[GRASS-dev] Re: gis.m, thematic layer: You must open a display monitor

[Moritz Lennert]

On 13/02/07 16:06, Michael Barton wrote:
> There should be a flag to send output to the gis manager. This flag should
> cause this conditional to be skipped.

The flag is $GIS_FLAG_S:

On line 860ff you have this:
# reset rendering for GUI use
      if [ "$GIS_FLAG_S"  -eq 1 ] ; then
[...]
        d.mon select=$currmon
[...]
fi

In my understanding, unless someone works in parallel with GIS Manager 
and x-mons, $currmon will be empty and this is where this error message 
comes from.

Maybe it's just a problem of an unequal number of if and fi calls...

But I don't have the time to go into the file in detail now...sorry.

Moritz


> 
> Michael
> 
> 
> On 2/13/07 7:58 AM, "Moritz Lennert" <mlennert at club.worldonline.be> wrote:
> 
>> On 13/02/07 15:29, Michael Barton wrote:
>>> I have no idea. Perhaps I'm imagining things, but I swear that I'm seeing a
>>> LOT more of these messages recently--with their deleterious effects on
>>> scripting. Hamish says they've been there all along. I'm suspecting that the
>>> code for the message has existed, but they did not get parsed for some
>>> reason. But that is just speculation.
>> Well in this case, there is a change by Daniel dated Jan. 28 which might
>> be linked:
>>
>> - Added quotes in occurences of 'if [ -z "$var" ]'
>>
>> But reverting back to no quotes gives the same error message.
>>
>> Looking at the logic, I don't actually understand how this worked
>> before. As I said, IIUC the x-mon test seems to be done whether you are
>> using GIS Manager or not. But under GIS Manager there should be no x-mon...
>>
>> Conditonalising the check as follows:
>>
>> if [ "$GIS_FLAG_S"  -eq 0 ] ; then
>>    if [ -z "$currmon" ] ; then
>>          echo ""
>>          echo "*** You must open a display monitor ***"
>>          echo ""
>>          cleanup
>>          exit 2
>>    fi
>> fi
>>
>> makes the error message go away, but I get a new one:
>>
>> Usage:  /usr/lib/grass/etc/mon.select monitor_name
>>
>> This time, however, the map and legend are displayed correctly (without
>> the "fix" only the legend is displayed). I can get this second error
>> message to go away by commenting line 862:
>>
>> #               d.mon select=$currmon
>>
>> This comes from the fact that $currmon is empty in GIS Manager. Again, I
>> don't understand why we have to deal with d.mon here if we are in the
>> GIS Manager...and why this did not cause any problems before.
>>
>> Moritz
>>
>>
>>> Michael
>>>
>>>
>>> On 2/13/07 6:19 AM, "Moritz Lennert" <mlennert at club.worldonline.be> wrote:
>>>
>>>> Michael,
>>>>
>>>> I am seeing a new error for d.vect.thematic. Trying to display a
>>>> thematic layer fails with the message: *** You must open a display
>>>> monitor ***
>>>>
>>>> I can reproduce this with any spearfish vector layer.
>>>>
>>>> The problem seems to be in lines 336-345 where it checks for the
>>>> existance of an x-monitor. It seems to do this unconditionally, so even
>>>> if you run it from the GIS Manager it still checks and fails if there is
>>>> not monitor.
>>>>
>>>> I can't find anything recent in the commit logs that would explain this,
>>>> so before trying to go deeper into it, I wanted to know if you had an
>>>> idea why this is suddenly happening.
>>>>
>>>> Moritz
>>> __________________________________________
>>> Michael Barton, Professor of Anthropology
>>> School of Human Evolution & Social Change
>>> Center for Social Dynamics & Complexity
>>> Arizona State University
>>>
>>> phone: 480-965-6213
>>> fax: 480-965-7671
>>> www: http://www.public.asu.edu/~cmbarton
>>>
>>>
> 
> __________________________________________
> Michael Barton, Professor of Anthropology
> School of Human Evolution & Social Change
> Center for Social Dynamics & Complexity
> Arizona State University
> 
> phone: 480-965-6213
> fax: 480-965-7671
> www: http://www.public.asu.edu/~cmbarton
> 
>