I?m new to grass on windows and I?m wandering if there?s any way to > have grass command running directly from cmd.exe without starting grass. > > I?ve this up&running on linux setting env variable. > > I tried the same on windows but can?t get it works. :-( > > I added : C:\grass\grass-6.3.cvs\bin and C:\grass\grass-6.3.cvs\lib to PATH > > Defined : GISBASE, GRASS_GUI, LOCATION_NAME and MAPSET variables > > Any ideas ? Use the same mechanism as on Linux; just change the syntax for setting environment variables from bash's "export" to cmd.exe's "set". Apart from PATH (which needs to contain the bin, scripts and lib subdirectories), you also need to set GISBASE and GISRC. Note that GISDBASE, LOCATION_NAME and MAPSET are GRASS variables, set in the file referenced by %GISRC%, not environment variables. -- Glynn Clements From fx.prunayre at oieau.fr Wed Sep 5 02:20:47 2007 From: fx.prunayre at oieau.fr (=?UTF-8?Q?Fran=C3=A7ois-Xavier_PRUNAYRE?=) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] Starting grass command from cmd.exe In-Reply-To: <18141.56796.864280.775834@cerise.gclements.plus.com> Message-ID: <200709050623.l856NsUu010101@ns1.oieau.fr> Thanks a lot Glynn, you're right, I made a gisrc file with LOCATION_NAME: data MAPSET: user1 DIGITIZER: none GISDBASE: c:\grass\data OVERWRITE: 1 GRASS_GUI: text Then start the command line and set set GISBASE=c:\grass\grass-6.3.cvs set GISRC=c:\grass\gisrc set PATH=%PATH%;C:\grass\grass-6.3.cvs\bin;C:\grass\grass-6.3.cvs\lib;C:\grass\grass-6.3.cvs\scripts;C:\grass\grass-6.3.cvs\etc And that's it. Cheers. Francois. -----Message d'origine----- De : Glynn Clements [mailto:glynn@gclements.plus.com] Envoy? : mercredi 5 septembre 2007 00:36 ? : Fran?ois-Xavier PRUNAYRE Cc : wingrass@grass.itc.it Objet : Re: [winGRASS] Starting grass command from cmd.exe Fran?is-Xavier PRUNAYRE wrote: > Hi list, I? new to grass on windows and I? wandering if there? any way to > have grass command running directly from cmd.exe without starting grass. > > I?e this up&running on linux setting env variable. > > I tried the same on windows but can? get it works. :-( > > I added : C:\grass\grass-6.3.cvs\bin and C:\grass\grass-6.3.cvs\lib to PATH > > Defined : GISBASE, GRASS_GUI, LOCATION_NAME and MAPSET variables > > Any ideas ? Use the same mechanism as on Linux; just change the syntax for setting environment variables from bash's "export" to cmd.exe's "set". Apart from PATH (which needs to contain the bin, scripts and lib subdirectories), you also need to set GISBASE and GISRC. Note that GISDBASE, LOCATION_NAME and MAPSET are GRASS variables, set in the file referenced by %GISRC%, not environment variables. -- Glynn Clements From mlennert at club.worldonline.be Wed Sep 5 04:18:38 2007 From: mlennert at club.worldonline.be (Moritz Lennert) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] native winGRASS pre-alpha testers needed In-Reply-To: <000e01c7eeee$12e2f780$9e476395@VALUED976296A0> References: <000e01c7eeee$12e2f780$9e476395@VALUED976296A0> Message-ID: <46DE665E.3040209@club.worldonline.be> On 04/09/07 14:21, Ted Nebitko wrote: > Is there an index of those modules that are working and those that > require further testing? There is no such systematic index. What we know not to work at this stage: - some modules dealing with attributes of vector data (confirmed for v.out.ogr, but others are affected): these work intermittently, but often cause deadlocks. We are currently busy trying to hunt down this bug. - all d.* commands from the command line: there are no x-monitors in windows so none of these commands will work. All display has to happen within the GIS Manager gui. - d.vect.thematic: this is a script which cannot easily be ported to windows as it uses d.mon commands. - There might be other scripts that don't work. The rest potentially works, but still needs extensive testing since whether a module works or not might depend on the parameters used. Moritz From mlennert at club.worldonline.be Fri Sep 14 05:59:53 2007 From: mlennert at club.worldonline.be (Moritz Lennert) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] native winGRASS pre-alpha testers needed In-Reply-To: References: Message-ID: <46EA5B99.7050800@club.worldonline.be> Dear Peitro, Pietro Calogero wrote: > Dear Moritz, > I have just subscribed to the winGRASS email list but have not yet > received a confirming email. I hope that the email confirmation will > instruct me how to submit questions to the list, because I have found no > way to do so from the outside. That is why I am emailing you directly. > Sorry; given the required instructions, I will not do this again. I imagine you will soon receive this confirmation mail. In the meantime, don't hesitate to ask me directly. > However I am at a total loss: how do you start winGRASS63? > I have unzipped wingrass63_cvs20070910 onto a testing PC at c:\grass > (rather than in Program Files, to avoid the incompatibility of winGRASS > with spaces in directory names). > I double-clicked on the c:\grass\bin\grass63.bat file, and the cmd > window flickers into and out of existence in a blink. Have you installed Tcl/Tk ? You can get it at http://www.activestate.com/products/activetcl/ > > I assume that this "pre-compiled binary" actually requires considerable > configuration. Actually it doesn't. > I have searched repeatedly through the GRASS website, the > wingrass archives, and Google to find instructions. As far as I can tell > they are not available on the internet. There is a README at the same page you downloaded the binary from: http://moritz.homelinux.org/grass/wingrass/. The main instructions you will need are: "INSTALL - Download wingrass63_cvsXXXXXXXX.zip (XXXXXXXX represents the date at which the current package was compiled) - Unzip (easiest to unzip in C:\, but if not adapt WINGISBASE in grass63.bat - see below). - Install ActiveState Tcl (http://www.activestate.com/products/activetcl/)" and "RUN - double-click on grass/bin/grass63.bat - you can put a shortcut to grass63.bat on you desktop and associate it with the grass logo (grasslogo_vector_small.ico) which you can find in this same directory." > There is plenty of commentary > about grass62 configured to run on Cygwin; that does not seem to apply. > Or perhaps part of it applies, but there is no way for me to figure out > which part applies. No, the Cygwin port functions quite differently. > > As you can see, I have done quite a bit of digging on this, so please do > not brush me off with RTFM. I recommend that such instructions should be > posted on a page such as: > http://grass.itc.it/platforms/wingrass.html > Where installation instructions for native winGRASS63 are conspicuously > absent. Good idea. I'll put parts of the readme there. > Assume that the installer is someone who has never used Cygwin (and will > never do so), and has never installed GRASS on a Windows platform. > Assume the installer can open the cmd window and is willing to use it, > but has no idea how Windows XP is configured under the hood. Normally, the installation is actually extremely easy. Just unzip the downloaded file, install tcl/tk and off you go. However, some elements still don't work correctly, notably vector attribute data access. Moritz From benjamin.ducke at ufg.uni-kiel.de Wed Sep 5 05:36:28 2007 From: benjamin.ducke at ufg.uni-kiel.de (benjamin.ducke@ufg.uni-kiel.de) Date: Wed Nov 14 17:04:10 2007 Subject: [GRASS-user] Re: [winGRASS] native winGRASS pre-alpha testers needed In-Reply-To: <46DE665E.3040209@club.worldonline.be> References: <000e01c7eeee$12e2f780$9e476395@VALUED976296A0> <46DE665E.3040209@club.worldonline.be> Message-ID: <2902.128.40.195.179.1188984988.squirrel@webmail.mail.uni-kiel.de> In addition, GRASS shell scripts and modules should be checked and hardened for cases where spaces may appear in file or directory names. On a unix system, users generally don't store stuff in paths with spaces, but Windows by its very design encourages this. E.g. the user's home folder may well be something like C:\Windows\Documents and Settings\Default User\ This can easily break scripts and modules that need to interact directly with the filesystem, e.g. for reading and writing ASCII files. If the user installs the WinGRASS binaries into a path with spaces, also something Windows encourages, there might be further sources of trouble. So I'd recommend to do just that: install WinGRASS into a dir with spaces, create a GRASS mapset, also in a path with spaces, and work from there to find all the vulnerable bits and pieces. Benjamin > On 04/09/07 14:21, Ted Nebitko wrote: >> Is there an index of those modules that are working and those that >> require further testing? > > There is no such systematic index. What we know not to work at this stage: > > - some modules dealing with attributes of vector data (confirmed for > v.out.ogr, but others are affected): these work intermittently, but > often cause deadlocks. We are currently busy trying to hunt down this bug. > > - all d.* commands from the command line: there are no x-monitors in > windows so none of these commands will work. All display has to happen > within the GIS Manager gui. > > - d.vect.thematic: this is a script which cannot easily be ported to > windows as it uses d.mon commands. > > - There might be other scripts that don't work. > > > The rest potentially works, but still needs extensive testing since > whether a module works or not might depend on the parameters used. > > > Moritz > > _______________________________________________ > grassuser mailing list > grassuser@grass.itc.it > http://grass.itc.it/mailman/listinfo/grassuser > > From neteler at itc.it Fri Sep 14 12:25:20 2007 From: neteler at itc.it (Markus Neteler) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] native winGRASS pre-alpha testers needed In-Reply-To: <46EA5B99.7050800@club.worldonline.be> References: <46EA5B99.7050800@club.worldonline.be> Message-ID: <12679025.post@talk.nabble.com> Pietro Calogero wrote: > > Dear Moritz, > I have just subscribed to the winGRASS email list but have not yet > received a confirming email. I hope that the email confirmation will > instruct me how to submit questions to the list, because I have found no > way to do so from the outside. ... > [Moritz, can you fwd this to Pietro?] Hi Pietro, the system is fully automated and should respond in less than one minute. Did you spam filter eat the confirmation message? Please contact me personally if needed. Markus -- View this message in context: http://www.nabble.com/Re%3A-native-winGRASS-pre-alpha-testers-needed-tf4441480.html#a12679025 Sent from the Grass - Win mailing list archive at Nabble.com. From iuri at ufrgs.br Fri Sep 14 22:32:38 2007 From: iuri at ufrgs.br (Iuri Gavronski) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] Performance comparisons between WinGRASS 6.2 and 6.3 Message-ID: <1daa4e8d0709141932k51fb760fp9bb2cc3322ee4915@mail.gmail.com> Hi, Has anyone made a performance comparison between the GRASS 6.2 running under Cygwin or GRASS 6.3 running natively? Is it possible to have both versions running on the same machine? Thanks, Iuri. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/grass-windows/attachments/20070914/c1a0da35/attachment.html From lponti at infinito.it Sat Sep 15 03:54:14 2007 From: lponti at infinito.it (Luigi Ponti) Date: Wed Nov 14 17:04:10 2007 Subject: [winGRASS] Performance comparisons between WinGRASS 6.2 and 6.3 In-Reply-To: <1daa4e8d0709141932k51fb760fp9bb2cc3322ee4915@mail.gmail.com> References: <1daa4e8d0709141932k51fb760fp9bb2cc3322ee4915@mail.gmail.com> Message-ID: <46EB8FA6.4050205@infinito.it> Iuri Gavronski wrote: > > > Is it possible to have both versions running on the same machine? > Please, see this post http://www.nabble.com/running-cygwin-grass-6.1,-6.2-and-native-6.3-on-the-same-xp-machine-t4234845.html Regards, Luigi