Re: [Qgis-user] How to combine multiple rasters into a tilesarray?

[Nyall Dawson]

On Fri, 22 Jun 2018 at 07:15, Joep Orbons  wrote:
>
> Hi Nicolas,
>
> Good idea, the virtual layer is a new option to me. I tried it on a multitude 
> of files and it produced a strange pallette image, mostly black. I'll try it 
> with some less files first and see what that brings, in small steps.
> But any good suggestion is welcome.

You probably had the "Layer stack" option checked (renamed in 3.2 to
"Place each input file into a separate band"). Try again with this
option unchecked.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] [QGIS-Developer] Gentle reminder: please test 3.2 RC and update Visual Changelog with your work items

[Nyall Dawson]

On Fri, 15 Jun 2018 at 19:41, Richard Duivenvoorde  wrote:
>
> Hi Devs,
>
> With QGIS 3.2 coming near...
>
> Please devs: add items too our visual changelog,
> so if downloads start we will have something to show (off):

Thanks for the reminder!

I've just trawled through all the commits since 3.0 was released and
added entries for all missing features. Some of these are very empty
though - just a heading.  I also may have missed sponsors if they
weren't indicated in the commit message - so please double check those
for any sponsored work which wasn't already added to the changelog.

So now it's over to some other magic elf to flesh out these entries
with nicer text and lovely screenshots. ;)



Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] GDBTable Files

[Nyall Dawson]

On Mon, 18 Jun 2018 at 11:47, Mike Flannigan  wrote:

> I followed your instructions and it worked perfectly.
> Thanks so much.  Very helpful.
>

* snip *
> So I am in a good place now.  I plan to change the CRS to NAD27,
> but I think I am good to go.  I have a lot of work ahead of me.

> I really appreciate your help.

No problem -- as a follow up, I'd suggest you wrap this process up in
a "Processing model" if you're going to be doing it more than once.

Your model could consist of

1. A Reproject algorithm to transform the layer to a the desired CRS (e.g. 4326)
2. The "Extract Nodes" algorithm to obtain a point for each node
3. The "Export/Add geometry columns" algorithm to add the fields with
x/y values for each node, and using the output from this algorithm as
the model output

Then you could just run the model in a single click to process the
files as desired!

Cheers,
Nyall



> Thank you Nyall.
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] GDBTable Files

[Nyall Dawson]

On Mon, 18 Jun 2018 at 07:50, Mike Flannigan  wrote:
>
>
> I understand GDBTable Files are part of a geodatabase.
> I am please to see that my QGIS 2.14.3 opens these
> just fine.  That was a pleasant surprise.

Hi Mike!

First off, I'd strongly suggest upgrading to 2.18 (or 3.0, if you like
bleeding edge!). You'll get lots of bug fixes and improvement
behaviour over the older, unsupported 2.14 release. If you're stuck on
2.14 for some reason, you definitely should be running the final patch
release for that version, which is 2.14.22.

> I have BLM land data and Federal land data for the USA
> in the GDB format.  I would like to obtain the lat/long
> point data for all the polygons.  If anybody knows a
> straight forward way to do that, please tell me.

Do you mean for each node in the polygons, or just for the centroid?

The easiest way is to use the built-in Processing toolbox.

If you're after a lat/long for the centroid:
1. Run the "Polygon centroids" algorithm to obtain a point for each centroid
2. Run the "Export/Add geometry columns" algorithm to add a field for
centroid x/y values
3. Export to XLS/CSV/etc

If you're after a lat/long for each node:
1. Run the "Extract Nodes" algorithm to obtain a point for each node
2. Run the "Export/Add geometry columns" algorithm to add a field for
centroid x/y values
3. Export to XLS/CSV/etc

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] "Pick color" freeze win10?

[Nyall Dawson]

On Tue, 12 Jun 2018 at 23:37, Elstermann, Mike
 wrote:
>
> QGIS 2.14 / 2.18(.20) under Windows 10:
> Windows freezes when the "Pick color" function is used. Only way out 
>  and "Abort". Ideas?

I see the same on Windows 7, but only with 2.18.

I've (fortunately) never experienced this with 3.0.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS3 Zonal Stat in raster Analysis fail count

[Nyall Dawson]

On Wed, 6 Jun 2018 at 02:25, Mensio Federico  wrote:
>
> But the raster and vector have the same CRS.
> The raster is 20X20m, the grid 100mX100m and the two files have the same 
> origin.

Please check with the next round of nightly releases - I've just
pushed many fixes and optimisations for zonal stats to the 3.2 builds.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Creating boundary map

[Nyall Dawson]

On 7 June 2018 at 16:37, PARESH CHAUKHANDE  wrote:
> Hello,
>  I want to create a boundary layer on a map in QGIS 3.0.3 with Mac OS.

Fantastic! You've chosen the right software to do this in.

Nyall

(If you've got a specific question, better ask it so we know how to help!)
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS3 Zonal Stat in raster Analysis fail count

[Nyall Dawson]

2018-05-31 22:34 GMT+10:00 Mensio Federico :
> Hi.
> I used zonal statistics for raster analysis both in Qgis 2.18 and Qgis 3.0.
> In the first it's work perfectly with the count of cell is correct, in the
> second is totally uncorrect.
> Some solution?

Can you share sample data and the parameters you use?

Nyall

>
> --
> Federico MENSIO
> IPLA spa
> Corso Casale, 476 - 10132 Torino
> Mobile +39.340.06.01.363
> TEL +39.011.432.0443 FAX +39.011.89.89.333
> men...@ipla.org - www.ipla.org
>
> -
> Questa e-mail è personale e i suoi contenuti contengono informazioni
> riservate ai destinatari. Pertanto la lettura, copia od altro uso da parte
> di altri o qualsiasi altra azione derivante dalla conoscenza di queste
> informazioni sono rigorosamente vietate e perseguibili ai sensi del Dlgs
> 196/03.
> Qualora abbiate ricevuto questo documento per errore siete cortesemente
> pregati di darne immediata comunicazione al mittente ai riferimenti indicati
> e provvedere immediatamente alla sua distruzione.
>
> Salvaguardiamo l'ambiente: se non ti è necessario, non stampare questa mail
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] How do I pass parameters to grass7 algorithms with Processing.run?

[Nyall Dawson]

On 14 May 2018 at 19:30, Stephen  wrote:
>
> Automatically setting output paths to temporary files would be a
> sensible convenience, I think.
>
> In any event, I used processing.tools.system.getTempFilename() to set a
> filename and that gets me past the 'Incorrect parameter value for
> output' error.
>
> However, I am still getting "Unable to execute algorithm".
>
> In AlgorithmExecutor, line 55:
> results, ok = alg.run(parameters, context, feedback)
>
> returns ({}, False)
>
> parameters contains:
>
> {'GRASS_REGION_CELLSIZE_PARAMETER': 0.008333, 'use': 1,
> 'GRASS_REGION_PARAMETER':
> '33.91415594546105,41.91384412832699,-4.679722671256611,5.467319665842772',
> 'input': ,
> 'GRASS_MIN_AREA_PARAMETER': 0.0001, 'GRASS_SNAP_TOLERANCE_PARAMETER':
> -1, 'type': 0, 'output':
> '/tmp/processing_da49d2e8c1bd4e249d1a5c6d42680149/5af9540a0b2212'}
>
> context contains:
>
> 
>
> and feedback contains:
>
> 
>
> Without a useful error message, it's very difficult to troubleshoot this
> on my own.
>
> Do you have any idea what might be wrong here?

I had another look at this. I suspect what might be happening is that
'tmp/processing_da49d2e8c1bd4e249d1a5c6d42680149/5af9540a0b2212' does
not exist, which makes GRASS bomb out with a 'No such file or
directory' error.

Are you running this script from within QGIS itself? You should see
the detailed log, including all the output from grass, by opening the
log window and switching to the "Processing" tab (alternatively
running from a terminal window will also show this output on the
command line).

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Contours generation problem.

[Nyall Dawson]

On 28 May 2018 at 06:32, Xero  wrote:
> Hello I have a problem when trying to generate a contour on a raster. Every
> time I start the processus I've got this error message :
> /Traitement de l'algorithme ...
> AlgorithmDialog
> Input parameters:
> { 'BAND' : 1, 'CREATE_3D' : False, 'FIELD_NAME' : 'ELEV', 'IGNORE_NODATA' :
> False, 'INPUT' : '/Users/arnomak/Desktop/RDB DEM/Agisoft/DEM.tif',
> 'INTERVAL' : 10, 'NODATA' : 0, 'OFFSET' : 0, 'OUTPUT' :
> '/var/folders/22/h094_46j6fs1y_hp6gf81r80gn/T/processing_5ac940d3667d4295b56d8931c0ec3a61/813de11bbc234ad792da768934852d05/OUTPUT.shp'
> }
>
> GDAL command:
> gdal_contour -b 1 -a ELEV -i 10.0 -snodata 0.0 -f "ESRI Shapefile"
> "/Users/arnomak/Desktop/RDB DEM/Agisoft/DEM.tif"
> "/var/folders/22/h094_46j6fs1y_hp6gf81r80gn/T/processing_5ac940d3667d4295b56d8931c0ec3a61/813de11bbc234ad792da768934852d05/OUTPUT.shp"
> GDAL command output:
> /bin/sh: gdal_contour: command not found

Here's the error ^^. QGIS can't run gdal_contour, so it's likely not
available on your workstation. Try running gdal_contour from a
terminal window and see if it works there. If not, I suspect you'r
missing a required gdal package.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Programmatically export layout with georeferenced file

[Nyall Dawson]

On 25 May 2018 at 03:02, DelazJ  wrote:
> Hi all,
>
> I have a layout on which I checked the "save world file" option in Layout
> --> Exports settings
> When I export this layout to image, I'd expect it to generate a world file
> next to it considering that the generateWorldFile is an attribute of the
> ImageExportSettings class [0]. But no world file.

The GUI setting isn't stored in the layout*  - you need to manually
set generateWorldFile to true
in your ImageExportSettings if you want the world file created.


> mylayout=QgsProject.instance().layoutManager().layoutByName('Composeur 1')
> exporter = QgsLayoutExporter(mylayout)
> exporter.exportToImage(filePath, QgsLayoutExporter.ImageExportSettings)

If you want the world file, this should be

exporter = QgsLayoutExporter(mylayout)
settings = QgsLayoutExporter.ImageExportSettings()
settings.generateWorldFile = True
exporter.exportToImage(filePath, settings)


> ls=QgsProject.instance().layoutManager().layouts()
> for elt in ls:
>   exporter = QgsLayoutExporter(elt)
>   imgSettings = exporter.ImageExportSettings()
>   print(elt.name(), imgSettings.generateWorldFile)
>
> which returns False for any layout, including the one I checked the option
> for. Btw, dpi returns -1.0, pages return an empty list... I guess I might
> have missed something but...No idea.

That's because imgSettings here is a default constructed
ImageExportSettings - which defaults to no world file, and dpi of -1
(as noted in the docs, the dpi in ImageExportSettings is an override
... if it's -1, then the layout's DPI will be used). An empty pages
list means export all pages. You need to manually populate this
ImageExportSettings as required for your export.

Nyall


* actually it is stored in a customProperty for the layout, but that's
not part of stable API

layout.customProperty('exportWorldFile')


> I use this kind of code
>
> mylayout=QgsProject.instance().layoutManager().layoutByName('Composeur 1')
> exporter = QgsLayoutExporter(mylayout)
> exporter.exportToImage(filePath, QgsLayoutExporter.ImageExportSettings)
>
> and then I check the image export settings with the next code
>
> ls=QgsProject.instance().layoutManager().layouts()
> for elt in ls:
>   exporter = QgsLayoutExporter(elt)
>   imgSettings = exporter.ImageExportSettings()
>   print(elt.name(), imgSettings.generateWorldFile)
>
> which returns False for any layout, including the one I checked the option
> for. Btw, dpi returns -1.0, pages return an empty list... I guess I might
> have missed something but...No idea.
> When is the generateWorldFile set to True (if ever it's what will trigger my
> world file export)?
>
> [0]
> https://qgis.org/pyqgis/master/core/Layout/QgsLayoutExporter.html#qgis.core.QgsLayoutExporter.ImageExportSettings
>
> Thanks for your help,
> Harrissou
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS 3 clipping raster with a polygon - problem

[Nyall Dawson]

On 24 May 2018 at 01:12, David Addy  wrote:
> I have had another go with another dataset.
> I have tried both these algorithms in QGIS 3.0.2:
> 1.  Gdal ->" Raster extraction" -> "Clip raster by mask layer"
> 2.  Saga -> "Vector <-> raster" -> "Clip raster with polygon"
> Both give an error message.
> In the case of GDAL it is :"GDAL command output:
> ERROR 6: SetColorTable() only supported for Byte or UInt16 bands in TIFF
> format."
> I must be completely misunderstanding both these processes.
> By reverting to QGIS 2.18.16 the GDAL algorithm works quite happily on the
> same geotiff using the same shp file polygon to clip it..
> In Q3 I do not understand why the output file entry box is labelled
> Clipped(mask) and the box contains (save to temporary file)
> Are there some other entries I need to make it work?

Hi David,

In  both 2.18 and 3.0 you can see the GDAL command which is run by the
algorithm. Can you see if there's any difference between the two? Can
you please share them?

Cheers,
Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Programmatically close a layout dialog

[Nyall Dawson]

On 22 May 2018 at 23:54, DelazJ  wrote:
> Hi,
>
> From the Python Console, how can I close a print layout dialog, knowing its
> name?

In 3.0 something like this will work:

designers = [d for d in iface.openLayoutDesigners() if
d.masterLayout().name() == 'Layout 1']
if designers:
designers[0].close()

Nyall



>
> Thanks for any hint,
> Harrissou
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] How do I pass parameters to grass7 algorithms with Processing.run?

[Nyall Dawson]

On 14 May 2018 at 19:30, Stephen  wrote:

> Without a useful error message, it's very difficult to troubleshoot this
> on my own.

Is this on master? The error messages have been greatly improved for
3.2, so I'd be interested to see what results you get there.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] Processing 3.0 - porting guide

[Nyall Dawson]

Hi all,

Just a quick heads up - I've finally had the chance to finish off the
guide to porting Processing 2.x algorithms to 3.x API and merged it
into the API docs.

You can find it here:

https://raw.githubusercontent.com/qgis/QGIS/master/doc/porting_processing.dox


Hope it helps -- of course, more contributions and greater detail
would be very much appreciated. So if in your algorithm porting you
come across some traps/tricks you think should be added, please open a
pull request adding these!

Cheers,
Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] Drill down (cascading) forms in QGIS crowdfund – final stretch!

[Nyall Dawson]

Hi all,

If you're not aware, North Road have been running a crowd funding
campaign to implement drill-down (cascading) field support within QGIS
forms.

We’re now nearing the final hours of this campaign, and thanks to
numerous generous backers we’re very close to hitting the funding
goal! This is a really exciting new feature which would help add
greater flexibility and power to QGIS feature forms, but in order to
implement it for QGIS 3.2 we need to hit the funding target by 11 May
2018.

As a result, we’re dropping the minimum contribution amount and
throwing open the campaign for payments of any amount. You can read
more about this (and the campaign itself) here:

https://north-road.com/2018/05/08/drill-down-cascading-forms-in-qgis-crowdfund-final-stretch/

Spread the word if you'd like to see this feature land in QGIS 3.2!

Cheers,
Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Clip Raster by Mask Layer algorithm not working when part of processing model

[Nyall Dawson]

On 18 April 2018 at 06:03, Phil Ribbens  wrote:
> Hi,
>
> I'm trying to include the “Clip raster by mask layer” algorithm in a
> processing model, but I can’t get it to work. What’s weird is that when I
> run the algorithm by itself (i.e. not as part of a model), it works fine.
>
> Below are the logs of each (when I run it by itself and when I run it as
> part of a model). From studying these it seems like the only difference (and
> what seems to be causing the problem) is that when the algorithm is run from
> within a model the path to the MASK layer isn’t being included in the GDAL
> command. Anybody have any idea why? Am I doing something wrong? Is it a bug?

This is a known issue in 3.0.1 - it will be fixed in 3.0.2

Nyall

>
> Thanks,
> Phil Ribbens
>
>
> Running Clip raster by mask layer by itself (works fine)
> Processing algorithm…
> AlgorithmDialog
> Input parameters:
> { 'INPUT' :
> '/var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_c893e43e9931410397ab42a423fd9d61/e27038e2ea2f4b42842aaacf3c6bf15c/OUTPUT.tif',
> 'MASK' : '/Users/philipribbens/Dropbox/work/QGIS plugin for
> Key-Log/lower_roanoke_subbasin.shp', 'NODATA' : 255, 'ALPHA_BAND' : False,
> 'CROP_TO_CUTLINE' : True, 'KEEP_RESOLUTION' : False, 'OPTIONS' : '',
> 'DATA_TYPE' : 0, 'OUTPUT' :
> '/var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_6c6a3f6013994c2eafbaf28a231297a3/44f299a602e34ef8a666e6848ff95658/OUTPUT.tif'
> }
>
> GDAL command:
> gdalwarp -ot Byte -of GTiff -cutline "/Users/philipribbens/Dropbox/work/QGIS
> plugin for Key-Log/lower_roanoke_subbasin.shp" -crop_to_cutline -dstnodata
> 255.0
> /var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_c893e43e9931410397ab42a423fd9d61/e27038e2ea2f4b42842aaacf3c6bf15c/OUTPUT.tif
> /var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_6c6a3f6013994c2eafbaf28a231297a3/44f299a602e34ef8a666e6848ff95658/OUTPUT.tif
> GDAL command output:
> Copying color table from
> /var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_c893e43e9931410397ab42a423fd9d61/e27038e2ea2f4b42842aaacf3c6bf15c/OUTPUT.tif
> to new file.
> Creating output file that is 3642P x 2860L.
> Processing input file
> /var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_c893e43e9931410397ab42a423fd9d61/e27038e2ea2f4b42842aaacf3c6bf15c/OUTPUT.tif.
> Using internal nodata values (e.g. 255) for image
> /var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_c893e43e9931410397ab42a423fd9d61/e27038e2ea2f4b42842aaacf3c6bf15c/OUTPUT.tif.
> 0...10...20...30...40...50...60...70...80...90...100 - done.
> Execution completed in 0.65 seconds
> Results:
> {'OUTPUT':  {'sink':/var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_6c6a3f6013994c2eafbaf28a231297a3/44f299a602e34ef8a666e6848ff95658/OUTPUT.tif,
> 'createOptions': {'fileEncoding': 'System'}}>}
>
> Loading resulting layers
> Algorithm 'Clip raster by mask layer' finished
>
>
>
> Running Clip raster by mask layer as part of a model (doesn’t work)
> Processing algorithm…
> AlgorithmDialog
> Input parameters:
> { 'inputrasterlayer' :
> '/var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_c893e43e9931410397ab42a423fd9d61/e27038e2ea2f4b42842aaacf3c6bf15c/OUTPUT.tif',
> 'maskvectorlayer' : '/Users/philipribbens/Dropbox/work/QGIS plugin for
> Key-Log/lower_roanoke_subbasin.shp', 'gdal:cliprasterbymasklayer_1:Final
> clipped raster' :
> '/var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_6c6a3f6013994c2eafbaf28a231297a3/3b9e2df02eca4aa6b8f6476df78cdf9f/gdal:cliprasterbymasklayer_1:Final
> clipped raster.tif' }
>
> Prepare algorithm: gdal:cliprasterbymasklayer_1
> Running Clip raster by mask layer [1/1]
> Input Parameters:
> { ALPHA_BAND: False, CROP_TO_CUTLINE: True, DATA_TYPE: 0, INPUT:
> '/var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_c893e43e9931410397ab42a423fd9d61/e27038e2ea2f4b42842aaacf3c6bf15c/OUTPUT.tif',
> KEEP_RESOLUTION: False, MASK: '/Users/philipribbens/Dropbox/work/QGIS plugin
> for Key-Log/lower_roanoke_subbasin.shp', NODATA: 255, OPTIONS: '', OUTPUT:
> '/var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_6c6a3f6013994c2eafbaf28a231297a3/3b9e2df02eca4aa6b8f6476df78cdf9f/gdal:cliprasterbymasklayer_1:Final
> clipped raster.tif' }
> GDAL command:
> gdalwarp -ot Byte -of GTiff -cutline -crop_to_cutline -dstnodata 255.0
> /var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_c893e43e9931410397ab42a423fd9d61/e27038e2ea2f4b42842aaacf3c6bf15c/OUTPUT.tif
> "/var/folders/10/qfg4mkl127gcxgmpr82tl1xcgq/T/processing_6c6a3f6013994c2eafbaf28a231297a3/3b9e2df02eca4aa6b8f6476df78cdf9f/gdal:cliprasterbymasklayer_1:Final
> clipped raster.tif"
> GDAL command output:
> ERROR 1: Cannot open -crop_to_cutline.
> OK. Execution took 0.196 s (1 outputs).
> Model processed OK. Executed 1 algorithms total in 0.372 s.
> Execution completed in 0.41 seconds
> Results:
> {'gdal:cliprasterbymasklayer_1:Final clipped raster':
>  

Re: [Qgis-user] QGIS3 importing from ARCGIS with custom coordinate in arc.

[Nyall Dawson]

On 17 April 2018 at 03:38, Basques, Bob (CI-StPaul)
 wrote:
> All,
>
> I have to connect QGIS3 to a ARCGIS server.
>
> Using the ArcGisFeatureServer connector:
>
>  I can see the catalog just fine, but when I try to add a layer I get the
> “AFSProvidor Could Not Parse Spatial Reference” error.
>
> I successfully added the Custom layer on the Arc Server to QGIS with a
> custom EPSG, but how to I get QGIS and Arcserver to agree on a title for the
> projection?  The “Spatial Refence” on the arc server is the actual PROJCS
> line.  Do I need to use this as the name in QGIS?

Got a link to the server we can look at?

Nyall

>
> Thanks
>
> bobb
>
>
>
>
> We’ve heard that a million monkeys at a million keyboards could produce the
> complete works of Shakespeare; now, thanks to the Internet, we know that is
> not true.
> —Robert Wilensky
>
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] Drill-down (cascading) forms in QGIS Value Relation Widgets crowdfunding

[Nyall Dawson]

Hi all,

North Road are proud to launch our latest QGIS crowd-funding campaign:
Drill-down (cascading) forms! This campaign covers extending the
functionality present in QGIS “Value Relation Widgets”, allowing the
implementation of complex, dynamic filters within QGIS attribute
forms. The functionality can be used to implement “drill-down” forms
within QGIS, where the values available in one field depend on the
values of other fields.

Full details are available at the campaign page:
https://north-road.com/drill-down-cascading-forms/

We need your help to make this feature a reality! We need backers, but
just as importantly, we need people to spread the word about this
campaign so that all interested users and organisations hear about it.
Here are some ideas:

- spread the campaign page via your social media accounts: Twitter,
Linked in, etc
- forward a link to the campaign to any relevant professional networks
you're part of
- publicise the campaign via your local QGIS user group
- have a chat to your manager... if this feature will make your work
easier, and you've saved money by implementing QGIS within your
organisation, consider contributing and let us make your job easier!

Stay tuned for more updates and posts exploring this feature as the
campaign progresses...

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Principal Components Analysis Problem

[Nyall Dawson]

On 11 April 2018 at 16:57, Grant Boxer  wrote:
> I am trying to do a PCA on ASTER satellite data using bands 4 through 9 (all
> 30 m pixel size). I have tried using the Processing Toolbox > Grass >
> Imagery>i.pca but it continually fails.
>
>
>
> I get this at the end of processing.
>
> _
>
> Traceback (most recent call last):
> File
> "C:/PROGRA~1/QGIS3~1.0/apps/qgis/./python/plugins\processing\algs\grass7\Grass7Algorithm.py",
> line 395, in processAlgorithm
> if isinstance(out, QgsProcessingOutputHtml):
> NameError: name 'QgsProcessingOutputHtml' is not defined

Unfortunately this is a known bug in 3.0.1, fixed in the upcoming
3.0.2 version (along with a TON of other Processing GRASS plugin
fixes).

As a short-term fix you could try downloading the qgis-rel-dev package
via osgeo4w64 to get access to these fixes before the official release
of 3.0.2.

Nyall


>
> Execution failed after 16.32 seconds
>
>
>
> Loading resulting layers
>
> Algorithm 'i.pca - Principal components analysis (PCA) for image
> processing.' finished
>
> _
>
> When using the SemiAutomatic Classification PCA calculator I get the
> following error message.
>
>
>
> 2018-04-11T14:51:00 WARNING Traceback (most recent call last):
>
> File
> "C:/Users/Grant/AppData/Roaming/QGIS/QGIS3\profiles\default/python/plugins\SemiAutomaticClassificationPlugin\maininterface\pcaTab.py",
> line 46, in calculatePCAAction
>
> self.calculatePCA()
>
> File
> "C:/Users/Grant/AppData/Roaming/QGIS/QGIS3\profiles\default/python/plugins\SemiAutomaticClassificationPlugin\maininterface\pcaTab.py",
> line 97, in calculatePCA
>
> self.PCACalculation(rD, bL, outF, numbComp, None, batch)
>
> File
> "C:/Users/Grant/AppData/Roaming/QGIS/QGIS3\profiles\default/python/plugins\SemiAutomaticClassificationPlugin\maininterface\pcaTab.py",
> line 140, in PCACalculation
>
> comp, totalVariance, totalVarianceCumulative, eigenValues =
> self.calculateEigenVectors(covM)
>
> File
> "C:/Users/Grant/AppData/Roaming/QGIS/QGIS3\profiles\default/python/plugins\SemiAutomaticClassificationPlugin\maininterface\pcaTab.py",
> line 264, in calculateEigenVectors
>
> val, vect = cfg.np.linalg.eigh(matrix)
>
> File
> "C:\PROGRA~1\QGIS3~1.0\apps\Python36\lib\site-packages\numpy\linalg\linalg.py",
> line 1291, in eigh
>
> w, vt = gufunc(a, signature=signature, extobj=extobj)
>
> File
> "C:\PROGRA~1\QGIS3~1.0\apps\Python36\lib\site-packages\numpy\linalg\linalg.py",
> line 96, in _raise_linalgerror_eigenvalues_nonconvergence
>
> raise LinAlgError("Eigenvalues did not converge")
>
> numpy.linalg.linalg.LinAlgError: Eigenvalues did not converge
>
> _
>
> Can anyone shed any light on what is happening here?
>
>
>
> Cheers Grant
>
>
>
> Grant Boxer (FAIG RPGeo)
>
> Maylands, Western Australia, 6931
>
>
>
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] How do I pass parameters to grass7 algorithms with Processing.run?

[Nyall Dawson]

On 10 April 2018 at 20:25, Stephen Bosch <post...@vodacomm.ca> wrote:
> Hi Nyall:
>
> On Tue, Apr 10, 2018 at 2:20 AM, Nyall Dawson <nyall.daw...@gmail.com>
> wrote:
>>
>> On 10 April 2018 at 01:03, Stephen Bosch <post...@vodacomm.ca> wrote:
>>
>> > I've tried calling the algorithm this way:
>> >
>> >> file_grid_r = processing.run("grass7:v.to.rast", {'INPUT': grid_v,
>> >> 'type':
>> >> 0, 'use': 1, 'GRASS_REGION_PARAMETER': extent}, feedback=None)
>> >
>>
>> Looks correct
>>
>> > But it fails. On the console, I see
>> >
>> >> sys:1: ResourceWarning: unclosed > >> family=AddressFamily.AF_INET, type=SocketKind.SOCK_STREAM, proto=0,
>> >> laddr=('127.0.0.1', 41358), raddr=('127.0.0.1', 37307)>
>>
>> That can be ignored - it's unrelated.
>
>
> (Messages like this occur only when the process exits with a code other than
> 0.)
>
> I should add that, without using the debugger, I wouldn't be able to see the
> error at all. Is that intended behaviour? Is there any way to turn it on?
> The absence of exceptions and tracebacks make things fail silently, which
> can have nasty consequences.
>
>>
>> >
>> > If I step through the code in PyCharm and trace the exception, I see an
>> > AttributeError: 'NoneType' object has no attribute 'messageBar'.
>>
>> Where is this thrown? I suspect the problem is because you're passing
>> "feedback=None" and somewhere there's code which is expecting a valid
>> feedback object. Can you let me know which line the exception is
>> thrown on and I'll fix?
>
>
> It is thrown in line 42 of MessageBarProgress.py:
> __init__ [MessageBarProgress.py:42]  id:94435226437000

Thanks, fixed in
https://github.com/qgis/QGIS/commit/d5573a5a7bf5a32490b50f528da10c6f58871eab

>
> After correcting the statement to the following:
>
>> file_grid_r = QgsRasterLayer()
>> processing.run("grass7:v.to.rast", {'input': grid_v, 'type': 0, 'use': 1,
>> 'GRASS_REGION_PARAMETER': extent,
>> 'output': file_grid_r,
>> 'GRASS_REGION_CELLSIZE_PARAMETER':
>> 0.008333,
>> 'GRASS_SNAP_TOLERANCE_PARAMETER': -1,
>> 'GRASS_MIN_AREA_PARAMETER': 0.000100},
>>  feedback=QgsProcessingFeedback())
>
>
>  I get the message
>
>> Unable to execute algorithm
>> Incorrect parameter value for output

The OUTPUT parameter should be set to the path to save the result,
e.g. 'd:\blahblah.tif":

 processing.run("grass7:v.to.rast", {'input': grid_v, 'type': 0, 'use': 1,
 'GRASS_REGION_PARAMETER': extent,
 'output': 'd:\blah.tif',
 'GRASS_REGION_CELLSIZE_PARAMETER':
 0.008333,
 'GRASS_SNAP_TOLERANCE_PARAMETER': -1,
 'GRASS_MIN_AREA_PARAMETER': 0.000100},
  feedback=QgsProcessingFeedback())

processing.run will return a dict a results, and inside that you'll
automatically have a QgsRasterLayer loaded corresponding to this
output.

E.g.

params = { ., 'OUTPUT': 'd:\blah.tif' }
results = processing.run('grass7:v.to.rast', params, feedback )
my_raster_layer = results['OUTPUT']
assert my_raster_layer.isValid()

Here my_raster_layer will be a QgsRasterLayer, ready for further
processing or adding to the project for visualisation.

I'm pondering whether we should automatically set output locations to
temporary files if they aren't specified. E.g. if you don't set OUTPUT
to 'd:\blah.tif', it'll fill it in automatically with a .tif file in
your temp folder...

Nyall

> I've tried a couple of variations (specifying 'output' as an index and
> assigning it to a variable while setting 'output': None in the dictionary,
> for example) and get the same error.
>
> In my 2.x code, I just used the index and assigned the return to a variable.
> The implementation of v.to.rast for QGIS 3 Processing seems to require a
> value for output, and simply putting an variable doesn't work either (it
> complains that the variable is undefined).
>
> With a few more hints, I'm sure I'll get the hang of this ;-)
>
> Thanks
>
> Stephen
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] How to Prevent GeoPackage file from updating modified date when viewed in a Qgis project when it is not being edited

[Nyall Dawson]

On 10 April 2018 at 16:47, Alessandro Pasotti  wrote:
> On Tue, Apr 10, 2018 at 8:33 AM, Borys Jurgiel 
> wrote:
>>
>> Dnia wtorek, 10 kwietnia 2018 08:11:48 CEST Alessandro Pasotti pisze:
>>
>>
>>
>> > Hi, I also faced this problem when Implementing gpkg support in the QGIS
>>
>> > browser panel: when the file browser scanned the folder, it opened the
>>
>> > gpkg, which in turn changed its file stats and made the directory
>> > watcher
>>
>> > think that something has changed in the directory, which in turns
>> > started
>>
>> > scanning again in a delicious endless loop :)
>>
>> >
>>
>> > I think I solved this with a kind of locker for the directory scan
>> > thread,
>>
>> > but it was't easy.
>>
>>
>>
>> Ah so this it the loop in 2.18... Could you recall that commit that fixed
>> it in 3.0? I understand it's not trivial to backport it to 2.18?
>
>
> Hi Borys,
>
> Are you sure that this is also a problem with 2.x?
>
> If I remember correctly 2.x browser panel does not open gpkg for scanning
> contents, this feature was added for 3.x.

I've seen this with recent versions of 2.18

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] How do I pass parameters to grass7 algorithms with Processing.run?

[Nyall Dawson]

On 10 April 2018 at 01:03, Stephen Bosch  wrote:

> The input types are described with terms like
> ""; what does this mean exactly? Do I
> need to pass class instances?

Nope, strings/ints/layer refs/etc are fine. The class is shown in the
help to indicate what values are acceptable (e.g. you can't use a
layer reference for a point parameter!)

> I've tried calling the algorithm this way:
>
>> file_grid_r = processing.run("grass7:v.to.rast", {'INPUT': grid_v, 'type':
>> 0, 'use': 1, 'GRASS_REGION_PARAMETER': extent}, feedback=None)
>

Looks correct

> But it fails. On the console, I see
>
>> sys:1: ResourceWarning: unclosed > family=AddressFamily.AF_INET, type=SocketKind.SOCK_STREAM, proto=0,
>> laddr=('127.0.0.1', 41358), raddr=('127.0.0.1', 37307)>

That can be ignored - it's unrelated.
>
> If I step through the code in PyCharm and trace the exception, I see an
> AttributeError: 'NoneType' object has no attribute 'messageBar'.

Where is this thrown? I suspect the problem is because you're passing
"feedback=None" and somewhere there's code which is expecting a valid
feedback object. Can you let me know which line the exception is
thrown on and I'll fix?

In the meantime try passing a feedback object:

   feedback = QgsProcessingFeedback()


Nyall

> I'm not sure if this is a bug in QGIS or whether there is something wrong
> with my parameters.
>
> Do you have any suggestions or ideas?
>
> Thanks
>
> Stephen
>
>
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS 3 : Run alg in an algorithm

[Nyall Dawson]

On 9 April 2018 at 06:23, Jean-Baptiste Desbas  wrote:
> Hello,
>
> I wan't to run a native algorithm (execute SQL) in an homemade algorithm :
> I try
> processing.run('qgis:executesql',{'INPUT_DATASOURCES':[sink],'INPUT_QUERY':'SELECT
> * FROM input1'},feedback=None)
>
> but i've got an error :
>
> line 140, in processAlgorithm
> a=processing.run('qgis:executesql',{'INPUT_DATASOURCES':[sink],'INPUT_QUERY':'SELECT
> * FROM input1'},feedback=None)
> File "/usr/share/qgis/python/plugins/processing/tools/general.py", line 84,
> in run
> return Processing.runAlgorithm(algOrName, parameters, onFinish, feedback,
> context)
> File "/usr/share/qgis/python/plugins/processing/core/Processing.py", line
> 135, in runAlgorithm
> feedback.reportError(msg)
> TypeError: reportError() missing 1 required positional argument:
> 'fatal_error'
>

Thanks, fixed in https://github.com/qgis/QGIS/commit/ad54c0

Note that you should be passing on the feedback object from
processAlgorithm to this sub-algorithm, or you'll lose progress
reports, cancelation support and feedback messages sent from the sub
algorithm.

E.g.

processing.run('qgis:executesql',{'INPUT_DATASOURCES':[sink],'INPUT_QUERY':'SELECT
* FROM input1'},feedback=feedback)

That will also avoid the error on QGIS < 3.0.2

Nyall


>
> how can I do that ?
>
> Thanks,
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] [QGIS-Developer] QGIS 3 Print Composer - Page Properties

[Nyall Dawson]

On 3 April 2018 at 07:06, Carlo A. Bertelli (Charta s.r.l.)
 wrote:
> I suggest to provide the same interface as version 2 calling this "default
> page size" and explaining that every page could get a different size,
> right-clicking on the page. Most users make one-page maps and they would
> never need this otherwise vital feature.

I've thought long on this. I'm reasonably convinced it's a short-term
issue as people adjust there 2.x workflows to 3.0.

There's two approaches I can see working to mitigate it:

1. Add a "pages panel", which allows you to pick from existing pages
(e.g. via a drop-down combo box) and then set the properties for that
page. It could also have the actions for adding new pages and deleting
existing pages there too.

and/or

2. Add a "new layout wizard", which shows when you create a new print
layout. This would prompt users to set the page size before the layout
is actually created. It could potentially replace the existing layout
manager approach to creating layouts from a template, e.g. by showing
a nice list of available templates via thumbnail graphics.

Unfortunately both approaches aren't trivial, and would require some
hours to implement.

Nyall


> c
>
> On Mon, Apr 2, 2018 at 10:01 PM, Alessandro Pasotti 
> wrote:
>>
>> The reason for this change is that your can now have different settings
>> for each page.
>>
>> The page size is not a property of the layout anymore but it's a property
>> of the page.
>>
>> But I agree that the page properties might be better exposed.
>>
>>
>> On Mon, Apr 2, 2018, 21:56 C Hamilton  wrote:
>>>
>>> I was just going to ask what happened to the "Page Size" properties in
>>> the QGIS 3 print composer, but I stumbled upon it accidentally by
>>> right-mouse clicking on the page and selecting "Page Properties." This was
>>> not easily found. In the QGIS 2 print compose it is a part of the layout. In
>>> QGIS 3 I just lucked out.
>>>
>>> Could the page properties be more obvious to find? At lease add them to
>>> the Layout or Edit menus or put them back under the layout settings.
>>>
>>> Thanks,
>>>
>>> Calvin
>
>
> --
> --
> Carlo A. Bertelli
>Charta servizi e sistemi per il territorio e la storia ambientale srl
>   Dipendenze del palazzo Doria,
>   vc. alla Chiesa della Maddalena 9/2 16124  Genova (Italy)
>   tel./fax +39(0)10 2475439  +39 0108566195  mobile:+39 393 1590711
>e-mail: berte...@chartasrl.eu  http://www.chartasrl.eu
> --
>
>
>
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS3: ogr2ogr canceling impossible

[Nyall Dawson]

On 31 March 2018 at 01:28, Tobias Wendorff
 wrote:
> Hi there,
>
> I'm using QGIS3, build f07389b421 on Windows.
>
> It's not possible to cancel ogr2ogr processing by pressing the button
> on my system. I had to write a script to cancel ogr2ogr manually.
>
> Is it just me or shall I file a bug-report?

It's kind-of a bug - the cancel button should be disabled because
there's currently no support for cancelling gdal/ogr algorithms.

What script did you use to add cancelation support?

Nyall

>
> Best regards,
> Tobias
>
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS3 Layout - North Arrow?

[Nyall Dawson]

On 28 March 2018 at 04:56, Bernd Vogelgesang  wrote:
> I think it is not worth the effort to ask for that.
> The devs had plenty of time to add a specific north arrow tool, but
> obviously had no great interest in it.
> The map canvas and the composer window are completely different things, so
> there is no automatism to have the same features within both.
> And actually it is quite clever to use the same tool for adding north arrows
> or any other image with one button ... it's just about telling the users
> where to look for.

Actually I disagree - I think there's room here to add a dedicated
"north arrow" button to layout designer which adds a picture item but
automatically defaults it to a nice north-arrow picture and sets to
the match the main map's rotation. There would be no code duplication
- it would still just be a standard picture item, but with some nice
defaults so that new users can easily create north arrows.

We do this approach in 3.0 with "line arrow" items too - effectively
they are just the same as polyline items, but default to having an
arrow marker on their end point.

Nyall



>
>
>
> Am 27.03.2018, 17:37 Uhr, schrieb knussear :
>
>> Well I have to go and find north arrows if I want them in my print layout.
>> But when I'm just viewing them in my map they are available under
>> "decorations". It would be great if I could access the "decorations" in
>> the
>> print layout - since they already exist somewhere within the app.
>>
>>
>>
>> --
>> Sent from: http://osgeo-org.1560.x6.nabble.com/QGIS-User-f4125267.html
>> ___
>> Qgis-user mailing list
>> Qgis-user@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
>
>
> --
> Bernd Vogelgesang
> Siedlerstraße 2
> 91083 Baiersdorf/Igelsdorf
> Tel: 09133-825374
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Bug Report - QGIS Crashes when label are Enabled for Counter SHP from Raster Extraction on Windows 10 & Mac

[Nyall Dawson]

On 27 March 2018 at 10:05, Elvin Roldan  wrote:
> Hello:
>
> Im writing to the user email to report a very unusual bug that I have 
> encountered on QGIS 3.0 for Windows and Mac version. I use a windows 10 mPC 
> at work and a Mac for my personal Desktop. The bug occurs when tables are 
> enabled for a Contour SHP created out of the Extraction (algorithm) tool on 
> the Raster menu. Once you enable the labels an zoom in or out and pan in 
> either direction the status bar keeps continuously active and the rendering 
> on the main screen never develops keeping a blank canvas until QGIS Crashes o 
> r a force quit has to be executed.

Sounds like a bug which was fixed in 3.0.1

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] [QGIS-Developer] Mac QGIS 3 GRASS fix

[Nyall Dawson]

On 23 March 2018 at 08:47, William Kyngesburye  wrote:
> Well, it seems that running any GRASS Processing coughs up this error.
> Mystery to me, permissions are correct for the app, nothing should be
> writing to the app, and if it is trying that's wrong.
>
> The error doesn't help me try to figure out what's going on.  Help would be
> appreciated.


I believe it's this bug:

https://issues.qgis.org/issues/17555

The permission errors is caused because it's actually trying to
execute /Applications/QGIS3.app/Contents/MacOS/grass/etc/colors/grass
instead of the correct grass binary.

The bug is caused by this function:

https://github.com/qgis/QGIS/blob/master/python/plugins/processing/algs/grass7/Grass7Utils.py#L130

It's trying to automatically generate the grass executable path, but
it's incorrectly finding the colors/grass folder and deciding that
it's the grass executable. I think inserting a os.path.isfile check
somewhere around line 145 should fix this, but not having access to an
osx install I haven't been able to test this.

Hope that helps -- would be great to have this one sorted for 3.0.1 if
you can confirm a fix!

Nyall

>
> On Mar 22, 2018, at 1:10 PM, William Kyngesburye 
> wrote:
>
> What are you doing?  GRASS plugin tools or Processing GRASS algorithms?
>
> Nothing should be trying to write to the GRASS dir.  There should not be
> permissions problem reading from the GRASS dir.
>
> On Mar 22, 2018, at 11:00 AM, jehduke...@gmail.com wrote:
>
> Thanks William.  I'm still getting some sort of error using grass.  Not sure
> if this is still related.
>
> /var/folders/d3/q41sw4cj5_s479lpjf1z3hwcgn/T/processing_7fd31b05f63f437c84dee066f864d9f9/grassdata/temp_location/PERMANENT:
> /Applications/QGIS3.app/Contents/MacOS/grass/etc/colors/grass: Permission
> denied
> /var/folders/d3/q41sw4cj5_s479lpjf1z3hwcgn/T/processing_7fd31b05f63f437c84dee066f864d9f9/grassdata/temp_location/PERMANENT:
> /Applications/QGIS3.app/Contents/MacOS/grass/etc/colors/grass: Permission
> denied
>
>
>
>
> On Thu, Mar 22, 2018 at 8:36 AM, William Kyngesburye 
> wrote:
>>
>> The non-working GRASS in the Mac QGIS 3 should be fixed now.  It appears
>> the GRASS internals still don't like spaces in the path to GRASS -- this
>> came from my adding "3" to the QGIS app name with a space.  The app is now
>> without a space, "QGIS3.app".
>>
>> It looks like the missing SAGA is also fixed by the rebuild (don't know
>> why, it just happened).
>>
>> REMINDER: My QGIS 3 package requires the python.org Python 3.  Other
>> distributions/installations of Python are not supported (ie Homebrew,
>> Macports, Anaconda...).
>>
>> -
>> William Kyngesburye 
>> http://www.kyngchaos.com/
>>
>> [Trillian]  What are you supposed to do WITH a maniacally depressed robot?
>>
>> [Marvin]  You think you have problems?  What are you supposed to do if you
>> ARE a maniacally depressed robot?  No, don't try and answer, I'm 50,000
>> times more intelligent than you and even I don't know the answer...
>>
>> - HitchHiker's Guide to the Galaxy
>>
>>
>> ___
>> Qgis-user mailing list
>> Qgis-user@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
>
>
> -
> William Kyngesburye 
> http://www.kyngchaos.com/
>
> All generalizations are dangerous, even this one.
>
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
>
> -
> William Kyngesburye 
> http://www.kyngchaos.com/
>
> "We are at war with them. Neither in hatred nor revenge and with no
> particular pleasure I shall kill every ___ I can until the war is over. That
> is my duty."
>
> "Don't you even hate 'em?"
>
> "What good would it do if I did? If all the many millions of people of the
> allied nations devoted an entire year exclusively to hating the  it
> wouldn't kill one ___ nor shorten the war one day."
>
>  "And it might give 'em all stomach ulcers."
>
> - Tarzan, on war
>
>
> ___
> QGIS-Developer mailing list
> qgis-develo...@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Newbie coding question

[Nyall Dawson]

On 21 March 2018 at 07:29,   wrote:
> So, I'm trying to do bunch of shortest distance calculations using two point
> layers and a line layer.  Reading the forums etc. I gather there isn't a GUI
> solution for this.  As a result, I'm trying to adapt the code here:
>
> https://gis.stackexchange.com/questions/171896/how-to-calculate-shortest-path-from-points-in-one-layer-to-points-in-another-lay
>
> I'm using Qgis version 3.0.2.

That code won't work in version 3 - it's using the old 2.x API.

In QGIS 3.0 there's two built-in processing algorithms "Shortest path
(layer to point)" and "Shortest path (point to layer)" which sound
like they almost do what you need (I gather you need "Shortest path
(layer to layer)"?). If so, I'd start with either of these and adapt
it to do full layer-to-layer joins:

https://github.com/qgis/QGIS/blob/master/python/plugins/processing/algs/qgis/ShortestPathLayerToPoint.py
https://github.com/qgis/QGIS/blob/master/python/plugins/processing/algs/qgis/ShortestPathPointToLayer.py

If you get close (or even if you get stuck), open a PR against the
QGIS github repo with your work and we'll guide you through. This
would be a handy algorithm to have available out-of-the-box.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Atlas Map Book Design

[Nyall Dawson]

On 19 March 2018 at 10:42, Springfield Harrison  wrote:
> Okay, thanks Phil. Actually I was wanting to only have the title block
> appear on selected pages, probably page 1 and not appear on the others. I'm
> not sure if this would need to be done through programming, as in "if page
> number equals 1, then visible". But changing for each page might be useful
> also.
>

Actually it's super-simple, in both 2.x and 3.0.

- select the item
- scroll to the "Rendering" section in Item Properties
- click the data defined button to the right of "Exclude item from
exports", hit "edit", and enter an expression to control whether the
item is hidden from that page, e.g. "@layout_page = 1"

Hope that helps!
Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS 3 OS X/macOS

[Nyall Dawson]

On 13 March 2018 at 04:29, William Kyngesburye  wrote:
> Sorry for the long wait.  It's been a hectic year - personal life changes 
> (marriage), but I'm finally getting back into gear.
>
> My QGIS 3 package for OS X/macOS is ready.

Champion! I'm sure you've made a lot of people very happy with this
announcement.

For those who know what you do, your work is very much valued! It's a
shame that packaging is somewhat underappreciated and under-supported
by the wider QGIS user community :(

Nyall


  Besides the big QGIS release, there are a couple other big changes
in the packaging.
>
> - Minimum OS X 10.10 Yosemite
>
> - Requires Python 3.6 from python.org (note it must be this and not homebrew 
> or other distribution).
>
> - except for what I include in the GDAL Complete package, all extra necessary 
> python modules are available from pypi with pip.  These are installed by the 
> QGIS installer and need an internet connection at that time.
>
> Make sure to install Python 3 first, otherwise the GDAL Complete python 
> components will not be installed (these are also required by QGIS).
>
> Currently the globe plugin is not included, but I'll get that out in an 
> update soon.  The new QGIS 3D features are included.
>
> GDAL format plugins will follow soon (ECW, MrSID, GRASS).
>
> -
> William Kyngesburye 
> http://www.kyngchaos.com/
>
> The equator is so long, it could encircle the earth completely once.
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] 3.0.0 Italic Text Issue

[Nyall Dawson]

On 8 March 2018 at 05:22, Taylor, Jeffrey R  wrote:
>
> Hello QGIS,
>
>
>
> I just downloaded v 3.0.0 and now I only have two choices for text style: 
> Regular and Bold. When hovering the cursor over the Italic text icon 
> (grayed-out), it shows, “Italic text (data defined only, overrides Style).” 
> How do I restore the Italic function?

Hi Jeff,

That's not normal (at least, not intended). Can you share your
platform and font? Do you see this for all font families, or just one?

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Why is pandas not part of the QGIS python package?

[Nyall Dawson]

On 8 March 2018 at 22:30, Ivan Dochev  wrote:
> Hi,
>
>
> I apologize if this was already discussed, but is there a specific reason
> why pandas isn't part of the QGIS python distribution? I saw a couple of
> discussions on how to get it installed, however every dependency
> installation lowers the "user-friendliness" - imagine I write a script,
> format it is a processing tool and send it someone, explaining that he/she
> has to install dependencies is an additional step, which could,
> theoretically, be avoided...
>
> Pandas is an extremely wide used library, so this must have already come up?

I think you're mixing some things up -- there's no such thing as a
"QGIS python distribution".

If you're running a linux distro then python libraries can be
installed using your normal distro package installer. If instead
you're referring to the Windows build then you'll need to raise this
with the osgeo4w project and get pandas packaged there.

If you're using a mac... I'd just dual boot and save yourself a
billion issues and much frustation.*

Regards,
Nyall

* My opinion. Not that of the project.
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Digitizing points with azimuth and distance

[Nyall Dawson]

On 5 March 2018 at 18:20, Paolo Cavallini  wrote:
> Hi all,
> I cannot recall an easy solution for this problem:
> * clicking on a point
> * adding azimuth an distance
> * automatically creating a point, in a second layer, at the distance and
> azimuth specified
> * repeat the procedure, adding more point to the same second layer.
> Anyone has a good solution to this?

Not sure if it helps - but 3.2 has an inbuilt "project points"
processing algorithm which allows you to shift the points from an
input layer by a given bearing and distance (also supports
data-defined/expression based bearing/distance values).

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Where is georeferencer plugin in QGIS 3 ?

[Nyall Dawson]

On 6 March 2018 at 01:58, Andreas Neumann  wrote:
> Hi,
>
> It is a core plugin. You can enable/disable it in menu "Plugins" --> "Manage
> and Install Plugins" --> "Installed".
>
> Once enabled, you find it in menu "Raster" --> "Georeferencer" -->
> "Georeferencer".

I've been wondering... is there any good reason we don't always enable
georeferencer and require users to manually check it on?

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Multipart Split

[Nyall Dawson]

On 1 March 2018 at 20:36, Vincent Picavet (ml) <vincent...@oslandia.com> wrote:
> Hello,
>
> On 26/02/2018 08:47, Nyall Dawson wrote:
>> On 26 February 2018 at 17:08, Andreas Neumann <a.neum...@carto.net> wrote:
>>> Hi Alex,
>>>
>>> If this is plugin is so useful and missed - wouldn't it make sense to have
>>> this functionality in QGIS core instead of a plugin?
>>>
>>> I am in favor of having more powerful editing tools in core - rather than
>>> forcing the user to install many plugins - which has its own problems.
>>
>> Stay tuned Andreas - North Road has a crowdfunding campaign we are
>> going to launch soon which will dramatically increase the power of
>> QGIS editing tools!
>
> Before doing the call for funding, do you intend to write a QEP with the
> proposed plan for modifications to these tools ?

See https://github.com/qgis/QGIS-Enhancement-Proposals/issues/114

>
> We have a lot of users depending on QGIS editing tools heavily, and they
> would like to have a clear visibility on what would be implemented, so
> as to get ready for it when it comes.
> They also have a lot of experience in CAD-oriented software, and would
> be able to give feedback and advices on editing tools behaviour.
> Following the QEP process would be good to gather these feedback.

This proposal doesn't actually alter any of the current editing tools
- rather it exposes a new way to use (certain) processing algorithms
to edit features in place -- such as separating a multipart feature
into single part features (or reversing line direction, translating
features, repairing geometries, etc).

Feedback and suggestions are definitely welcome!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Multipart Split

[Nyall Dawson]

On 26 February 2018 at 17:08, Andreas Neumann  wrote:
> Hi Alex,
>
> If this is plugin is so useful and missed - wouldn't it make sense to have
> this functionality in QGIS core instead of a plugin?
>
> I am in favor of having more powerful editing tools in core - rather than
> forcing the user to install many plugins - which has its own problems.

Stay tuned Andreas - North Road has a crowdfunding campaign we are
going to launch soon which will dramatically increase the power of
QGIS editing tools!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS SWOT at community meeting Madeira

[Nyall Dawson]

On 23 February 2018 at 01:19, Vincent Picavet (ml)
 wrote:
> Hi all,
>
> This is the current state of the SWOT :
> https://share.oslandia.net/public/679ef4/dl/20180222_123736.jpg
>
> Do not hesitate to send notes to add.
>
> Nyall, too bad you will not be available to share, timezones are hard to
> conciliate. We will publish the recording anyway so that you can still
> react afterwards.

Ahhh screw it... I really want to attend. Tim/others can you ping me
before the session and if I can find enough coffee I'll be there...

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] [QGIS-Developer] Data defined @symbol_color outlines

[Nyall Dawson]

On 22 February 2018 at 01:31, McDonaldR  wrote:
> Hi List
>
>
>
> This issue (https://issues.qgis.org/issues/15132) refers to using the
> @symbol_color function to make the outline of graduated/categorised features
> the same colour as the fill.  There is a fix by Nyall from over a year ago
> but I wonder if there has been a regression?
>
>
>
> I am using both the latest 2.99 master and 2.18.16 64-bit versions on
> Windows 7 and if I render a polygon layer using the categorised or graduated
> renderer and set the outline colour using @symbol_color then the legend
> doesn’t reflect the style shown on the map. The features are rendered
> correctly (I like using “darker(@symbol_color,150)”) on the map canvas and
> the print composer but the legend in the table of contents and on the
> composer layout show the default outline colour.

This has never been supported - all data defined properties don't
currently play well with the layer tree or map legends (and never
have). Maybe something for 3.2...

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS SWOT at community meeting Madeira

[Nyall Dawson]

On 22 February 2018 at 04:27, Vincent Picavet (ml)
 wrote:

> We will have a meeting at the community meeting on Friday 16:00 local
> time, to discuss about the items gathered, try to summarize some ideas
> and then write a report.

I'd have loved to take part, but this time is totally incompatible
with my timezone! (UTC+11).

Looking forward to the recording!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] Hi-dpi 3.0 beta tests, your feedback wanted...

[Nyall Dawson]

Hi all,

If you're running a QGIS 3.0 beta and have a hidpi/retina display, I'm
after some feedback.

Can you please run the following commands in the Python console:

ss = iface.mainWindow().styleSheet()
iface.mainWindow().setStyleSheet( ss + ' QToolBar > QToolButton
{padding: 8px;} ')

and see if the result is an improvement? Try running just the second
command and tweaking the padding amount, and let me know:

1. your platform
2. the "icon size" setting in QGIS options
3. what you think the ideal padding should be.

For me, on Win 10, 48px icon size, setting the padding to 8px makes a
HUGE improvement in the appearance of the toolbars. But I think this
padding needs tweaking by platform/icon size, which is why I need to
feedback...

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] View freezing and style

[Nyall Dawson]

On 14 February 2018 at 23:38, Enrico Fiore <enricofi...@libero.it> wrote:
>
>> Il 14 febbraio 2018 alle 12.43 Nyall Dawson <nyall.daw...@gmail.com> ha 
>> scritto:
>> >
>>
>> Let me guess - it's a point layer?
>
> yes, it is
>
>> >
>> > Can be a bug?
>>
>> Yes, it is.
>
> I have to open a ticket?

I think there's already an open ticket. I looked into this a while
ago. The culprit is this line (added in 2.16, or 2.18. I can't
recall):

https://github.com/qgis/QGIS/blob/master/src/core/layertree/qgslayertreemodel.cpp#L1524

Which is very slow to calculate, and is called for every symbol in
your layer (so 2381 times) at every map pan/zoom operation.

I only had a brief opportunity to look into it and could not find a
quick fix for this.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] View freezing and style

[Nyall Dawson]

On 14 February 2018 at 22:26, Enrico Fiore  wrote:
> Hi,
> I have noticed that QGIS 2.18.16 has strange behavior, that I'm going to try
> to explain.
>
> The view freezes for 3 - 4 seconds after zooming or dezooming action.
> View freezing happen only if in the Project there is a specific layer with
> categorized style (the same layer with single simbol style doesn't cause the
> problem).
> The layer has 130421 feature (point) and 2381 items in the legend (after
> have applied the style) and the datasource is db (The problem raised happens
> both with data deriving from Oracle and PostgreSQL database).
>

Let me guess - it's a point layer?


> The behaviour didn't happen in qgis 2.8.4.
>
> Has anyone experienced the same problem ?
>
> Can be a bug?

Yes, it is.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Google Satellite Images v3

[Nyall Dawson]

On 12 February 2018 at 18:43, Tobias Wendorff
 wrote:
> Hi Stefan,
>
> Am Mo, 12.02.2018, 09:30 schrieb Stefan Giese:
>> I don't think so, but I am not sure. I didn't find anything about
>> this in the term of usage. If you publish something, sure you
>> have to follow the terms of use and put the attribution on your
>> maps...
>
> Yeah, I think that's the point. It's not part of the documentation.
> It's an old backfall for outdated Google Maps applications and where
> Javascript isn't or wasn't support. The TOS ("terms of usage" was
> the wrong name) allow only to use the static API nowerdays.
>
> @Jürgen: I think, that would be the legit way to go for QGIS.
> Static API allows to show a fixed image even inside QGIS (some code
> needed of course).

Keep in mind this is a 3rd party plugin - nothing at all to do with
the QGIS project.

In fact, speaking personally, I'd very much like to see this plugin
(and a handful of others with serious known issues) deprecated and NOT
be made available for 3.0. The fact is that there's some plugins for
2.x which have long standing, extremely serious issues and I'd like to
avoid users of 3.x being exposed to these...

But that's a separate discussion, bound to cause all sorts of conflict!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] [Qgis-psc] Nominations for the QGIS Board and a short announcement from your Board Chair

[Nyall Dawson]

On 7 February 2018 at 01:16, Tim Sutton  wrote:
> On a personal note I would like to share that I will be standing down from 
> the Board and PSC at the end of my tenure as chair

Tim -

It's with much sadness that I read this, but I can also totally
understand your decision in doing so. Under your leadership the QGIS
project has grown and gone from strength to strength, and I believe
that a HUGE part of this is directly due to your guidance of the
project. You've been everything we could possibly dream of in a chair.

I'd also like to say that I've found your welcoming attitude a great
inspiration. Just watching you at hackfests I've seen endless
occasions of how you always give everyone time and attention, how you
actively seek out new members of the community and take time to learn
their stories, and how courteous and respectful you are of everyone's
opinions.

The whole open-source software ecosystem would be a much better place
if we had more people like you in leadership!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Print Composer Templates in v2_99

[Nyall Dawson]

On 3 February 2018 at 09:28, Grant Boxer  wrote:
> I am trying to apply/use a template in print composer from v.2.18.16 but
> v2_99 does not seem to able to display it. Do we need to re-create our
> templates for the new version?
>

It's just an oversight. I'll fix.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Profiles

[Nyall Dawson]

On 3 February 2018 at 17:48, Saber Razmjooei
 wrote:
> Hi Paolo,
>
> Each of those profile tools uses a different library for plotting.
>
> I think for migration to QGIS 3, it is best to use Qt Charts module which
> has been GPLed since Qt 5.7:
>
> http://doc.qt.io/QtCharts/
>
> IMO, profile tool should be a core QGIS function.

+1 billion.

Related features which should be added at the same time:
- Allow the chart to show a profile of a line from existing z (or m)
values (e.g. no raster dem required)
- A processing algorithm to set z/m values in a line/point/polygon
from a raster band

Nyall


>
> Regards
> Saber
>
> On 1 Feb 2018 12:53, "Paolo Cavallini"  wrote:
>>
>> Hi all,
>> while revising featured plugins, I checked the current situation with
>> plugins dealing with profiles.
>> Currently searching for profile
>> http://plugins.qgis.org/search/?q=profile
>>
>> returns the following:
>>
>>  [87] Profile tool
>>
>> [625] ProfileFromPoints
>>
>> [789] swath profile
>>
>> [609] Temporal/Spectral Profile Tool
>>
>> [33] Profile from Line
>>
>> [294] qProf
>>
>> [376] VoGIS-ProfilTool
>>
>> [127] EarthBeat
>>
>> [1240] Geo Picker
>>
>> [407] geopunt4Qgis
>>
>> There is obviously a large overlap between many of them. We already
>> analysed this long ago, and things have not improved.
>> I will be grateful if users:
>> * could confirm which of the bunch they find most useful
>> * could confirm whether some of them should be deprecated because it is
>> not working well
>> * would help identifying areas of overlap, and help me convincing
>> authors to merge their work, possibly in just one plugin, better if
>> within Processing.
>> All the best, and thanks.
>> --
>> Paolo Cavallini - www.faunalia.eu
>> QGIS & PostGIS courses: http://www.faunalia.eu/training.html
>> https://www.google.com/trends/explore?date=all=IT=qgis,arcgis
>> ___
>> Qgis-user mailing list
>> Qgis-user@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Idea: Crowd-sourcing a batch-georeferencer plugin withQR code ?

[Nyall Dawson]

On 4 February 2018 at 06:32, Alex Mandel  wrote:
> I think you undervalue what's been done. See the code base, and
> re-use/adapt whatever components you want. You can run your own custom
> version.
>
> https://github.com/fieldpapers
>
> Could even integrate it with QGIS as a plugin, or flip that around and
> use QGIS server to create the tile layer used in the background.

Hmm, let's do the math:

- QGIS 3.0 composer adds support for plugin-based item types, allowing
Python plugins to add new items to a layout
- fieldpapers is open source, so easy to see and reproduce the logic
regarding what's embedded in those QR codes
- there's existing Python libraries for QR codes as images, which
could then easily be rendered inside a QGIS 3.0 QgsLayoutItem subclass

Go forth magic QGIS plugin authoring elves! Do your thing!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS3: browse to path by text

[Nyall Dawson]

On 20 January 2018 at 04:30, Tobias Wendorff
 wrote:
> Hi there!
>
> QGIS3's browser doesn't seem to have a simple input, where to
> directly enter a path like "Z:\geodata\subdir\anywhere\else\"
> in text, does it?
>
> I tried fast scanning "Z:\" and using the filter on it - without
> success.
>
> Am I doing something wrong, is there a workaround or was it simply
> not implemented?

It's not implemented yet. The filter suggestion Stefan made only
filters currently expanded directories, so doesn't help here.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Presets in Layers Panel

[Nyall Dawson]

On 30 December 2017 at 08:07, Lene Fischer  wrote:
>
> Hi,
>
> I´m trying to use the ‘Preset’ function in the Layers Panel.
>
> But I can only add one preset and then ‘Add Preset’ is greyed out.
>
>
> It might be me who use it wrong, but could I get someone to try it out before 
> I file a ticket.

Yep, you're using it wrong! :)

Think of creating presets like a "snapshot" of the current
configuration. The action is disabled if the current state already
matches an existing preset. So you need to make some change to the
configuration (e.g. toggling some layers), and then you'll be allowed
to create another preset.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Cool Features of QGIS

[Nyall Dawson]

On 6 December 2017 at 12:11,   wrote:
> Hi,
>
>
>
> I think this also can help you for the presentation material
> http://monde-geospatial.com/10-differences-between-commercial-and-open-source-gis-software/
>

Just be careful with articles from monde-geospatial. I've seen a lot
of very poorly researched posts on that site, especially when they are
comparing software.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Cool Features of QGIS

[Nyall Dawson]

On 6 December 2017 at 10:59, Clifford Snow  wrote:
> I'm doing a short presentation on Open Source GIS software this Friday. My
> audience are 99% arcgis users. I'd like to tell them about a couple WOW
> features in QGIS that ESRI doesn't provide or is just a pain to do.
>
> I have never used ArcGIS so I'm struggling. Any suggestions would be most
> appreciated. If you can point me to a website that I can get some images
> from.
>
> Sorry for the short notice, but I was just added to the agenda moments ago.
>

This could be a good place to start:

https://gisunchained.wordpress.com/2016/01/28/qgis-features-i-long-for-while-using-arcgis/

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Processing Model issue

[Nyall Dawson]

On 1 December 2017 at 23:10, Richard McDonnell  wrote:
>
> Hi,
> Ive an issue, which Im hoping someone could have a solution or point me in a 
> direction of a solution.
> We have developed a Processing model on QGIS 2.8, which interpolates a raster 
> from points and subtracts the resultant from a DEM to produce a depth raster.
> This tool works flawlessly in 2.8, but in 2.14, 2.18 and 2.99 it fails at the 
> Raster Calculator stage, producing the same error. I have ran this using both 
> GDAL and SAGA raster calculators, but no luck.
> See attached error Logs.
> I would appreciate any help in this.
> Regards,

Seems like the error is coming from SAGA:

Error: grid: memory allocation failed [1545.94mb]

If you run the SAGA command manually (via the command line), does it
work? Maybe the error is due to a newer SAGA version, and unrelated to
QGIS.
Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] Recordings from QGIS Australia Meetup 2017

[Nyall Dawson]

Hi all,

Last week we held the first ever QGIS Australia Meetup in Sydney. The
recordings from this meetup have just been released online and can all
be viewed under:

https://www.youtube.com/playlist?list=PLQbaw-XC9jWFsTbZlIyoYLGsE4iZEUjFM

There's stacks of great content available there - I encourage you all
to check it out!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Issue with coordinates in QGIS 2.10

[Nyall Dawson]

On 31 October 2017 at 16:48, Andreas Neumann  wrote:

> But not sure if this existed in version 2.10 - which is
> already quite old.


Yes - I'd STRONGLY suggest upgrading. 2.10 was one of the buggiest
QGIS releases in the 2.x cycle!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] minidump after running script

[Nyall Dawson]

On 20 October 2017 at 18:30, Luigi Pirelli  wrote:
> probably the problem is the facto that you use:
> https://qgis.org/api/2.18/classQgsFeature.html#ab0a934a1b173ce5ad8d13363c20ef3c8
> without cloning geometry... and then you remove the layer => the
> geometry pointer is no more valid.
>
> try to do
> geom = QgsGeometry( feature.geometry() )
> or better
> geom = QgsGeometry( feature.constGeometry() )


Just for future reference - this applies to 2.x only. In 3.0 it's safe
to use the original code and you don't need to make these copies.

Nyall


> Luigi Pirelli
>
> **
> * Boundless QGIS Support/Development: lpirelli AT boundlessgeo DOT com
> * LinkedIn: https://www.linkedin.com/in/luigipirelli
> * Stackexchange: http://gis.stackexchange.com/users/19667/luigi-pirelli
> * GitHub: https://github.com/luipir
> * Mastering QGIS 2nd Edition:
> * 
> https://www.packtpub.com/big-data-and-business-intelligence/mastering-qgis-second-edition
> **
>
>
> On 20 October 2017 at 10:08, Reginald  
> wrote:
>> Hi,
>>
>> I am trying to make a report by using a script.
>> The script looks as follows:
>>
>> from qgis.PyQt.QtCore import *
>> from qgis.PyQt.QtGui import *
>> from qgis.PyQt.QtWidgets import *
>> import processing
>> # selectiepolygoon layer contains one multipolygone
>> layer=None
>> for lyr in QgsMapLayerRegistry.instance().mapLayers().values():
>> if lyr.name() == "selectiepolygoon":
>> layer = lyr
>> break
>>
>> # Get the first feature from the layer
>> feature = layer.getFeatures().next()
>> # fetch geometry
>> geom = feature.geometry()
>> # remove selectiepolygoon from legend
>> my_id = layer.id()
>> QgsMapLayerRegistry.instance().removeMapLayers(my_id)
>>
>> layers = iface.legendInterface().layers()
>>
>> for layer in layers:
>> layerType = layer.type()
>> if layerType == QgsMapLayer.VectorLayer:
>> for feature in layer.getFeatures():
>> if feature.geometry().intersects(geom):
>> attr =feature.attributes()
>> for att in attr:
>> print (att.value)
>>
>> When I run this script in my project it terminates with a minidump.
>> I suspect that the geom is empty at some point?
>> Any idea what is wrong with this script?
>>
>> Regards,
>>
>>
>>
>> --
>> Sent from: http://osgeo-org.1560.x6.nabble.com/QGIS-User-f4125267.html
>> ___
>> Qgis-user mailing list
>> Qgis-user@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] Upcoming QGIS Australia Meetup (Sydney)

[Nyall Dawson]

Hi all,

Just a quick heads-up about an upcoming QGIS user group meetup in
Sydney, Australia on the 17th November.

Attendance is free, and we've got a great line up of speakers and
workshops planned. Numbers are strictly limited (due to venue size),
and tickets are going very fast.

You can find full details and secure your place here:

https://www.eventbrite.com.au/e/qgis-australia-meetup-2017-tickets-37996172575?utm-medium=discovery=social=attendeeshare=escb=cp=listing

Cheers,

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Please, shapefile of the world

[Nyall Dawson]

On 31 August 2017 at 20:51, labiancamaril...@libero.it
 wrote:
> Dear All,
> I need a shapefile (to modify) of the World.
> Can you help me?
> Thanks a lot

Here you go: http://www.naturalearthdata.com/downloads/10m-cultural-vectors

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Test

[Nyall Dawson]

On 30 August 2017 at 16:42,   wrote:
> Hi all,
>
>
>
> can anyone confirm if my mails to the list are received? Usually I always
> got copies of my own posts via the list but now not anymore. The config is
> OK, at “Receive your own posts”  “yes” is checked. I receive all posts from
> other users. Still I can see my posts in the archive. I just want to find
> out if the spam filter in my company behaves strange or if there is any
> problem with the list. Thank you!

Confirmed
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] QEP notification: Rework datum transform handling

[Nyall Dawson]

Hi all,

Following on from discussion at the recent hackfest, please see QEP 100 at

https://github.com/qgis/QGIS-Enhancement-Proposals/issues/100

regarding proposed changes to datum transformation handling within
QGIS. Anyone with interest in how QGIS handles datum transforms
(including the looming introduction of temporal dependant datum
transforms, such as GDA 2020) should read and comment on the proposed
changes.

Timing is tight for this one - but at the same time it's a "3.0 or
never" type change!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Selecting All and Selecting None

[Nyall Dawson]

On 21 July 2017 at 12:02, Johanna Botman  wrote:
> Thanks Nyall,
>
> But the only options I have there are:
> Copy
> Paste
> Change Colour
> Change Transparency
> Change Output Unit
>  ... and I was aware of those and use them often. No option to select or 
> deselect all.
>
> I'm using 2.18.9

ah - try in the layer tree, not the style dock

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Selecting All and Selecting None

[Nyall Dawson]

On 21 July 2017 at 11:45, Johanna Botman  wrote:
> Hi,
>
>
>
> I have a layer of suburbs and I style them using the Categorise function so
> that they are all different colours. There are 43 of them.
>
> By default they are then all selected to show on my canvas.
>
>
>
> Is there a way that I can deselect them all at once and then reselect them
> all? I can only see that I have to click each check box (43 times).

Right click one of them, you'll get a menu with these options.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Area Calculations - what the h... is going on?

[Nyall Dawson]

On 7 July 2017 at 00:38, magerlin  wrote:
> I normally use $area in the field calculator to calculate areas of polygon
> but now this is really weird.
>
> I have 5 polygons in the same shape file and tries to calculate the areas -
> but two of the polygon get areas of about 100 times as big as they should be
> while the 3 remaining polygons get ok areas. If I use area($geometry)
> instead I get OK areas:

Thanks for reporting this - I've identified the issue and it's fixed
now in master and the 2.18 branch (so fix will be in 2.18.11)

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Mini dump when adding SQL layers QGIS 2.18.8

[Nyall Dawson]

On 26 June 2017 at 08:54, Johanna Botman  wrote:
>
> Thanks Martin, for getting back to me.
>
>
>
> I have isolated that it this behaviour occurs when I load a layer from the 
> browser window. If I add it through the Layer | Add New Layer menu items I 
> don’t have a problem. So that confirms what you are saying.
>
> A shame, I like being able to drag and drop layers onto my projects. I’ll 
> have to investigate updating to the latest version.


That's precisely the issue which 2.18.9 fixed :)

Update and you'll be right...

Nyall

>
>
>
>
>
>
>
>
>
>
>
> Johanna Botman
>
> Asset & GIS Officer | Melton City Council
>
>P: 03 9747 7209
>
>E: johan...@melton.vic.gov.au
>
>W: melton.vic.gov.au
>
>
>
> From: Martin Bain [mailto:martin.b...@lismore.nsw.gov.au]
> Sent: Monday, 26 June 2017 8:47 AM
> To: Johanna Botman; qgis-user@lists.osgeo.org
> Subject: RE: Mini dump when adding SQL layers QGIS 2.18.8
>
>
>
> Hi Johanna,
>
> 2.18.8 was quickly replaced by 2.18.9 I think it had some issues – I don’t 
> know if any of them relate to your problem.
>
>
>
> Are you adding the layers from the browser window or the add layer menu?
>
>
>
> Are you using a named ODBC DSN or putting the SQL Server Host details 
> directly into QGIS?  I had mini dumps when I was using named ODBC connections.
>
>
>
> I’m afraid the only thing I can suggest, if you haven’t already tried, is to 
> try switching the way you are connecting and adding the layers.
>
>
>
> Cheers,
>
> Martin.
>
>
> ***
> The information in this message is privileged and confidential, intended only 
> for the use of the individual or entity named above. If you are not the 
> intended recipient you are hereby notified to delete the message and that any 
> dissemination, copying or use of its content is strictly prohibited. If you 
> have received this message in error please notify the Melton City Council.
> ***
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Geodesic Measure Tool With Bearing Information

[Nyall Dawson]

On 16 June 2017 at 06:36, C Hamilton  wrote:

> There are several missing methods to QgsDistanceArea
>
> 1. Return the shortest path line between two points. The input would be the
> two points and the maximum distance of travel before a new line segment is
> added to the path. It may be good to add an attribute that is the maximum
> number of line segments that the line can have.
>
> 3. Calculate the intersection of two paths given their starting points and
> bearing.

These sound very useful! I'd love to see them implemented.

> 2. Given a starting point, bearing, and distance, return the destination
> point.

This was recently added (in master) by Mark Johnson.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Geodesic Measure Tool With Bearing Information

[Nyall Dawson]

On 13 June 2017 at 00:20, C Hamilton  wrote:

>>
>> > Right now it could use the Vincinty algorithms in QGIS to calculate the
>> > distances, but eventually I would like to see the geographiclib
>> > algorithms
>> > used.
>>
>> I'd start with the bearing changes (since they are self contained, and
>> the methods for calculating bearing are already present in the QGIS
>> api). Then using geographiclib could be discussed as a separate set of
>> changes, to be decided on its own merit. If you mix these two changes
>> it's likely to get bogged down as one or other part of the changes are
>> debated.
>
>
> Ok.

This looks relevant:

https://github.com/sourcepole/kadas-albireo/commit/df64318e5198f40f0e96d27fcf9e179fbbf682fd

There's a lot of unrelated changes in there, but it also shows the
basics for using geographiclib within the QGIS c++ code.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] problems with 'geometry creator'

[Nyall Dawson]

On 13 June 2017 at 02:53, DelazJ  wrote:
> Btw, giving a deeper look to the help provided in the code, I wonder if
> "geom_to_wkt(line_merge(geom_from_wkt('MULTILINESTRING((0 0, 1 1),(1 1, 2
> 2))')))"
> and especially MULTILINESTRING((0 0, 1 1),(1 1, 2 2))') is a valid
> multilinestring (a multipart feature whose parts already touch themselves).
> Shouldn't it be two separate linestrings?

That's exactly what line_merge does - it takes a single
multilinestring geometry and any touching parts are converted to a
single connected linestring.

> Also, in the application function help line_merge(geometry) should instead
> be line_merge(geometry1, geometry2,.., geometry n), shouldn't it?

No, because it takes just a single multilinestring input, not multiple inputs.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] [QGIS-Developer] Geodesic Measure Tool With Bearing Information

[Nyall Dawson]

On 12 June 2017 at 18:43, Neumann, Andreas  wrote:
> +1 for getting these improvements in QGIS core, rather than in a plugin.
>
> fyi: https://map.geo.admin.ch/ has an interesting "all-in-one" measure tool,
> which could server as an inspiration for QGIS.
>
> If you start with a new measure line, you first get a circle with distance
> and bearing. It also shows markers with labels along the line with round
> distances (e.g. 500m, 1000m, etc.). If you close the polygon by going back
> to the first vertex, it autosnaps and shows the area. So it is really just
> one single tool, but it helps to show distances, bearings and areas - along
> with an elevation profile.

That's quite nice! I wonder why the bearing/radius circle is only
shown for the first line segment though...

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Geodesic Measure Tool With Bearing Information

[Nyall Dawson]

On 10 June 2017 at 02:30, C Hamilton  wrote:
> Several weeks ago I had proposed a geodesic measure tool that also includes
> the bearing from point A to B and bearing from B to A. Nyall suggested that
> this capability be added to the QGIS measure tool. The question is would you
> want this with the QGIS measure tool or should it be a separate plugin?

My strong vote would be towards extending the QGIS measure tool. To me
that's a better solution then requiring users to download a plugin,
and having to reimplement all the (complex) logic of the measure tool
again in a python plugin.

>
> 1. Where the current measure tool has 'Segments' there would be two more
> columns with 'Bearing to' and 'Bearing from'.

Sounds good!

>
> 2. I would like to see a check box or selection widget of some sort that
> chooses between using Geodesic math or the current method of calculating the
> distances which I am still unsure on how it is doing its calculations.

If you expand down the info box you'll get a detailed description of
exactly how QGIS is calculating the distances. Is there something
missing there?

> Right now it could use the Vincinty algorithms in QGIS to calculate the
> distances, but eventually I would like to see the geographiclib algorithms
> used.

I'd start with the bearing changes (since they are self contained, and
the methods for calculating bearing are already present in the QGIS
api). Then using geographiclib could be discussed as a separate set of
changes, to be decided on its own merit. If you mix these two changes
it's likely to get bogged down as one or other part of the changes are
debated.

> If this is of interest then I will attempt to get a QGIS development
> environment running to help make the changes to the core measure tool;
> otherwise, I will just create a plugin to do this. If I create a plugin I
> will use the geographiclib python library to do the calculations.
>
> Thoughts?

Go the core! Let us know if you need help setting up a dev
environment. It's actually much easier for 3.0 then it was for 2.x
(especially on windows).

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS 2.18.9 crash using Python console

[Nyall Dawson]

On 7 June 2017 at 21:20, johnrobot  wrote:
> Hi
> There is no additional code. I didn´t expect the first example to work, but
> I do consider the crash a bug. I get the same result in QGIS 2.14.6 on
> Windows 7. I will open a ticket for it.

Likely what's happened is you've run this in the Python console, and
somewhere in the session a previous command has stored a temporary
object in the geom variable. Otherwise Python would error out with a
message like ''NoneType' object has no attribute 'asPoint'.

I suspect it's a result of https://issues.qgis.org/issues/777, which
is fixed for 3.0

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS 2.18.9 crash using Python console

[Nyall Dawson]

On 7 June 2017 at 06:18, johnrobot  wrote:
> Hi
> I had a look at the python examples at
> http://www.qgistutorials.com/en/docs/getting_started_with_pyqgis.html. I
> noticed that when I run
>
> layer = iface.activeLayer()
> for f in layer.getFeatures():
>   print geom.asPoint()
>
> QGIS crashes and creates a dump file. Is this a known issue? Running
>

Where's geom set here? Can you show the rest of your code?

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] Layout and Reporting Engine Crowdfunding Campaign – a success!

[Nyall Dawson]

Hi all,

Just a quick update to let you know that our recent crowdfunding
campaign to rework QGIS' print composer for 3.0 was a huge success!

I'd like to thank every member of the QGIS community who assisted with
promoting this campaign and bringing it to the attention of the
campaign backers. Without your efforts we would not have been able to
reach these backers and the campaign would not have been successful.

Of course, our thanks also go to the many financial sponsors for this work:

- The Swiss QGIS User group (funded preliminary work in refactoring
how compositions are handled and stored within projects)
- The QGIS Grant Program (2016) (funded preliminary work allowing for
a flexible property framework which will be used within the project)
- Land Vorarlberg, Austria
- Tudor Bărăscu
- Oester Messtechnik GmbH, Thun, Switzerland
- Ferdinando Urbano, Italy
- Alta ehf, Reykjavik, Iceland
- Trage Wegen, Belgium
- QGIS Denmark User Group, Denmark
- Debuco Techniek B.V., Netherlands
- The QWAT Project
- QGIS Sweden User Group, Sweden
- Andreas Neumann
- Kristianstads kommun, Sweden
- City of Vevey, Switzerland
- Kanton Thurgau, Switzerland
- Ricardo Pinho, Portugal
- Portugese QGIS User Group, Portugal
- Waymotion, Portugal
- Lloyd Analytics LLC, USA
- Canton of Neuchâtel, Switzerland
- Hartmut Dickel, Switzerland

We’ve also received numerous anonymous contributions in addition to
these – please know that the QGIS community extends their gratitude
for your contributions too!

This campaign was also successful thanks to The Agency for Data Supply
and Efficiency, Denmark, who stepped up and have funded an initial
component of this project directly.

I'll be posting updates as this work progresses over at
https://north-road.com/blog/

Many thanks!
Nyall


...
(the scary bozo)
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Selecting displaced symbols

[Nyall Dawson]

On 1 June 2017 at 09:26, Martin Bain <martin.b...@lismore.nsw.gov.au> wrote:
> Cool!  Maybe I didn't imagine it after all.

No - you did ;) That commit isn't merged into master, it's sitting
orphaned and alone and rotting away along with a bunch of other
abandoned experiments...

Nyall

>
> -Original Message-
> From: Nyall Dawson [mailto:nyall.daw...@gmail.com]
> Sent: Thursday, 1 June 2017 9:21 AM
> To: Neumann, Andreas <a.neum...@carto.net>
> Cc: Martin Bain <martin.b...@lismore.nsw.gov.au>; qgis-user 
> <qgis-user@lists.osgeo.org>
> Subject: Re: [Qgis-user] Selecting displaced symbols
>
> On 30 May 2017 at 19:11, Neumann, Andreas <a.neum...@carto.net> wrote:
>> Hi Martin,
>>
>> As far as I know, picking is totally independent of the rendering. So
>> the "Info tool" doesn't know about the displacement renderer. I know
>> it is not ideal, but it is a technical limitation and there are no known 
>> workarounds.
>
> Well - there's this half finished commit I wrote but never merged:
> https://github.com/nyalldawson/QGIS/commit/2405c9cc2c9d07bd0975d60ce36db5c5e8d8b612
>
> I can't remember exactly why I didn't complete this. I think it was written 
> during feature freeze once and I never got around to verifying that it didn't 
> cause regressions in the rendering speed after freeze was lifted. From memory 
> it worked well for info tool and rotate point symbol tool.
>
> From the commit message:
>
> "Register rendered point features into a spatial index so that we can later 
> retrieve features by their actual rendered location.
> (The rendered location may be different in numerous circumstances, eg when 
> using the displacement renderer, or when marker symbols have an offset set) 
> This index could then be used to later retrieve the features for a specific 
> canvas coordinate, for use eg in identify tool, rotate point symbols, etc."
>
> Theoretically this could be cleaned up, finished off and tested, and then 
> later extended to apply to selections also.
>
> Nyall
>
> This email is intended for the named recipient only.  The information it 
> contains may be confidential.  If you are not the intended recipient you must 
> not reproduce or distribute any part of this email, disclose its contents to 
> any other party or take any action in reliance on it.  If you have received 
> this email in error, please contact the sender immediately and delete the 
> message.
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Snapping: Move Features vs Node Tool

[Nyall Dawson]

On 1 June 2017 at 09:20, Martin Bain  wrote:
> Hi Randy,
>
> I've come to accept that's just the way it is.  It would be nice if the move 
> tool did snapping.  I guess it might be tricky to implement - with the node 
> tool the cursor position and node being moved are one and the same, but with 
> the move tool that’s not always the case.

Not long to wait now - snapping (and the advanced digitizing dock)
works with the move tool in 3.0

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Can we make QGIS stop lying?

[Nyall Dawson]

On 26 May 2017 at 07:37, Nyall Dawson <nyall.daw...@gmail.com> wrote:
> On 26 May 2017 at 07:25, Tobias Wendorff <tobias.wendo...@tu-dortmund.de> 
> wrote:
>> Am Do, 25.05.2017, 22:39 schrieb Redoute:
>> <http://qgis.org/api/2.18/classQgsDistanceArea.html#a9877c70b2a9c90e89cf7211fb82176b1>
>>
>> Oops, there it is:
>> QgsDistanceArea::ellipsoidSemiMajor
>> QgsDistanceArea::ellipsoidSemiMinor
>>
>
> Looks like there's a fix for this in 3.0 already - I'll backport now.

Ok - here's the correct fix:
https://github.com/qgis/QGIS/commit/7e089a1c01c3996a43a90c79caa6fb549a737bdd

You were very close - the tricky part here is reading the project
ellipsoid setting. That's not an easy thing to do in QGIS 2.x

The equivalent fix in 3.0 looks like:
https://github.com/qgis/QGIS/blob/master/python/plugins/processing/algs/qgis/HubDistancePoints.py#L118

(This part of the API has been cleaned up a lot!)

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] A Couple of Plugin Ideas

[Nyall Dawson]

On 26 May 2017 at 05:53, C Hamilton  wrote:

> I would like a measure tool that provides the distance between points A & B,
> the bearing from point A to B, and the bearing from point B to A.
>

This certainly sounds useful!

I'm wondering - instead of adding this as a plugin, how about adding
it into the default QGIS measure tool instead? I'm happy to help out
with mentoring c++ changes if you need it.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Can we make QGIS stop lying?

[Nyall Dawson]

On 26 May 2017 at 07:25, Tobias Wendorff  wrote:
> Am Do, 25.05.2017, 22:39 schrieb Redoute:
> 
>
> Oops, there it is:
> QgsDistanceArea::ellipsoidSemiMajor
> QgsDistanceArea::ellipsoidSemiMinor
>

Looks like there's a fix for this in 3.0 already - I'll backport now.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS Composer Rewrite and Layout Engine crowdfund – half way there!

[Nyall Dawson]

On 22 May 2017 at 18:24, John Helly <hel...@ucsd.edu> wrote:
> Get out of the fund-raising mode.  this is not about money. Scarey bozo.

Hi John,

Thanks for that valuable contribution. It's a very interesting way to
reflect the values of the University of California on a public forum.
With insights like this it's easy to see how you were selected for the
position of Laboratory Director at the UCSD Supercomputer Center...

Nyall


>
>
>
>
> On 5/21/17 8:33 PM, Nyall Dawson wrote:
>>
>> Hi all,
>>
>> If you’ve been following North Road's blog posts, you’ll be aware that
>> we are currently running a crowd funding campaign to extend the
>> capabilities of QGIS’ print composer. You can read full details about
>> this over at the campaign page
>>
>> (https://north-road.com/2017/04/26/qgis-composer-rewrite-and-layout-engine-crowdfund-launched/)
>>
>> The good news is that we’ve just hit the mid way point of the funds!
>> Many generous backers have stepped up with contributions and we’re
>> well on the way to reaching the funding goal. However, we still need
>> your help make this work a reality.
>>
>> Right now, what we need most is interested users and community members
>> who will reach out to their local QGIS users and seek more backing for
>> the campaign. We need to publicise the campaign beyond the regular
>> online QGIS community, to the thousands of enterprises and
>> organisations which rely on QGIS for their daily mapping operations.
>> We need community members who can get in contact with these
>> organisations and help convince them that investing back into the open
>> source software they utilise is beneficial (and often will even SAVE
>> them money in the long run, due to the increased productivity that
>> changes like our composer improvements will bring!).
>>
>> So, while social media reshares have been vital to reaching the
>> current stage, we now need more “hands on” helpers who will take this
>> on. If you know of any organisations which depend on QGIS for their
>> mapping outputs, now’s the time to get in contact with them directly
>> and advise them of this campaign!
>>
>> Thanks in advance,
>> Nyall
>> ___
>> Qgis-user mailing list
>> Qgis-user@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
>
> --
> John Helly, University of California, San Diego / San Diego Supercomputer
> Center / Scripps Institution of Oceanography / 760 840 8660 mobile /
> http://www.sdsc.edu/~hellyj
> ORCID ID: orcid.org/-0002-3779-0603
>
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] QGIS Composer Rewrite and Layout Engine crowdfund – half way there!

[Nyall Dawson]

Hi all,

If you’ve been following North Road's blog posts, you’ll be aware that
we are currently running a crowd funding campaign to extend the
capabilities of QGIS’ print composer. You can read full details about
this over at the campaign page
(https://north-road.com/2017/04/26/qgis-composer-rewrite-and-layout-engine-crowdfund-launched/)

The good news is that we’ve just hit the mid way point of the funds!
Many generous backers have stepped up with contributions and we’re
well on the way to reaching the funding goal. However, we still need
your help make this work a reality.

Right now, what we need most is interested users and community members
who will reach out to their local QGIS users and seek more backing for
the campaign. We need to publicise the campaign beyond the regular
online QGIS community, to the thousands of enterprises and
organisations which rely on QGIS for their daily mapping operations.
We need community members who can get in contact with these
organisations and help convince them that investing back into the open
source software they utilise is beneficial (and often will even SAVE
them money in the long run, due to the increased productivity that
changes like our composer improvements will bring!).

So, while social media reshares have been vital to reaching the
current stage, we now need more “hands on” helpers who will take this
on. If you know of any organisations which depend on QGIS for their
mapping outputs, now’s the time to get in contact with them directly
and advise them of this campaign!

Thanks in advance,
Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] Remaining "blocking" operations

[Nyall Dawson]

Hi all,

With a recent PR merge by Matthias, another one of the remaining
interface blocking operations is gone and moved to task manager. This
means that task manager is now being used for:
- saving vector/raster layers
- importing data to browser destinations via drag and drop
- saving maps as rasters/pdf from canvas (not composer)
- calculating feature counts for layer tree

There's also been some recent changes in composer which prevents the
interface blocking while refreshing map item, and a funded change by
the Swiss user group which dramatically decreases project load times
while compositions are present

With these changes QGIS master is feeling extremely responsive. So now
I'm wondering what other "blocking" operations remain in master.

A quick list off the top of my head:
- updating the snapping index
- exporting compositions/atlases
- running processing algs (work underway to fix this)
- calculating categorized/graduated renderer classes for large layers

What have I missed? What other tasks currently "lock up" the QGIS
interface while performing?

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Crashing

[Nyall Dawson]

On 10 May 2017 at 13:54, Allan Younger  wrote:
> I am having major problems with the latest QGIS 2.18.7 crashing all the time
> with only a very small number of layers open.
> Have uninstalled & re-installed 3 times.
> Previous versions handled huge amount more data.
> Running a 64 bit 16gb wind 7pro system.
> Allan Younger

Can you please check:
1. do you have on the fly reprojection enabled?
2. are any layers in your project in a different projection to the
project itself?
3. if so, does removing these layers from the project help?

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] Announcing crowd funding campaign for composer rewrite and reporting framework

[Nyall Dawson]

Hi all,

I'm proud to announce the launch of North Road's latest QGIS crowd
funding campaign: the "QGIS Layout and Reporting Engine Campaign".

This campaign covers stage 1 of a large, ongoing project to modernise
and expand on QGIS’ print composer and layout facilities. Over time
QGIS’ composer functionality has grown extensively and now is capable
of creating flexible, high quality cartographic outputs. However,
we’ve now hit a limit where the current code architecture is
prohibiting further improvements and important fixes. In order to add
a reporting framework to QGIS, it is necessary for us to refactor and
improve large sections of the composer code.

You can read more about the campaign and what we're planning at

https://north-road.com/2017/04/26/qgis-composer-rewrite-and-layout-engine-crowdfund-launched/

Please reshare this link via social media, your local user groups, etc
- the more exposure we get the more likely this is to be successful!

Thanks,

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Number of vertices

[Nyall Dawson]

On 24 April 2017 at 11:52, Martin Bain  wrote:
> Hi Anke,
>
> The asPolyline() function of a geometry returns a list of points.  You can
> then simply get the length of the list:
>
>
>
>
>
> lyr=iface.activeLayer()
>
> sel=lyr.selectedFeatures()
>
> obj=sel[0].geometry()
>
> verts=obj.asPolyline()
>
> print verts
>
> [(526002,6.81305e+06), (526011,6.81309e+06), (526016,6.81311e+06)]
>
>
>
> print len(verts)
>
> 3

While this will work, it's quite slow as it requires copying all the
geometry's points into a new list. A faster approach is:

feature.geometry().geometry().nCoordinates()

In case you're wondering, the extra ".geometry()" gives you access to
the QgsAbstractGeometry interface containing all the low level methods
for working with geometry primitives.

Nyall


>
>
>
> Cheers,
>
> Martin.
>
>
>
> From: Qgis-user [mailto:qgis-user-boun...@lists.osgeo.org] On Behalf Of Anke
> zur Gölden
> Sent: Sunday, 23 April 2017 4:39 PM
> To: QGIS User List 
> Subject: [Qgis-user] Number of vertices
>
>
>
> Good morning
>
>
>
> is there a way to get the number of vertices for a Polyline?
>
> I am looking for the python method to get if possible
>
>
>
> thx
>
> Anke
>
>
>
>
> This email is intended for the named recipient only. The information it
> contains may be confidential. If you are not the intended recipient you must
> not reproduce or distribute any part of this email, disclose its contents to
> any other party or take any action in reliance on it. If you have received
> this email in error, please contact the sender immediately and delete the
> message.
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] New layer from current view

[Nyall Dawson]

On 18 April 2017 at 20:50, Phil (The Geek) Wyatt  wrote:
> Hi Pat,
>
>
>
> · select just the parts of the current layer that you want to save
> (CTRL and select the parts)
>
> · right click on the layer name and save as
>
> · tick the box for 'save only selected features'

There's an easier way -

- right click the layer you want to save, click 'save as'
- tick the 'extent' check box and click 'map view extent'

done!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS Server Process Crash

[Nyall Dawson]

On 3 April 2017 at 08:48, Nyall Dawson <nyall.daw...@gmail.com> wrote:
> On 3 April 2017 at 08:29, Simon Greener <si...@spatialdbadvisor.com> wrote:
>> The installation can only be on Microsoft Windows Server.
>>
>> AFAIK it is a supported platform.
>
> Unfortunately, while it may be "supported" platform, it's not *well*
> supported. There's severe known issues running server under Windows
> which have been unaddressed for a long time, eg
> http://hub.qgis.org/issues/13293.
>
> I'm personally not aware of anyone using QGIS server under Windows in
> a production environment.
>
>> The supplied information is all that we can find that describes the error.
>>
>> The Apache and QGIS error logs show nothing before the process crash.
>>
>> The QGIS Projects are all OK when they are used in QGIS Desktop: no crash
>> occurs.
>>
>> The crashing has consistently occurred in both 2.14 LTR 32bit and 2.18.4
>> 64bit.
>> The crashing has stopped the application from going live.
>>
>> If anyone can point to some "standard" settings or things that need to be
>> checked, please supply.
>
> My suggestion - either Alessandro or the team at Oslandia are the most
> knowledgeable about QGIS server setup and operation. I'd engage them
> for this project and sponsor fixes through them for any blocking
> blocking bugs in the QGIS server operation on Windows.

3liz should also be in that list - sorry!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS Server Process Crash

[Nyall Dawson]

On 3 April 2017 at 08:29, Simon Greener  wrote:
> The installation can only be on Microsoft Windows Server.
>
> AFAIK it is a supported platform.

Unfortunately, while it may be "supported" platform, it's not *well*
supported. There's severe known issues running server under Windows
which have been unaddressed for a long time, eg
http://hub.qgis.org/issues/13293.

I'm personally not aware of anyone using QGIS server under Windows in
a production environment.

> The supplied information is all that we can find that describes the error.
>
> The Apache and QGIS error logs show nothing before the process crash.
>
> The QGIS Projects are all OK when they are used in QGIS Desktop: no crash
> occurs.
>
> The crashing has consistently occurred in both 2.14 LTR 32bit and 2.18.4
> 64bit.
> The crashing has stopped the application from going live.
>
> If anyone can point to some "standard" settings or things that need to be
> checked, please supply.

My suggestion - either Alessandro or the team at Oslandia are the most
knowledgeable about QGIS server setup and operation. I'd engage them
for this project and sponsor fixes through them for any blocking
blocking bugs in the QGIS server operation on Windows.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Alternative labels on each iteration?

[Nyall Dawson]

On 18 March 2017 at 06:40, Tobias Wendorff
 wrote:

> Since I've developed a way, how to shorten label the automatic way
> for ArcGIS, Mapnik and Mapserver, I'm missing this feature in QGIS.
> I love working with QGIS and telling other people, how to use it.

Thanks for the clarification - it's hard to read tone over email sometimes :)

I should add... I also would like this feature. In fact, I've already
sketched out a plan to implement it. Basically QGIS' labeling engine
needs a way to register multiple labels for a feature, with a
customised cost for each. The placement engine then needs a concept of
"mutually exclusive" labels - i.e. label candidates where only a
single candidate in the mutually exclusive group can be used in the
final labeling solution. This way multiple approaches could be used to
specify the list of possible alternative labels + their corresponding
cost: this could be via python, via data defined overrides, or just a
simple dropdown "alternative label" combo. Not *huge* changes, but
anything touching on the placement engine brings a certain degree of
complexity just by the nature of that code.

That said, I'm unlikely to take this any further myself unless I get a
complex mapping job which just can't be done without this feature (or
sponsorship). It falls firmly into the "too long to just whip up for
the sake of it, too complex to be an enjoyable coding job just for
coding's sake" category!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Alternative labels on each iteration?

[Nyall Dawson]

On 18 March 2017 at 02:57, Tobias Wendorff
 wrote:

> Actually, I don't have money to pay the developers. But I know some
> corporations, which would like to switch from ArcGIS to QGIS to save
> money on license costs. But since those are into cartography, QGIS
> isn't really useful for them right now. I might convince them to
> spend some money on QGIS, but I bet, the answer will be "no" :(
>

That's odd logic. To me this reads that QGIS developers should spend
hours of volunteer time to make QGIS the hypothetical "perfect"
cartographic tool, and only then maybe organisations will start to
support QGIS. That's not how open source works.

Let me suggest an alternative approach:

A very similar conversation occurred recently regarding missing
functionality with discrete raster styling. You can follow the
conversation here:
http://lists.osgeo.org/pipermail/qgis-developer/2017-January/046547.html

Basically, a user identified an important missing feature, but alone
was not able to develop this feature or fund its development. Instead,
the user organised a group of other users who were wanting this
functionality, and together they developed the specifications for the
change and fundraised for its development. The end result is that this
missing feature is going to land very shortly in QGIS master!

I suspect that you'll find many users would be interested in this
labeling enhancement, and that following a similar approach will lead
to the same successful end result.

> But since those are into cartography, QGIS isn't really useful for them right 
> now.

I may be biased, but I strongly disagree with this statement! Fact is,
the "ideal" cartography tool doesn't exist yet. But every time I have
to use one of the commercial alternatives I'm reminded just how much
MORE powerful QGIS symbology is. QGIS is already a cartographic
powerhouse, it just may not yet have every single function required
for all users. But the beauty of open source is that it's possible for
anyone to fix this!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] select by radius [v2.18.4, possible bug]

[Nyall Dawson]

On 13 March 2017 at 07:51, Tobias Wendorff
 wrote:
> Hi Martin,
>
> thanks for your reply.
>
> Am So, 12.03.2017, 22:43 schrieb Martin Bain:
>>
>> Its working for me on 2.18.4 32 bit Win 7.  Select by radius creates
>> a new selection, with the shift key it adds to existing selection,
>> control key removes from selection, alt key changes to select
>> only objects fully within radius.
>
> Yeah, this works for all the other modes (freehand etc.), but doesn't
> work for radius here. Let me install QGIS to a sandbox to get a clean
> installation.

Check your geometries. This has been experienced in the past as a
result of a layer with invalid geometries.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

[Qgis-user] Changing behaviour of "show selected features" mode for attribute table

[Nyall Dawson]

Hi all,

I'd like to raise discussion about changing the behaviour of the "show
selected features" mode in the attribute table.

Over the last couple of weeks I've pushed fixes to both 3.0 and
(shortly) 2.18 to improve the performance of the attribute table when
this mode is selected (Thanks to Faunalia and ENEL for sponsoring
this!). With these changes *only* the selected features are fetched
from providers to show in the attribute table, vs the current
behaviour of fetching *everything* and then filtering out to the
selection. It makes a huge difference for working with large layers.

Now - there's one last piece of this I'd like to land, but it changes
the behaviour of this mode. Currently if you have the table set to
"show selected features", but there's nothing selected, then ALL
features are shown.

This kills the benefit of setting the table to show in this mode. If
you accidentally open the table for a large layer with no selection,
it'll force every feature to be fetched again.

I'd like to change this, so that no selection = nothing shows in the
table. This means that users can safely set the attribute table to
always show in "selected features" mode and be confident that they'll
never hit the situation where every feature is fetched (unless of
course they have selected *every* feature!).

Does anyone object to this change landing for 3.0 and 2.18?

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] print composer

[Nyall Dawson]

On 4 March 2017 at 15:16, Boaz Bar Ilan <boazpro...@gmail.com> wrote:
> hallo
>
> you are right
>
> well, I am trying to export qgis produce into ppt . I want to save the image
> with a legend. I'm trying to activate composer manager in order to produce
> map and  legend and to export it as image to the ppt.  there is not reapond
> from the composer. an it be because of too many layers ?

Thanks for the extra detail!

How many layers have you got? Are any of them remote sources (eg a wms
or wfs server)?

Could you make a cut-down version of your project which demonstrates
this issue so that we can investigate?

Cheers,
Nyall


> thanks
> boaz
> p.s - qgis 2.14
>
> 2017-03-04 6:40 GMT+02:00 Nyall Dawson <nyall.daw...@gmail.com>:
>>
>> On 4 March 2017 at 13:54, Stephen Wandling <swandl...@gmail.com> wrote:
>> > Are you saying that cultural insensitivity is part of list etiquette?
>>
>> I'm just saying that this list is manned by humans. A terse message
>> with no detail at all provides no value to anyone. You only get out of
>> this list what you put into it.
>>
>> Nyall
>>
>>
>> >
>> > Stephen
>> >
>> >
>> > On 03/03/2017 6:50 PM, Nyall Dawson wrote:
>> >>
>> >> On 4 March 2017 at 03:30, Boaz Bar Ilan<boazpro...@gmail.com>  wrote:
>> >>
>> >>>
>> >>> print composer sopped responsing
>> >>>
>> >>> what to do ?
>> >>>
>> >>>
>> >>
>> >> I'd start by learning some mailing list etiquette - this isn't a chat
>> >> bot.
>> >>
>> >> Nyall
>> >> ___
>> >> Qgis-user mailing list
>> >> Qgis-user@lists.osgeo.org
>> >> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> >> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> >
>> > ___
>> > Qgis-user mailing list
>> > Qgis-user@lists.osgeo.org
>> > List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> > Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> ___
>> Qgis-user mailing list
>> Qgis-user@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
>
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] [Qgis-developer] QGIS loaded millions of entries

[Nyall Dawson]

On 4 March 2017 at 18:41, Alessandro Pasotti  wrote:
> On Sat, Mar 4, 2017 at 9:25 AM, Nathan Woodrow  wrote:
>>
>> Hmm are you sure we don't do this already?
>>
>> I suspect we do. Each provider can handle the sorting, giving to it by
>> QgsFeatureRequest, and hand the results back to the caller.  I thought the
>> attribute table was already implemented this way.
>>
>> Regards,
>> Nathan
>
>
> Hi,
>
> IIRC the last time I looked into it (way more than 1 year ago) the
> sorting/filtering on PostGIS was not yet implemented on the back-end, that's
> why the provider (invoked from the attribute table) was fetching all the
> records.
>
> Please correct me if I'm wrong, but I think that the sorting/filtering is
> now implemented in the backend (at least for PostGIS, not sure about the
> status of every provider)

Filtering yes, sorting no.

Every provider has *some* functionality for filtering on the provider
side, but with different levels of compatibility. Most providers
(exception of oracle and mssql) we've pushed this about as far as we
can, and the remaining expression functions just can't be mapped
across.

> but the attribute table will still fetch all
> records unless you limit the view to the map/project extent.

A change I just pushed to master also prevents this full table load if
you use the "show selected features" mode. So that helps a bit.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS loaded millions of entries

[Nyall Dawson]

On 4 March 2017 at 18:03, Tobias Wendorff
 wrote:
> Am Sa, 4.03.2017, 08:34 schrieb Alessandro Pasotti:
>>
>> Yes, that would be a strategy, but you might run into problem with
>> sorting and filtering if the data provider does not support for it.
>
> All database-like formats, even Shapefile's DBF, support sorting
> on database level. I've never understood, why some applications
> download the whole content first and

Because it depends on whether the front-ends filtering and sorting
functions match up with the backend's. In the case of QGIS our
expression engine supports a lot of functions which just can't be
translated to Postgres/PostGIS functions. In this case you have no
choice but to fetch everything and do the filtering/sorting on the
client side.

QGIS does have robust support for "compiling expressions", which is
basically translating them from QGIS' expression engine to functions
supported by the backend. This works great for simple expressions, and
depending on the provider even some complex ones (eg postgres provider
has large coverage of mapping qgis functions->provider functions. OGR
does not and can only handle simple field=value compilation). Some
providers also have support for compiling the sort order to perform on
the database side - unfortunately postgres does not (see
https://github.com/qgis/QGIS/blob/master/src/providers/postgres/qgspostgresfeatureiterator.cpp#L130).

I'm sure Matthias would love to re-enable that code and give postgres
a performance boost if someone wants to step up and sponsor it!

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] print composer

[Nyall Dawson]

On 4 March 2017 at 13:54, Stephen Wandling <swandl...@gmail.com> wrote:
> Are you saying that cultural insensitivity is part of list etiquette?

I'm just saying that this list is manned by humans. A terse message
with no detail at all provides no value to anyone. You only get out of
this list what you put into it.

Nyall


>
> Stephen
>
>
> On 03/03/2017 6:50 PM, Nyall Dawson wrote:
>>
>> On 4 March 2017 at 03:30, Boaz Bar Ilan<boazpro...@gmail.com>  wrote:
>>
>>>
>>> print composer sopped responsing
>>>
>>> what to do ?
>>>
>>>
>>
>> I'd start by learning some mailing list etiquette - this isn't a chat bot.
>>
>> Nyall
>> ___
>> Qgis-user mailing list
>> Qgis-user@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] status of bug report

[Nyall Dawson]

On 4 March 2017 at 05:35, Tobias Wendorff
 wrote:
> Hi there!
>
> About three weeks ago, I've posted a possible bug in the WMS client
> of QGIS. This bug hasn't been reviewed or assigned to a developer:
> https://hub.qgis.org/issues/16180
>
> Did I miss something to make or did I post it in the wrong section?

No - you haven't missed anything. It's just important to note that the
bug tracker is just a tracker - basically a giant list of issues.
Filing new ones won't automatically result in a fix.

A while ago I wrote up a blog post to try to clarify the situation
with how you can get things fixed in QGIS. It's just my opinion, and
not reflective of an official project view, but you can read it here:
http://nyalldawson.net/2016/08/how-to-effectively-get-things-changed-in-qgis/.

Hope that helps!
Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] print composer

[Nyall Dawson]

On 4 March 2017 at 03:30, Boaz Bar Ilan  wrote:
> print composer sopped responsing
>
> what to do ?
>

I'd start by learning some mailing list etiquette - this isn't a chat bot.

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] QGIS loaded millions of entries

[Nyall Dawson]

On 4 March 2017 at 05:39, Tobias Wendorff
 wrote:
> Hi there!
>
> I'm using QGIS 2.18.3 (Windows 7, 64bit with 16 GB of RAM).
>
> I've loaded a PostgreSQL table with more than 3 * 10^6 entries.
> When accesing the attributes, QGIS downloads all the attributes!
> QGIS first gets instable and crashes afterwards.
>
> Actually, it would be more useful to download only the attributes,
> which are in the current view.
>
> Is there an option for this or is it a bug/missing feature?

Yep - it's really annoying for large layers. If you go to settings ->
options -> data sources you can change "attribute table behaviour" to
"show features visible on map". That'll help a bit.

I've got changes incoming which also improve the performance if you
have the "show selected features" option enabled
(https://github.com/qgis/QGIS/pull/4164).

Nyall
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Question on Spatialite

[Nyall Dawson]

On 2 March 2017 at 08:33, Randal Hale <rjh...@northrivergeographic.com> wrote:
> Attribute! Used Field Calculator to split an attribute to get the zipcode
> into a standalone column. I then hit save and watched the fun.

Can you try doing this outside of the attribute table? Ie just click
the main window field calculator toolbar button and see if the
performance is the same?

Nyall

>
> And yes that seems to be it. Good, I'm not crazy (on this issue).
>
> Randy
>
>
>
> On 03/01/2017 05:23 PM, Nyall Dawson wrote:
>>
>> On 2 March 2017 at 08:08, Randal Hale <rjh...@northrivergeographic.com>
>> wrote:
>>>
>>> I'm using 2.18.4
>>>
>>> I'm working with Address point data. My typical process is to import the
>>> data into PostGIS/Postgresql and work with it. I decided to work with
>>> spatialite and do some data cleanup.
>>>
>>> I have 222519 points. I made a change to all of them and hit save.
>>
>> What change did you make? Attribute or geometry? How was this change made?
>>
>> It possibly relates to this ongoing work:
>> https://github.com/qgis/QGIS/pull/4175
>>
>>
>> Nyall
>>
>>> So for
>>> the next 20 minutes I watched a very small journal file appear and
>>> disappear
>>> and QGIS just locked up until it was done. QGIS returned to normal and
>>> I'm
>>> working along again.
>>>
>>> So my question is - Can I avoid this happening easily or is this just a
>>> "it's what happens with spatialite" problem? Am I missing a setting
>>> somewhere?
>>>
>>> Randy
>>>
>>>
>>> --
>>> 
>>> Randal Hale
>>> North River Geographic Systems, Inc
>>> http://www.northrivergeographic.com
>>> 423.653.3611 rjh...@northrivergeographic.com
>>> twitter:rjhale http://about.me/rjhale
>>> http://www.northrivergeographic.com/introduction-to-quantum-gis
>>> https://www.facebook.com/NRGSInc
>>>
>>> ___
>>> Qgis-user mailing list
>>> Qgis-user@lists.osgeo.org
>>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
>
> --
> 
> Randal Hale
> North River Geographic Systems, Inc
> http://www.northrivergeographic.com
> 423.653.3611 rjh...@northrivergeographic.com
> twitter:rjhale http://about.me/rjhale
> http://www.northrivergeographic.com/introduction-to-quantum-gis
> https://www.facebook.com/NRGSInc
>
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Question on Spatialite

[Nyall Dawson]

On 2 March 2017 at 08:08, Randal Hale  wrote:
> I'm using 2.18.4
>
> I'm working with Address point data. My typical process is to import the
> data into PostGIS/Postgresql and work with it. I decided to work with
> spatialite and do some data cleanup.
>
> I have 222519 points. I made a change to all of them and hit save.

What change did you make? Attribute or geometry? How was this change made?

It possibly relates to this ongoing work: https://github.com/qgis/QGIS/pull/4175


Nyall

> So for
> the next 20 minutes I watched a very small journal file appear and disappear
> and QGIS just locked up until it was done. QGIS returned to normal and I'm
> working along again.
>
> So my question is - Can I avoid this happening easily or is this just a
> "it's what happens with spatialite" problem? Am I missing a setting
> somewhere?
>
> Randy
>
>
> --
> 
> Randal Hale
> North River Geographic Systems, Inc
> http://www.northrivergeographic.com
> 423.653.3611 rjh...@northrivergeographic.com
> twitter:rjhale http://about.me/rjhale
> http://www.northrivergeographic.com/introduction-to-quantum-gis
> https://www.facebook.com/NRGSInc
>
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user