[GRASSLIST:57] Re: [GRASS5] Re: New GIS Manager adds transparency to layers

Michael Barton michael.barton at asu.edu
Mon Mar 13 15:32:53 EST 2006 

Thomas,

I'm not quite sure where the problem lies, but am glad that you solved it.
I'm simply using the default working GRASS .tmp directory that you get when
you run g.tempfile pid=$$ from the command line. This is located in:

$GISBASE/$LOCATION_NAME/$MAPSET/.tmp/[devicename]/

I use this because it *ought* to be standard and writeable on any system.

In order to neatly put things back the way you set them, I also identify the
current directory (i.e., the one you started GRASS from) so that I can cd
back to that directory when finished writing and reading display files.

Is the error you get a TclTk error? If so, please press the detail button,
copy the text, and send it to me. I might be able to help troubleshoot your
problem.

Michael


On 3/13/06 11:21 AM, "Thomas Adams" <Thomas.Adams at noaa.gov> wrote:

> Michael,
> 
> Sorry; I should have tried this first… no problem when I startup from my
> $HOME (login) directory. Typically I like to startup GRASS from my GRASS
> database directory. Is this the anticipated behavior or should it be
> considered a 'bug'?
> 
> Thanks for your feedback and great work!
> 
> Tom
> 
> 
> Thomas Adams wrote:
>> Michael,
>> 
>> Where is GRASS/new display manager going to look for the .tmp
>> directory? In the directory GRASS is started from; one's $HOME
>> directory; other?
>> 
>> Thanks,
>> Tom
>> 
>> 
>> Michael Barton wrote:
>>> Something similar has just been reported on some lab computers
>>> running GRASS
>>> under Cygwin. Maybe it's a different issue, though because it only
>>> happened
>>> with a version of init.sh that automatically runs gism2. It seems to
>>> run OK
>>> if launched manually.
>>> 
>>> Your GRASS in a non-standard location might be an issue. I'm not sure
>>> where
>>> the working directory error comes from. Is it a TclTk error (buttons
>>> run,
>>> cancel, details) or another kind of error (xwindows, bash, etc..)?
>>> 
>>> The GIS Manager display works by creating PPM files in a temp directory
>>> within your working GRASS mapset--set by g.tempfile. If you don't have
>>> access to such a .tmp directory, you will have problems.
>>> 
>>> Let me know if any of these gives you a clue to what is wrong.
>>> 
>>> Michael
>>> 
>>> 
>>> On 3/10/06 7:22 PM, "Thomas Adams" <Thomas.Adams at noaa.gov> wrote:
>>> 
>>>  
>>>> Michael,
>>>> 
>>>> I have installed the GRASS 6.1 CVS release dated 2006_03_04
>>>> 
>>>> and followed your instructions in the README in
>>>> _http://www.public.asu.edu/~cmbarton/files/grass_gismgr/
>>>> <http://www.public.asu.edu/%7Ecmbarton/files/grass_gismgr/>_gism_rc4_200603
>>>> 07. 
>>>> 
>>>> tgz.
>>>> 
>>>> First off, the installation using the old d.m works fine ‹ no errors.
>>>> When I try the new gis.m, the interface tries to come up, but it stalls
>>>> with an error Dialog that reads "Error getting working directory name:
>>>> no such file or directory". If I select the Error Dialog 'OK' button,
>>>> the GIS manager completes its startup. However, if I select a raster
>>>> map
>>>> to display and click on the refresh button (labeled: "Display active
>>>> layers in the current region") in the "Map Display 1" window, I get the
>>>> same Error Dialog as I got when I first started up the new GIS Manager
>>>> (gis.m). Also, from the time of the startup of the GIS Manager, the
>>>> 'status indicator' in the lower left of the "Map Display 1" window
>>>> shows: "please wait..." ‹ this never goes away.
>>>> 
>>>> Interestingly, if I select the NVIZ button, there is no problem
>>>> displaying the raster I selected; so, NVIZ seems OK, but the raster map
>>>> will never display in 2-D in the "Map Display 1" windowŠ
>>>> 
>>>> Also, when I try to use the Init.sh file from gism_rc4_20060307.tgz, I
>>>> get shared library problems: "libgrass_gis.so cannot open shared object
>>>> file: No such file or directory" ‹ even though the library is in the
>>>> GRASS lib directory. I should mention that due to system constraints,
>>>> the entire GRASS distribution and other libraries (GDAL, PROJ, etc.)
>>>> are
>>>> in a non-standard location.
>>>> 
>>>> A final question: should I be using the gis.m file from
>>>> gism_rc4_20060307.tgz? I have tried both with and withoutŠ
>>>> 
>>>> I'm running on a Redhat Linux system, FWIWŠ
>>>> 
>>>> Tom
>>>> 
>>>> Michael Barton wrote:
>>>>    
>>>>> I¹ve just finished phase 2 of the GIS Manager update and it¹s pretty
>>>>> exciting. At first glance, it looks a lot like the one I¹ve been
>>>>> working on and have in the CVS now, but with a couple of important
>>>>> exceptions. One is that transparency is now supported for all GRASS
>>>>> layers.
>>>>> 
>>>>> The second will probably generate the most discussion, so I¹m ready to
>>>>> duck and run for cover. Now GRASS is consistent with other GIS
>>>>> programs and more intuitive (as much as I am skeptical about that
>>>>> term) in that the uppermost layer in the layer tree is also the top
>>>>> map in the display. Related to this, new layers insert directly above
>>>>> the currently selected layer instead of at the bottom of the layer
>>>>> tree.
>>>>> 
>>>>> The main changes come Œunder the hood¹ of this version of the GIS
>>>>> Manager. It uses g.pnmcomp to composite a set of maps (each rendered
>>>>> as *.PPM files) into a visual display. One of the results is that now
>>>>> GRASS is significantly faster in adding new layers, turning layers on
>>>>> and off, and changing layer attributes‹only the new or changed layer
>>>>> needs to be rendered; the others will just be re-composited into a
>>>>> display from existing PPM files. This all happens in a GRASS .tmp
>>>>> directory, so that it doesn¹t clutter up your home directories. The
>>>>> other important thing about this new architecture is that it makes
>>>>> GRASS even less dependent on the xdriver and PNGdriver for
>>>>> visualization. GRASS can become more easily portable to different OS¹s
>>>>> and GUI frameworks. It is also more scriptable. You can write a script
>>>>> that will create a complete, multilayered map, and output it to a
>>>>> graphics file.
>>>>> 
>>>>> One consequencer of this new architecture is that the GIS Manager does
>>>>> not currently support display output to PNG files. With GDAL (required
>>>>> for many GRASS features) and ghostscript installed, it does support
>>>>> output to PPM/PNM, TIF, JPG, BMP, EPS, and PDF.
>>>>> 
>>>>> Command line usage is still supported in several ways‹the old
>>>>> xterminal is still there, there is the new command console, and you
>>>>> can add command layers to the layer tree and display them.
>>>>> 
>>>>> You can see a screen shot of semi-transparent soil vector areas
>>>>> overlaying a raster shaded relief map at
>>>>> 
>>>>> _http://www.public.asu.edu/~cmbarton/files/grass_screenshots/gism2-2.jpg
>>>>> 
>>>>> <http://www.public.asu.edu/%7Ecmbarton/files/grass_screenshots/gism2-2.jpg
>>>>> > 
>>>>> 
>>>>> _
>>>>> I¹ve committed all changes to the CVS. For those of you with binary
>>>>> versions of GRASS who would like to try it, I¹ve posted packages you
>>>>> can use to replace your current GIS Manager on my website. For full
>>>>> functionality, you need a version of grass that is dated
>>>>> 21-February-2006 or later. For those of you with a slightly earlier
>>>>> version, I made a second package with all functionality minus
>>>>> transparency. Here are the links.
>>>>> 
>>>>> New GIS Manager files
>>>>> 
>>>>> <_http://www.public.asu.edu/~cmbarton/files/grass_gismgr/
>>>>> <http://www.public.asu.edu/%7Ecmbarton/files/grass_gismgr/>_gism_rc4_20060
>>>>> 307 
>>>>> 
>>>>> .tgz>
>>>>> 
>>>>> New GIS Manager files for people with recent, but pre-21 February 2006
>>>>> binaries.
>>>>> 
>>>>> <_http://www.public.asu.edu/~cmbarton/files/grass_gismgr/
>>>>> <http://www.public.asu.edu/%7Ecmbarton/files/grass_gismgr/>_gism_rc4_20060
>>>>> 307 
>>>>> 
>>>>> pre_02-21.tgz>
>>>>> 
>>>>> Enjoy
>>>>> Michael
>>>>> 
>>>>> ______________________________
>>>>> Michael Barton, Professor of Anthropology
>>>>> School of Human Evolution and Social Change
>>>>> Arizona State University
>>>>> Tempe, AZ 85287-2402
>>>>> USA
>>>>> 
>>>>> voice: 480-965-6262; fax: 480-965-7671
>>>>> www: http://www.public.asu.edu/~cmbarton
>>>>> <http://www.public.asu.edu/%7Ecmbarton>
>>>>>       
>>> 
>>> ___________________________
>>> Michael Barton, Professor of Anthropology
>>> School of Human Evolution & Social Change
>>> Arizona State University
>>> Tempe, AZ  85287
>>> 
>>> WWW - http://www.public.asu.edu/~cmbarton
>>> Phone: 480-965-6262
>>> Fax: 480-965-7671
>>> 
>>>   
>> 
>> 
> 

___________________________
Michael Barton, Professor of Anthropology
School of Human Evolution & Social Change
Arizona State University
Tempe, AZ  85287

WWW - http://www.public.asu.edu/~cmbarton
Phone: 480-965-6262
Fax: 480-965-7671

More information about the grass-user mailing list