Re: [Qgis-developer] Reliable way to determine a groupIndex in the legend tree

2013-10-09 Thread Andreas Neumann 
Am 09.10.2013 06:16, schrieb Bernhard Ströbl:
> Hi
> 
> Am 09.10.2013 08:12, schrieb Denis Rouzaud:
>> Hi,
>>
>> I am not sure we should make any publicity to the legend API which is
>> very bad.
>>
>> I hope these methods will deprecated soon ;)

So is there anyone working on fixing these legend API problems?

We should probably come up with a Wiki page where we could collect ideas
what we would expect from a proper legend API.

Andreas
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [Qgis-developer] Reliable way to determine a groupIndex in the legend tree

2013-10-09 Thread Denis Rouzaud 

I heard some discussions... from Matthias, Martin and Nathan.

but there are probably others!

On 10/09/2013 09:30 AM, Andreas Neumann wrote:

Am 09.10.2013 06:16, schrieb Bernhard Ströbl:

Hi

Am 09.10.2013 08:12, schrieb Denis Rouzaud:

Hi,

I am not sure we should make any publicity to the legend API which is
very bad.

I hope these methods will deprecated soon ;)

So is there anyone working on fixing these legend API problems?

We should probably come up with a Wiki page where we could collect ideas
what we would expect from a proper legend API.

Andreas
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer


___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

[Qgis-developer] PyQGIS API code completion not working in PyCharm

2013-10-09 Thread Jorge Gil 

Hi everyone,

I'm new to QGIS plug-in developemtn and I'm setting up a development 
environment on Mac OS X.


At the moment I'm going with PyCharm and got the remote debugging to 
work. Thanks to this list!


One thing I can't get to work is code completion for the PyQGIS API.

I've added '/Applications/QGIS.app/Contents/Resources/python/qgis/' and 
'/Applications/QGIS.app/Contents/Resources/python/' to the interpreter's 
path. PyCharm can find PyQGIS and ads it to the external libraries. This 
way I can import the qgis modules without errors, in the Python console 
for example.

However, code completion doesn't work in the editor nor in the console.

Thanks in advance for your help,
Jorge


___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

[Qgis-developer] PyQGIS and checking WMS layers have data

2013-10-09 Thread Stott James 
Hi all,

I have a loop in my code that makes PDFs for a number of polygons and the base 
mapping comes from an aerial photography WMS. So for each polygon, the python 
code moves the map to the extent of that polygon and the makes a PDF.

However, some of my maps end up with white backgrounds where the aerial 
photography hasn't loaded/displayed.

Is it possible to make sure that a WMS has loaded its data before a PDF is made 
in python? Or does the problem lie with the network connection at the time that 
PDF is made? Is there any way I can catch these instances and maybe load the 
WMS again?

The last time I ran the script, of the 1465 PDFs it created, about 21 had a 
white base map, so generally it works really really well. I have checked that 
there is actually data for these 21 areas, and there is. It also seems to be 
different areas that do not load each time I run the script.

Many thanks

James
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [Qgis-developer] PyQGIS API code completion not working in PyCharm

2013-10-09 Thread Larry Shaffer 
Hi Jorge,

On Wed, Oct 9, 2013 at 5:44 AM, Jorge Gil  wrote:

> Hi everyone,
>
> I'm new to QGIS plug-in developemtn and I'm setting up a development
> environment on Mac OS X.
>
> At the moment I'm going with PyCharm and got the remote debugging to work.
> Thanks to this list!
>

Welcome to developing with QGIS. PyCharm is an excellent choice for IDE,
especially now that 3.0 has a free community version and more support for
PyQt.


> One thing I can't get to work is code completion for the PyQGIS API.
>
> I've added '/Applications/QGIS.app/**Contents/Resources/python/**qgis/'
> and '/Applications/QGIS.app/**Contents/Resources/python/' to the
> interpreter's path. PyCharm can find PyQGIS and ads it to the external
> libraries. This way I can import the qgis modules without errors, in the
> Python console for example.
> However, code completion doesn't work in the editor nor in the console.
>

You only need the '/Applications/QGIS.app/Contents/Resources/python' path
added (assuming you are using a Kyngchaos.com QGIS install), so I recommend
removing the other path . However, I have noticed that sometimes PyCharm on
Mac does not properly rebuild the 'skeletons' for code completion (I'm
using PyCharm 2.7.3). Try relaunching the app and removing/re-adding the
path, then clicking the reload button. You should see a progress bar and
eventually notice the qgis.* modules being parsed. If not there may be an
issue.

For plugin development, I add external commands, several of which are for
running my plugin's Makefile [0]. While the new PyCharm 3.0 has an embedded
terminal, I find assigning a key shortcut (say Ctrl-M) to the make 
command to be very quick for deploying the plugin to
~/.qgis2/python/plugins/. Deploying is the best-practice means of testing
your plugin, instead of symlinking it, but that also works.

Also, I recommend using the Plugin Reloader PyQGIS plugin to reload your
plugin within QGIS, so you do not have to relaunch QGIS every time you
propagate a change in your plugin.

[0] http://drive.dakotacarto.com/qgis/pycharm-ext-cmd_make.png

Regards,

Larry

Thanks in advance for your help,
> Jorge
>
>
> __**_
> Qgis-developer mailing list
> Qgis-developer@lists.osgeo.org
> http://lists.osgeo.org/**mailman/listinfo/qgis-**developer
>
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

[Qgis-developer] Check

2013-10-09 Thread Paolo Cavallini 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Hi.
Who is responding to this address? I see it is reported as the one of
the maintainer of the Debian package, so I assume Juergen is
listening, right?
All the best, and sorry for the noise.
- -- 
Paolo Cavallini - Faunalia
www.faunalia.eu
Full contact details at www.faunalia.eu/pc
Nuovi corsi QGIS e PostGIS: http://www.faunalia.it/calendario
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.14 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAlJVgnIACgkQ/NedwLUzIr6SegCgmpqDq/SY1nX3+oKPTnsS6MhC
iKkAn2CDAdEYakdrXZZKzYlJ5DQ6+HjW
=ssfh
-END PGP SIGNATURE-
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [Qgis-developer] Check

2013-10-09 Thread Paolo Cavallini 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Il 09/10/2013 18:21, Paolo Cavallini ha scritto:
> Who is responding to this address? I see it is reported as the one
> of the maintainer of the Debian package

oops, sorry, the entire list.
- -- 
Paolo Cavallini - Faunalia
www.faunalia.eu
Full contact details at www.faunalia.eu/pc
Nuovi corsi QGIS e PostGIS: http://www.faunalia.it/calendario
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.14 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAlJVgtYACgkQ/NedwLUzIr6wYwCeNp1oq/bK5VktqhSK0DReb4Wv
qSIAnj0Ee8ryoiTflvBixXemAezKO0gs
=LbnC
-END PGP SIGNATURE-
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [Qgis-developer] vector rendering improvement patch

2013-10-09 Thread kimaidou 
Waou :)
+1 for having some rendering guru reviewing this patch, as it seems
brilliant ;)

Michael


2013/10/8 Paolo Cavallini 

> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> Il 07/10/2013 21:32, Giovanni Manghi ha scritto:
>
> > I would really appreciate if one of the core devs can review it
> > and eventually commit it (or suggest modifications), so we can
> > start test it as this would be a much needed improvement.
>
> agreed, seems worth a close look: anyone available?
> all the best-
>
> - --
> Paolo Cavallini - Faunalia
> www.faunalia.eu
> Full contact details at www.faunalia.eu/pc
> Nuovi corsi QGIS e PostGIS: http://www.faunalia.it/calendario
> -BEGIN PGP SIGNATURE-
> Version: GnuPG v1.4.14 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
>
> iEYEARECAAYFAlJTo+QACgkQ/NedwLUzIr5pqwCeK6+LPv7+XnyrsACT1fDk8lEe
> nhgAoLDs+WwZ+piEqBquakSehNMaxTJL
> =iJfk
> -END PGP SIGNATURE-
> ___
> Qgis-developer mailing list
> Qgis-developer@lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/qgis-developer
>
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

[Qgis-developer] Possible bug in Processing>GRASS

2013-10-09 Thread Paolo Cavallini 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Hi all.
When running a GRASS command through Processing, the import creates a
"cat" column. If the same layer is used as input for another GRASS
commend, the cat column is renamed as cat_, and a new cat column is
created. If this second layer is used once more as input, the renaming
of cat in cat_ fails, because the column already exists. Following
analyses may fail.
Did I interpret it correctly? Should I open a ticket?
All the best.
- -- 
Paolo Cavallini - Faunalia
www.faunalia.eu
Full contact details at www.faunalia.eu/pc
Nuovi corsi QGIS e PostGIS: http://www.faunalia.it/calendario
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.14 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAlJVk9YACgkQ/NedwLUzIr658wCeO4+RSsZBbRsSjtFSA/Vy1CwE
FBIAn3JlvROlUXBCu/2BHYVynxEQlXMF
=k6Yv
-END PGP SIGNATURE-
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [Qgis-developer] Reliable way to determine a groupIndex in the legend tree

2013-10-09 Thread Gino Pirelli 
we did a little face to face meeting in Brighton with Martin D. and Marco
H... QgsLegend classes need a refactory due to structural limitation.

Marco H. did an estimation of the refactoring, but probably, as I can see
there are more that need this refactoring, could be interesting collect
features that this re-write should involve
and obviously funding :)

ciao Luigi Pirelli (luigi.pire...@faunalia.it)



On 9 October 2013 09:35, Denis Rouzaud  wrote:

> I heard some discussions... from Matthias, Martin and Nathan.
>
> but there are probably others!
>
>
> On 10/09/2013 09:30 AM, Andreas Neumann wrote:
>
>> Am 09.10.2013 06:16, schrieb Bernhard Ströbl:
>>
>>> Hi
>>>
>>> Am 09.10.2013 08:12, schrieb Denis Rouzaud:
>>>
 Hi,

 I am not sure we should make any publicity to the legend API which is
 very bad.

 I hope these methods will deprecated soon ;)

>>> So is there anyone working on fixing these legend API problems?
>>
>> We should probably come up with a Wiki page where we could collect ideas
>> what we would expect from a proper legend API.
>>
>> Andreas
>> __**_
>> Qgis-developer mailing list
>> Qgis-developer@lists.osgeo.org
>> http://lists.osgeo.org/**mailman/listinfo/qgis-**developer
>>
>
> __**_
> Qgis-developer mailing list
> Qgis-developer@lists.osgeo.org
> http://lists.osgeo.org/**mailman/listinfo/qgis-**developer
>
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [Qgis-developer] PyQGIS and checking WMS layers have data

2013-10-09 Thread Gino Pirelli 
probally solution is connecting to the slots of the QgsRasterLayer class

  public slots:
void showStatusMessage( const QString & theMessage );

/** \brief Propagate progress updates from GDAL up to the parent app */
void updateProgress( int, int );

/** \brief receive progress signal from provider */
void onProgress( int, double, QString );

ciao Luigi Pirelli (luigi.pire...@faunalia.it)
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [Qgis-developer] Reliable way to determine a groupIndex in the legend tree

2013-10-09 Thread Andreas Neumann 
Hi Luigi,

Do you have a written proposal already? Is it available on Google Docs
so we can comment/append?

Personally, I am also interested in enhancing group options in the layer
tree. Groups are not just stupid containers with only toggling of
visibility. They could also have a group opacity and blending and group
metadata. Later on we may also have clipping and masking on groups. So
we really should have an extensible mechanism for groups as new group
features appear.

Yes, if we have a written proposal we will have more organization that
may help to fund this effort.

Andreas

Am 09.10.2013 18:19, schrieb Gino Pirelli:
> we did a little face to face meeting in Brighton with Martin D. and Marco
> H... QgsLegend classes need a refactory due to structural limitation.
> 
> Marco H. did an estimation of the refactoring, but probably, as I can see
> there are more that need this refactoring, could be interesting collect
> features that this re-write should involve
> and obviously funding :)
> 
> ciao Luigi Pirelli (luigi.pire...@faunalia.it)
> 
> 
> 
> On 9 October 2013 09:35, Denis Rouzaud  wrote:
> 
>> I heard some discussions... from Matthias, Martin and Nathan.
>>
>> but there are probably others!
>>
>>
>> On 10/09/2013 09:30 AM, Andreas Neumann wrote:
>>
>>> Am 09.10.2013 06:16, schrieb Bernhard Ströbl:
>>>
 Hi

 Am 09.10.2013 08:12, schrieb Denis Rouzaud:

> Hi,
>
> I am not sure we should make any publicity to the legend API which is
> very bad.
>
> I hope these methods will deprecated soon ;)
>
 So is there anyone working on fixing these legend API problems?
>>>
>>> We should probably come up with a Wiki page where we could collect ideas
>>> what we would expect from a proper legend API.
>>>
>>> Andreas
>>> __**_
>>> Qgis-developer mailing list
>>> Qgis-developer@lists.osgeo.org
>>> http://lists.osgeo.org/**mailman/listinfo/qgis-**developer
>>>
>>
>> __**_
>> Qgis-developer mailing list
>> Qgis-developer@lists.osgeo.org
>> http://lists.osgeo.org/**mailman/listinfo/qgis-**developer
>>
> 

___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [Qgis-developer] Reliable way to determine a groupIndex in the legend tree

2013-10-09 Thread Nyall Dawson 
>
> Personally, I am also interested in enhancing group options in the layer
> tree. Groups are not just stupid containers with only toggling of
> visibility. They could also have a group opacity and blending and group
> metadata. Later on we may also have clipping and masking on groups. So
> we really should have an extensible mechanism for groups as new group
> features appear.
>

One thing I'd love to see is the ability to make the layers in a group
"mutually exclusive". (ie, imagine radio buttons instead of checkboxes for
controlling the visibility of layers within that group). I find I'm often
going through a process of switching one layer on, then switching a bunch
of other layers off for comparison purposes. A mutually exclusive layer
group which automatically switched the other child layers off when I select
a layer would make this process much nicer!

Nyall
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer