Re: [QGIS-Developer] Bug #21460?

[Paolo Cavallini]

Hi all,

On 09/04/19 01:19, Nyall Dawson wrote:
> On Tue, 9 Apr 2019 at 09:03, Cory Albrecht  wrote:
>>
>> Hi Paolo,
>>
>> Might I suggest that you consider that not everybody has time to go an learn 
>> the codebase of QGIS to be able to know where to start for making changes 
>> themselves, even if they happen to know C++? Especially if they have to 
>> learn the Qt framework at the same time.
>>
>> That consideration might be better than a snarky reply to a polite inquiry. 
>> Just saying.
> 
> I don't believe Paolo's reply was ever intended to be snarky. Instead,
> he's just trying to inform you of how development occurs in
> open-source projects like QGIS in order to empower you to be able to
> get this bug fixed.
sorry Cory, I did not mean to be impolite. I think the page cited
depicts accurately the sheer reality behind bug fixing, and effectively
answers your question.
A Peter pointed out, thanks to our sponsors and donors we are able to
fix many bugs before each release, but there cannot be a guarantee. The
best option is always come and collaborate with core team, in the way
that is most practical to you.
Cheers.
-- 
Paolo Cavallini - www.faunalia.eu
QGIS.ORG Chair:
http://planet.qgis.org/planet/user/28/tag/qgis%20board/
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] Bug #21460?

[Peter Petrik]

Hi,

we have bug-fix period before each release which is funded by QGIS.org from
donations. So usually bugs that are critical/regression/crash/corrupt data
are fixed during that period for LTR and PR.

Cheers,
Peter

On Tue, Apr 9, 2019 at 3:41 AM Nyall Dawson  wrote:

> On Tue, 9 Apr 2019 at 10:52, Cory Albrecht  wrote:
> >
> > Hello Nyall,
> >
> > Oh, I am well familiar with the vicissitudes of getting work done in
> open source projects - It's why I have donated to many opensource projects
> over the decades that I have been using them, including QGIS. But just
> telling people "go fix it yourself" is not helpful. To be frank, it makes
> me wonder if I should have made that donation elsewhere rather than making
> me willing to contribute to the project by trying to fix a bug or two
> instead. How do I know that I wouldn't just get another snarky brush off
> were I to ask for help because my lack of familiarity with the QGIS
> codebase and the Qt framework?
>
> Hi Cory,
>
> The message of that blog post is definitely not intended to be that
> the answer is "go fix it yourself". It's that fixing it yourself is
> ONE option, and there's a number of other reliable avenues to follow
> too.
>
> If you can suggest any ways to improve the wording of that post (while
> keep the core message intact), please let me know.
>
> Nyall
>
>
> >
> >
> > On Mon, Apr 8, 2019 at 7:19 PM Nyall Dawson 
> wrote:
> >>
> >> On Tue, 9 Apr 2019 at 09:03, Cory Albrecht 
> wrote:
> >> >
> >> > Hi Paolo,
> >> >
> >> > Might I suggest that you consider that not everybody has time to go
> an learn the codebase of QGIS to be able to know where to start for making
> changes themselves, even if they happen to know C++? Especially if they
> have to learn the Qt framework at the same time.
> >> >
> >> > That consideration might be better than a snarky reply to a polite
> inquiry. Just saying.
> >>
> >> I don't believe Paolo's reply was ever intended to be snarky. Instead,
> >> he's just trying to inform you of how development occurs in
> >> open-source projects like QGIS in order to empower you to be able to
> >> get this bug fixed.
> >>
> >> Nyall
> >>
> >>
> >>
> >> >
> >> > On Mon, Apr 8, 2019 at 4:30 AM Paolo Cavallini 
> wrote:
> >> >>
> >> >> Hi Cory,
> >> >>
> >> >> On 08/04/19 00:57, Cory Albrecht wrote:
> >> >> > I was wondering when Bug #21460
> >> >> >  was going to get some
> attention?
> >> >> > It's about  QGIS not forming proper PostGIS SQL updates correctly,
> >> >> > specifically for null values.
> >> >>
> >> >> may I suggest you an interesting read?
> >> >>
> https://nyalldawson.net/2016/08/how-to-effectively-get-things-changed-in-qgis/
> >> >> Cheers.
> >> >> --
> >> >> Paolo Cavallini - www.faunalia.eu
> >> >> QGIS.ORG Chair:
> >> >> http://planet.qgis.org/planet/user/28/tag/qgis%20board/
> >> >> ___
> >> >> QGIS-Developer mailing list
> >> >> QGIS-Developer@lists.osgeo.org
> >> >> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> >> >> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> >> >
> >> > ___
> >> > QGIS-Developer mailing list
> >> > QGIS-Developer@lists.osgeo.org
> >> > List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> >> > Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> >
> > ___
> > QGIS-Developer mailing list
> > QGIS-Developer@lists.osgeo.org
> > List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> > Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] Why was selection tool behaviour changed in 3.x?

[Jo]

I guess the rationale was making it easier on the tendons in the carpal
tunnel. Click, hold/move, click became click, move, click.

I would like to see a mode where the node that will be moved is
autoselected based on proximity to the mouse pointer. Then it would become
move, click, move, click. Obviously this needs to be guided by a
rubberband, whowing which node will be moved.

In JOSM this "improve way accuracy" also allows Ctrl-Click for adding
nodes, I mean vertices.

Polyglot

On Tue, Apr 9, 2019 at 2:54 AM Cory Albrecht  wrote:

>
> Cory Albrecht 
> 8:14 PM (37 minutes ago)
> to Régis
> Hello Régis,
>
> Sorry for not being clear - I mean when using the selection tool in
> freehand mode. I am definitely not talking about the identification tool,
> assuming you're referring to the same thing that I am thinking of?
> Ctrl+Shift+I, or the icon that is the cursor with a the letter "i" in a
> sold blue circle. I'm not sure I would call that new as it's been part of
> QGIS since I started using it in about 2015. Perhaps you're an old salt
> from the 1.x days? ;-)
>
> As a principle of UX design, ideally, the user should do the same action -
> click and drag - for any type of selection, both to maintain internal
> consistency in the application and with common ways of doing things in the
> broader computer universe. This lets people learn new software quickly by
> having a set of transferable actions that can get them up and running and
> doing rudimentary things quickly. It also helps reduce unintended errors
> caused by using common actions that get unexpectedly interpreted different
> than the user is used to. Things like this contribute to how easy or
> frustrating an application is to use, both for new and long time users.
>
> 1. For the "Select Feature(s)", click and drag to indicate the diagonally
> opposite corners of a selection rectangle.
> 2. For the "Select Features by Freehand", click and drag to create an
> irregular blob of selection area.
> 3. For the "Select Features by Radius", click and drag to indicate the
> centre of a selection circle and it's radius.
>
> In 2.x the answer to all of those was yes, but in 3.x it's yes, no, no.
>
> In vector and raster drawing applications, drawing rectangles, circles and
> blobs is done by click and drag, as is selecting rectangular, circular or
> irregular blobby areas. If you release and click elsewhere then drag, you
> start drawing a new object, or you discard the first selection and start
> outlining a new one. Word processing and text section, video editors and
> frame selection, sound editors and lengths of time in a track, they all
> have the user do these conceptually similar tasks in the same way - click
> and drag to create a selection , new click discards old selection.
>
> Another principle of UX design is that you don't change how a user does
> something unless there is clear benefit that outweighs the trouble of
> relearning, especially for action concepts that are common in the broader
> sphere. When you make changes without benefits you cause friction in your
> user flows (some call those "point points"), and that means people find
> that task (and potentially the application as a whole) difficult and
> frustrating to use.
>
> For those three methods of selection there's nothing to be gained by
> making QGIS 3.x the odd one out in how this is done. There's no benefit
> added by extra functionality in these selection methods. All it does is
> create pain points, both by being different from everybody else and by
> being inconsistent internally.
>
> The exception to this is the poly gone selection tool. I've never
> encountered it outside of QGIS and ArcGIS. Drawing applications have
> polygon drawing tools in which you sequentially click the polygon's points,
> just like how you create features on polygon or line layers in QGIS, but
> there's no polygon selection analogue. As such it makes sense to take the
> feature creation method of sequential clicks over for use in a polygon
> selection tool rather than coming up with a whole new user flow like click
> and drag and tapping the space bar for the points.
>
> And so I wonder - what was the rationale behind making this change?
>
> On Mon, Apr 8, 2019 at 6:00 AM Régis Haubourg 
> wrote:
>
>> Hi Cory,
>> I must say I didn't notice any difference on the selection tool behavior
>> on my side.
>> I don't think there was any explicit attempt to homogenize the selection
>> behavior with the node tool new ergonomy.
>>
>> Just a check, in the maptool dropdown list for selection tool, are your
>> using the freehand selection tool or the classical clic and drag selection
>> tool?
>>
>> I've seen similar surprising issues with the new "identify" tool that now
>> can interrogate features in a polygon. Users got confused when they changed
>> this behavior by mistake. Could that be your case?
>> Cheers
>> Régis
>>
>> Le lun. 8 avr. 2019 à 01:09, Cory Albrecht  a
>> écrit :
>>
>>> I was wondering why 

Re: [QGIS-Developer] Bug #21460?

[Nyall Dawson]

On Tue, 9 Apr 2019 at 10:52, Cory Albrecht  wrote:
>
> Hello Nyall,
>
> Oh, I am well familiar with the vicissitudes of getting work done in open 
> source projects - It's why I have donated to many opensource projects over 
> the decades that I have been using them, including QGIS. But just telling 
> people "go fix it yourself" is not helpful. To be frank, it makes me wonder 
> if I should have made that donation elsewhere rather than making me willing 
> to contribute to the project by trying to fix a bug or two instead. How do I 
> know that I wouldn't just get another snarky brush off were I to ask for help 
> because my lack of familiarity with the QGIS codebase and the Qt framework?

Hi Cory,

The message of that blog post is definitely not intended to be that
the answer is "go fix it yourself". It's that fixing it yourself is
ONE option, and there's a number of other reliable avenues to follow
too.

If you can suggest any ways to improve the wording of that post (while
keep the core message intact), please let me know.

Nyall


>
>
> On Mon, Apr 8, 2019 at 7:19 PM Nyall Dawson  wrote:
>>
>> On Tue, 9 Apr 2019 at 09:03, Cory Albrecht  wrote:
>> >
>> > Hi Paolo,
>> >
>> > Might I suggest that you consider that not everybody has time to go an 
>> > learn the codebase of QGIS to be able to know where to start for making 
>> > changes themselves, even if they happen to know C++? Especially if they 
>> > have to learn the Qt framework at the same time.
>> >
>> > That consideration might be better than a snarky reply to a polite 
>> > inquiry. Just saying.
>>
>> I don't believe Paolo's reply was ever intended to be snarky. Instead,
>> he's just trying to inform you of how development occurs in
>> open-source projects like QGIS in order to empower you to be able to
>> get this bug fixed.
>>
>> Nyall
>>
>>
>>
>> >
>> > On Mon, Apr 8, 2019 at 4:30 AM Paolo Cavallini  
>> > wrote:
>> >>
>> >> Hi Cory,
>> >>
>> >> On 08/04/19 00:57, Cory Albrecht wrote:
>> >> > I was wondering when Bug #21460
>> >> >  was going to get some attention?
>> >> > It's about  QGIS not forming proper PostGIS SQL updates correctly,
>> >> > specifically for null values.
>> >>
>> >> may I suggest you an interesting read?
>> >> https://nyalldawson.net/2016/08/how-to-effectively-get-things-changed-in-qgis/
>> >> Cheers.
>> >> --
>> >> Paolo Cavallini - www.faunalia.eu
>> >> QGIS.ORG Chair:
>> >> http://planet.qgis.org/planet/user/28/tag/qgis%20board/
>> >> ___
>> >> QGIS-Developer mailing list
>> >> QGIS-Developer@lists.osgeo.org
>> >> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>> >> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>> >
>> > ___
>> > QGIS-Developer mailing list
>> > QGIS-Developer@lists.osgeo.org
>> > List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>> > Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] Why was selection tool behaviour changed in 3.x?

[Cory Albrecht]

Cory Albrecht 
8:14 PM (37 minutes ago)
to Régis
Hello Régis,

Sorry for not being clear - I mean when using the selection tool in
freehand mode. I am definitely not talking about the identification tool,
assuming you're referring to the same thing that I am thinking of?
Ctrl+Shift+I, or the icon that is the cursor with a the letter "i" in a
sold blue circle. I'm not sure I would call that new as it's been part of
QGIS since I started using it in about 2015. Perhaps you're an old salt
from the 1.x days? ;-)

As a principle of UX design, ideally, the user should do the same action -
click and drag - for any type of selection, both to maintain internal
consistency in the application and with common ways of doing things in the
broader computer universe. This lets people learn new software quickly by
having a set of transferable actions that can get them up and running and
doing rudimentary things quickly. It also helps reduce unintended errors
caused by using common actions that get unexpectedly interpreted different
than the user is used to. Things like this contribute to how easy or
frustrating an application is to use, both for new and long time users.

1. For the "Select Feature(s)", click and drag to indicate the diagonally
opposite corners of a selection rectangle.
2. For the "Select Features by Freehand", click and drag to create an
irregular blob of selection area.
3. For the "Select Features by Radius", click and drag to indicate the
centre of a selection circle and it's radius.

In 2.x the answer to all of those was yes, but in 3.x it's yes, no, no.

In vector and raster drawing applications, drawing rectangles, circles and
blobs is done by click and drag, as is selecting rectangular, circular or
irregular blobby areas. If you release and click elsewhere then drag, you
start drawing a new object, or you discard the first selection and start
outlining a new one. Word processing and text section, video editors and
frame selection, sound editors and lengths of time in a track, they all
have the user do these conceptually similar tasks in the same way - click
and drag to create a selection , new click discards old selection.

Another principle of UX design is that you don't change how a user does
something unless there is clear benefit that outweighs the trouble of
relearning, especially for action concepts that are common in the broader
sphere. When you make changes without benefits you cause friction in your
user flows (some call those "point points"), and that means people find
that task (and potentially the application as a whole) difficult and
frustrating to use.

For those three methods of selection there's nothing to be gained by making
QGIS 3.x the odd one out in how this is done. There's no benefit added by
extra functionality in these selection methods. All it does is create pain
points, both by being different from everybody else and by being
inconsistent internally.

The exception to this is the poly gone selection tool. I've never
encountered it outside of QGIS and ArcGIS. Drawing applications have
polygon drawing tools in which you sequentially click the polygon's points,
just like how you create features on polygon or line layers in QGIS, but
there's no polygon selection analogue. As such it makes sense to take the
feature creation method of sequential clicks over for use in a polygon
selection tool rather than coming up with a whole new user flow like click
and drag and tapping the space bar for the points.

And so I wonder - what was the rationale behind making this change?

On Mon, Apr 8, 2019 at 6:00 AM Régis Haubourg 
wrote:

> Hi Cory,
> I must say I didn't notice any difference on the selection tool behavior
> on my side.
> I don't think there was any explicit attempt to homogenize the selection
> behavior with the node tool new ergonomy.
>
> Just a check, in the maptool dropdown list for selection tool, are your
> using the freehand selection tool or the classical clic and drag selection
> tool?
>
> I've seen similar surprising issues with the new "identify" tool that now
> can interrogate features in a polygon. Users got confused when they changed
> this behavior by mistake. Could that be your case?
> Cheers
> Régis
>
> Le lun. 8 avr. 2019 à 01:09, Cory Albrecht  a
> écrit :
>
>> I was wondering why the selection tool behaviour in 3.x was changed from
>> the implementation in 2.18?
>>
>> In 2.18.x when you wanted to select features in a layer, you clicked the
>> primary mouse button, held it, and moves the mouse cursor over the items
>> you wanted to select - known as "click and drag". To help, a shape was
>> drawn on screen for the user to know what they had already dragged the
>> mouse over top of. To add to the selection you used shift plus click and
>> drag, to remove, Ctrl plus click and drag. It the way select tools work
>> broadly across computer world and is intuitive because of it's ubiquity -
>> learn it once, use it everywhere.
>>
>> In 3.x, however, instead of using 

Re: [QGIS-Developer] Bug #21460?

[Cory Albrecht]

Hello Nyall,

Oh, I am well familiar with the vicissitudes of getting work done in open
source projects - It's why I have donated to many opensource projects over
the decades that I have been using them, including QGIS. But just telling
people "go fix it yourself" is not helpful. To be frank, it makes me wonder
if I should have made that donation elsewhere rather than making me willing
to contribute to the project by trying to fix a bug or two instead. How do
I know that I wouldn't just get another snarky brush off were I to ask for
help because my lack of familiarity with the QGIS codebase and the Qt
framework?


On Mon, Apr 8, 2019 at 7:19 PM Nyall Dawson  wrote:

> On Tue, 9 Apr 2019 at 09:03, Cory Albrecht  wrote:
> >
> > Hi Paolo,
> >
> > Might I suggest that you consider that not everybody has time to go an
> learn the codebase of QGIS to be able to know where to start for making
> changes themselves, even if they happen to know C++? Especially if they
> have to learn the Qt framework at the same time.
> >
> > That consideration might be better than a snarky reply to a polite
> inquiry. Just saying.
>
> I don't believe Paolo's reply was ever intended to be snarky. Instead,
> he's just trying to inform you of how development occurs in
> open-source projects like QGIS in order to empower you to be able to
> get this bug fixed.
>
> Nyall
>
>
>
> >
> > On Mon, Apr 8, 2019 at 4:30 AM Paolo Cavallini 
> wrote:
> >>
> >> Hi Cory,
> >>
> >> On 08/04/19 00:57, Cory Albrecht wrote:
> >> > I was wondering when Bug #21460
> >> >  was going to get some
> attention?
> >> > It's about  QGIS not forming proper PostGIS SQL updates correctly,
> >> > specifically for null values.
> >>
> >> may I suggest you an interesting read?
> >>
> https://nyalldawson.net/2016/08/how-to-effectively-get-things-changed-in-qgis/
> >> Cheers.
> >> --
> >> Paolo Cavallini - www.faunalia.eu
> >> QGIS.ORG Chair:
> >> http://planet.qgis.org/planet/user/28/tag/qgis%20board/
> >> ___
> >> QGIS-Developer mailing list
> >> QGIS-Developer@lists.osgeo.org
> >> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> >> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> >
> > ___
> > QGIS-Developer mailing list
> > QGIS-Developer@lists.osgeo.org
> > List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> > Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] No LTR on SUSE repos

[Nyall Dawson]

On Mon, 8 Apr 2019 at 18:35, Alessandro Pasotti  wrote:
>
> On Mon, Apr 8, 2019 at 10:25 AM Paolo Cavallini  wrote:
>>
>> Hi Yann,
>>
>> On 08/04/19 10:01, Yann POUFFARIX wrote:
>> > Back in time.
>> >
>> > No answer from maintener and of course nothing listed here
>> >
>> > https://software.opensuse.org/package/qgis-ltr
>> >
>> > what to do x)
>>
>> I think you are left with the choice of either doing the package
>> yourself, or hire someone to do it. Of course you can always compile
>> qgis yourself, it shouldn't be too hard if the dependencies are properly
>> packaged.
>> Cheers.
>>
>
> Just curious: why should we invest money for Apple packaging and not for SuSE 
> packaging?
>
> I think we should be inclusive and supportive for all users on all platforms.

Because SuSE has 4 users, yet MacOS has 27*

Nyall

* Not scientific numbers ;)
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] Any plan for alternates to Python 3.6 dependency on macOS QGIS 3?

[Nyall Dawson]

On Tue, 9 Apr 2019 at 03:59, Anthony DeBarros  wrote:
>
> Hi, folks,
>
> Longtime QGIS user but recent listserv subscriber here with a question about 
> macOS installs of QGIS 3.x. I could not find this question in recent 
> archives, so here goes:
>
> When I installed QGIS 3.x recently on macOS Mojave, the required Python 3.6 
> install from Python.org severely mucked up my existing Homebrew Python 3. 
> I've discussed similar issues with colleagues who've experienced the same.
>
> I'm curious whether it's possible to direct QGIS 3 to reference an existing 
> Python 3 install other than the one from Python.org? Would this be possibly 
> considered as a new feature, if not?
>
> Beyond that, if anyone has found a good guide to managing Homebrew Python 
> alongside the Python.org install, I'd appreciate that.
>
> I've been told that installing QGIS 3 via Homebrew is also an alternative, 
> but I haven't tried and would appreciate thoughts there.

Have you tried using the Lutra all-in-one installers?
https://lutraconsulting.github.io/qgis-mac-packager/

Nyall
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] Bug #21460?

[Nyall Dawson]

On Tue, 9 Apr 2019 at 09:03, Cory Albrecht  wrote:
>
> Hi Paolo,
>
> Might I suggest that you consider that not everybody has time to go an learn 
> the codebase of QGIS to be able to know where to start for making changes 
> themselves, even if they happen to know C++? Especially if they have to learn 
> the Qt framework at the same time.
>
> That consideration might be better than a snarky reply to a polite inquiry. 
> Just saying.

I don't believe Paolo's reply was ever intended to be snarky. Instead,
he's just trying to inform you of how development occurs in
open-source projects like QGIS in order to empower you to be able to
get this bug fixed.

Nyall



>
> On Mon, Apr 8, 2019 at 4:30 AM Paolo Cavallini  wrote:
>>
>> Hi Cory,
>>
>> On 08/04/19 00:57, Cory Albrecht wrote:
>> > I was wondering when Bug #21460
>> >  was going to get some attention?
>> > It's about  QGIS not forming proper PostGIS SQL updates correctly,
>> > specifically for null values.
>>
>> may I suggest you an interesting read?
>> https://nyalldawson.net/2016/08/how-to-effectively-get-things-changed-in-qgis/
>> Cheers.
>> --
>> Paolo Cavallini - www.faunalia.eu
>> QGIS.ORG Chair:
>> http://planet.qgis.org/planet/user/28/tag/qgis%20board/
>> ___
>> QGIS-Developer mailing list
>> QGIS-Developer@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] Bug #21460?

[Cory Albrecht]

Hi Paolo,

Might I suggest that you consider that not everybody has time to go an
learn the codebase of QGIS to be able to know where to start for making
changes themselves, even if they happen to know C++? Especially if they
have to learn the Qt framework at the same time.

That consideration might be better than a snarky reply to a polite inquiry.
Just saying.

On Mon, Apr 8, 2019 at 4:30 AM Paolo Cavallini 
wrote:

> Hi Cory,
>
> On 08/04/19 00:57, Cory Albrecht wrote:
> > I was wondering when Bug #21460
> >  was going to get some attention?
> > It's about  QGIS not forming proper PostGIS SQL updates correctly,
> > specifically for null values.
>
> may I suggest you an interesting read?
>
> https://nyalldawson.net/2016/08/how-to-effectively-get-things-changed-in-qgis/
> Cheers.
> --
> Paolo Cavallini - www.faunalia.eu
> QGIS.ORG Chair:
> http://planet.qgis.org/planet/user/28/tag/qgis%20board/
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

[QGIS-Developer] Any plan for alternates to Python 3.6 dependency on macOS QGIS 3?

[Anthony DeBarros]

Hi, folks,

Longtime QGIS user but recent listserv subscriber here with a question
about macOS installs of QGIS 3.x. I could not find this question in recent
archives, so here goes:

When I installed QGIS 3.x recently on macOS Mojave, the required Python 3.6
install from Python.org severely mucked up my existing Homebrew Python 3.
I've discussed similar issues with colleagues who've experienced the same.

I'm curious whether it's possible to direct QGIS 3 to reference an existing
Python 3 install other than the one from Python.org? Would this be possibly
considered as a new feature, if not?

Beyond that, if anyone has found a good guide to managing Homebrew Python
alongside the Python.org install, I'd appreciate that.

I've been told that installing QGIS 3 via Homebrew is also an alternative,
but I haven't tried and would appreciate thoughts there.

Thanks,
Anthony DeBarros
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

[QGIS-Developer] Plugin [1391] Asistente LADM_COL approval notification.

[noreply]

Plugin Asistente LADM_COL approval by pcav.
The plugin version "[1391] Asistente LADM_COL 1.6.1" is now approved
Link: http://plugins.qgis.org/plugins/asistente_ladm_col/
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] [LTR 3.4.6] Background online services not working (OpenLayers & QuickMapServices)

[Matthias Kuhn]

Hi Niccolò,

Have a look at this gis.stackexchange.com answer for a QGIS only, no
plugin possibility. This also offers maximum compatibility with
printing, rotating and QField.

https://gis.stackexchange.com/questions/20191/adding-basemaps-from-google-or-bing-in-qgis/217670#217670

Best regards

Matthias

https://gis.stackexchange.com/questions/20191/adding-basemaps-from-google-or-bing-in-qgis/217670#217670

On 4/8/19 5:43 PM, Niccolò Marchi wrote:
>
> Hi all,
>
> I need to add an online service (Bing/google aerial) as a background
> map in my project.
>
>  
>
> Using those offered in the OpenLayers plugin I get the following error
> “Printing and rotating of Javascript API based layer is currently not
> supported” and, therefore I’m not able to keep any of these layers
> saved in my project.
>
>  
>
> Using QuickMapServices, instead, I’m not able  to download the
> “Contributed pack” because I get this error: “Attribute error: module
> ‘sys’ has no attribute ‘exec_type’ “.
>
>  
>
>  
>
> It is a pain for normal use but this has quite bad consequences when I
> decide to use that project on Qfield. Am I missing something? Any
> other suggestion?
>
> I’m using LTR 3.4.6 on W10.
>
>  
>
> Sorry for the re-post, but I got no reply on users-list.
>
>  
>
> Many thanks in advance for your help
>
>  
>
> Nic
>
>
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
-- 
Matthias Kuhn
matth...@opengis.ch 
+41 (0)76 435 67 63 
OPENGIS.ch Logo 
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

[QGIS-Developer] [LTR 3.4.6] Background online services not working (OpenLayers & QuickMapServices)

[Niccolò Marchi]

Hi all,
I need to add an online service (Bing/google aerial) as a background map in my 
project.

Using those offered in the OpenLayers plugin I get the following error 
“Printing and rotating of Javascript API based layer is currently not 
supported” and, therefore I’m not able to keep any of these layers saved in my 
project.

Using QuickMapServices, instead, I’m not able  to download the “Contributed 
pack” because I get this error: “Attribute error: module ‘sys’ has no attribute 
‘exec_type’ “.


It is a pain for normal use but this has quite bad consequences when I decide 
to use that project on Qfield. Am I missing something? Any other suggestion?
I’m using LTR 3.4.6 on W10.

Sorry for the re-post, but I got no reply on users-list.

Many thanks in advance for your help

Nic
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] No LTR on SUSE repos

[Régis Haubourg]

Hi Yann
We are tallking about going towards official docker for desktop use here :
https://github.com/qgis/QGIS-Enhancement-Proposals/issues/142

It might be an option, I have all my QGIS versions running with this option
on Ubuntu, just because running different versions at the same time is a
pain in Linux. We do this frequently for production uses with QGIS server,
and it seems to be working with Desktop, though we still need to test and
document it thoroughly. Of course other container options are here (flapak,
snap, conda) and I just can't give any advice on these other options as I
don't use them.
The other option is to support (in the long term) someone to package QGIS
for Suse.

Cheers Régis

Le lun. 8 avr. 2019 à 17:22, Yann POUFFARIX  a
écrit :

> For now, I'm actually trying to do it myself... for up to 200+ users in my
> company.
>
> A valid rpm package was available on repositories and suddenly disapear
> and i've found no explication.
>
> Some dependencies required by Qgis were provided by Qgis repository and...
> disapear.
>
> There is no objection to dev or anything we can do to improve Qgis but
> there is a point that we can't count, the fluctuant repository to install
> packages and that mean if we finnaly build a proper rpm , we'll distribute
> it by our on way because we can't predict what apend in the future.
> Le lun. 8 avr. 2019 à 10:35, Alessandro Pasotti  a
> écrit :
>
>> On Mon, Apr 8, 2019 at 10:25 AM Paolo Cavallini 
>> wrote:
>>
>>> Hi Yann,
>>>
>>> On 08/04/19 10:01, Yann POUFFARIX wrote:
>>> > Back in time.
>>> >
>>> > No answer from maintener and of course nothing listed here
>>> >
>>> > https://software.opensuse.org/package/qgis-ltr
>>> >
>>> > what to do x)
>>>
>>> I think you are left with the choice of either doing the package
>>> yourself, or hire someone to do it. Of course you can always compile
>>> qgis yourself, it shouldn't be too hard if the dependencies are properly
>>> packaged.
>>> Cheers.
>>>
>>>
>> Just curious: why should we invest money for Apple packaging and not for
>> SuSE packaging?
>>
>> I think we should be inclusive and supportive for all users on all
>> platforms.
>>
>>
>>
>> --
>>> Paolo Cavallini - www.faunalia.eu
>>> QGIS.ORG Chair:
>>> http://planet.qgis.org/planet/user/28/tag/qgis%20board/
>>> ___
>>> QGIS-Developer mailing list
>>> QGIS-Developer@lists.osgeo.org
>>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>>
>>
>>
>> --
>> Alessandro Pasotti
>> w3:   www.itopen.it
>> ___
>> QGIS-Developer mailing list
>> QGIS-Developer@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] No LTR on SUSE repos

[Yann POUFFARIX]

For now, I'm actually trying to do it myself... for up to 200+ users in my
company.

A valid rpm package was available on repositories and suddenly disapear and
i've found no explication.

Some dependencies required by Qgis were provided by Qgis repository and...
disapear.

There is no objection to dev or anything we can do to improve Qgis but
there is a point that we can't count, the fluctuant repository to install
packages and that mean if we finnaly build a proper rpm , we'll distribute
it by our on way because we can't predict what apend in the future.
Le lun. 8 avr. 2019 à 10:35, Alessandro Pasotti  a
écrit :

> On Mon, Apr 8, 2019 at 10:25 AM Paolo Cavallini 
> wrote:
>
>> Hi Yann,
>>
>> On 08/04/19 10:01, Yann POUFFARIX wrote:
>> > Back in time.
>> >
>> > No answer from maintener and of course nothing listed here
>> >
>> > https://software.opensuse.org/package/qgis-ltr
>> >
>> > what to do x)
>>
>> I think you are left with the choice of either doing the package
>> yourself, or hire someone to do it. Of course you can always compile
>> qgis yourself, it shouldn't be too hard if the dependencies are properly
>> packaged.
>> Cheers.
>>
>>
> Just curious: why should we invest money for Apple packaging and not for
> SuSE packaging?
>
> I think we should be inclusive and supportive for all users on all
> platforms.
>
>
>
> --
>> Paolo Cavallini - www.faunalia.eu
>> QGIS.ORG Chair:
>> http://planet.qgis.org/planet/user/28/tag/qgis%20board/
>> ___
>> QGIS-Developer mailing list
>> QGIS-Developer@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>
>
>
> --
> Alessandro Pasotti
> w3:   www.itopen.it
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

[QGIS-Developer] Plugin [1479] QWeather approval notification.

[noreply]

Plugin QWeather approval by pcav.
The plugin version "[1479] QWeather 3.1" is now approved
Link: http://plugins.qgis.org/plugins/QWeather/
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] rasters in FileGDB?

[Even Rouault]

On lundi 8 avril 2019 12:28:50 CEST Paolo Cavallini wrote:
> Hi all,
> is this limitation still valid?

Yes

> https://gis.stackexchange.com/questions/110533/open-raster-stored-in-esri-v1
> 0-0-geodatabase-in-qgis-v2-4 I guess we depend fully on GDAL capabilities,
> right?

Yes, there was some prototyping done a couple years ago that confirmed there 
was a high chance to be able to produce a GDAL raster driver to open 
FileGeodatabase raster (without using the ESRI SDK, that doesn't offer this 
capability). Mostly waiting for an interested party to develop/fund it.

Even

-- 
Spatialys - Geospatial professional services
http://www.spatialys.com
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

[QGIS-Developer] Djikstra in Qgis3 still not working properly

[Christoph Dohm]

Hello,

 

I wrote a while back a mail here and posted something on Stackexchange.

 

https://gis.stackexchange.com/questions/308244/dijkstra-finding-shortest-path-gives-back-a-tree-with-only-1-in-it

 

My main issue is, that in Qgis 2 there was a fairly simple way to perform a dijkstra analysis. I think you needed like 6 lines overall.

Migrating the same principle to Qgis 3 doesn't work, it always sees the path as not connected although it is. My guess is that the graph isn't created correctly, but I'm not an expert programmer.

 

I know by now, that there is a workaround using the processing algorithm, but that gets way more complicated and as i mentioned earlier I'm not an expert and that is above my skill-level.

 

I really would like to have a straightforward approach like there was in Qgis2, where you could create a graph from existing layers, define some properties and then perform the dijkstra on that graph.

 

Here is the code that I migrated from Qgis2 that doesn't work but that would be great if it could work that way again:

 


director = QgsVectorLayerDirector(layer, -1, '', '', '', 3)

strategy = QgsNetworkDistanceStrategy()

director.addStrategy(strategy)

builder = QgsGraphBuilder(crs) #crs = project.crs() which is in this case WGS84

tiedPoints = director.makeGraph(builder, [startPt, endPt])
graph = builder.graph()

startVertex = graph.findVertex(startPt)

(tree, cost) = QgsGraphAnalyzer.dijkstra(graph, startVertex,0)

 

Sincerely

Christoph Dohm

___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] rasters in FileGDB?

[Nyall Dawson]

On Mon, 8 Apr 2019 at 20:29, Paolo Cavallini  wrote:
>
> Hi all,
> is this limitation still valid?
> https://gis.stackexchange.com/questions/110533/open-raster-stored-in-esri-v10-0-geodatabase-in-qgis-v2-4
> I guess we depend fully on GDAL capabilities, right?

Yep. There's a tool which may be of use:

https://github.com/r-barnes/ArcRasterRescue


> Thanks.
> --
> Paolo Cavallini - www.faunalia.eu
> QGIS.ORG Chair:
> http://planet.qgis.org/planet/user/28/tag/qgis%20board/
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

[QGIS-Developer] rasters in FileGDB?

[Paolo Cavallini]

Hi all,
is this limitation still valid?
https://gis.stackexchange.com/questions/110533/open-raster-stored-in-esri-v10-0-geodatabase-in-qgis-v2-4
I guess we depend fully on GDAL capabilities, right?
Thanks.
-- 
Paolo Cavallini - www.faunalia.eu
QGIS.ORG Chair:
http://planet.qgis.org/planet/user/28/tag/qgis%20board/
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] Why was selection tool behaviour changed in 3.x?

[Régis Haubourg]

Hi Cory,
I must say I didn't notice any difference on the selection tool behavior on
my side.
I don't think there was any explicit attempt to homogenize the selection
behavior with the node tool new ergonomy.

Just a check, in the maptool dropdown list for selection tool, are your
using the freehand selection tool or the classical clic and drag selection
tool?

I've seen similar surprising issues with the new "identify" tool that now
can interrogate features in a polygon. Users got confused when they changed
this behavior by mistake. Could that be your case?
Cheers
Régis

Le lun. 8 avr. 2019 à 01:09, Cory Albrecht  a écrit :

> I was wondering why the selection tool behaviour in 3.x was changed from
> the implementation in 2.18?
>
> In 2.18.x when you wanted to select features in a layer, you clicked the
> primary mouse button, held it, and moves the mouse cursor over the items
> you wanted to select - known as "click and drag". To help, a shape was
> drawn on screen for the user to know what they had already dragged the
> mouse over top of. To add to the selection you used shift plus click and
> drag, to remove, Ctrl plus click and drag. It the way select tools work
> broadly across computer world and is intuitive because of it's ubiquity -
> learn it once, use it everywhere.
>
> In 3.x, however, instead of using that common method, it has changed to
> click and release and move the mouse around. This is a common UI method to
> set focus to an item for subsequent actions but still be able to move the
> mouse around without selecting or affecting any other items. I know things
> would work slightly different in QGIS because of having a distinct
> selection tool that one must activate, but this removes intuitiveness from
> the application and makes it more difficult to use without any
> corresponding gain in functionality.
>
> A similar change has also happened in the vertex editor where in 2.18.x
> single clicking on a vertex used to mean select, and you had to drag (click
> and hold) to move it. Now, if you click and release, it unexpectedly drags
> the vertex around as you move the mouse.
>
> QGIS having it's own, non-standard mouse actions for tasks that are common
> (select, copy, delete, etc…) across all types of data (text in a
> wordprocessor, frames in a movie editor, features in a map editor, etc…) is
> counter-intuitive and confusing, especially if those non-standard actions
> are already commonly used for other common user interface actions.
>
> It's almost like the QGIS development team has decided that Ctrl+V will
> now mean "Cut", Ctrl+X will mean "Copy", and to copy have to use Alt+F1 for
> "Paste". Extending common user interface actions for something in QGIS that
> has no exact parallel but is still conceptually similar to that common
> action, like how Ctrl+Alt+V means paste what was copied into the buffer
> into a brand new layer, that makes sense. But ignoring decades of common UI
> actions that are in the muscle memory of probably all users makes the
> programme frustrating and tedious to use as one has to constantly remind
> themselves that QGIS is different.
>
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

[QGIS-Developer] Plugin [1688] QLockLayers approval notification.

[noreply]

Plugin QLockLayers approval by pcav.
The plugin version "[1688] QLockLayers 0.1" is now approved
Link: http://plugins.qgis.org/plugins/QLocklayers/
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] Bug #21460? | is QGIS non-commercial?

[Nyall Dawson]

On Mon, 8 Apr 2019 at 19:22, Benjamin Ducke  wrote:
>
> Hi Paolo,
>
> On 08/04/2019 10:30, Paolo Cavallini wrote:
> > Hi Cory,
> >
> > On 08/04/19 00:57, Cory Albrecht wrote:
> >> I was wondering when Bug #21460
> >>  was going to get some attention?
> >> It's about  QGIS not forming proper PostGIS SQL updates correctly,
> >> specifically for null values.
> >
> > may I suggest you an interesting read?
> > https://nyalldawson.net/2016/08/how-to-effectively-get-things-changed-in-qgis/
>
> There is a (common) problem with that text
> and your reply that is increasingly giving
> my headaches when thinking how to advertise
> QGIS among colleagues:
>
> Quote (from the URL above):
>
> "unlike the commercial GIS offerings"
>
> This suggests that QGIS is a non-commercial
> project.
>
> But a few lines further down we read:
>
> "either organisations who offer commercial QGIS
> support"
>
> "if you hire a core QGIS developer to make your
> changes"

Thanks, fixed.

Nyall

>
> These are clearly commercial offerings/activities,
> and people are being advised to make use of them.
>
> To be clear: I don't think there is anything wrong
> with that. But I think that it is misleading to
> present QGIS as "non-commercial". It is open
> source and non-proprietary, in the sense that no
> single individual or institution owns the rights
> to it. But commerce is clearly part of QGIS now.
>
> As I see it, there are two "honest" options:
>
> (1) Stop calling QGIS non-commercial and keep
> advising people to hire developers for fixing
> bugs and implementing features.
>
> (2) Keep calling QGIS a non-commercial project
> and stop advising people to hire developers
> to fix bugs or develop features for them.
>
> Best,
>
> Ben
>
>
> > Cheers.
> >
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

[QGIS-Developer] Crashes on closing QGIS

[Andreas Neumann]

Hi, 


We always have crashes on closing QGIS (on Windows) - regardless of
version - it happens on 3.4, 3.6 and master. 


Seems to be related to the authentication manager. Is this a known
issue? Are there ideas how to fix the issue? 

Here is the crashlog: 


Crash ID: 0f25782d491a42462ec5c96af91c45b3e9859bd3

Stack Trace

QMutex::lock :
QMutexLocker::QMutexLocker :
::operator() qgsauthmanager.cpp:145
QtPrivate::FunctorCall,QtPrivate::List,void, >::call
qobjectdefs_impl.h:128
QtPrivate::Functor,0>::call,void> qobjectdefs_impl.h:239
QtPrivate::QFunctorSlotObject,0,QtPrivate::List,void>::impl
qobjectdefs_impl.h:427
QMetaObject::activate :
QThread::finished :
QThread::currentThreadId :
QThread::start :
BaseThreadInitThunk :
RtlUserThreadStart :

QGIS Info
QGIS Version: 3.7.0-Master
QGIS code revision: 5a85e206f6
Compiled against Qt: 5.11.2
Running against Qt: 5.11.2
Compiled against GDAL: 2.4.1
Running against GDAL: 2.4.1

System Info
CPU Type: x86_64
Kernel Type: winnt
Kernel Version: 6.1.7601___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] Bug #21460? | is QGIS non-commercial?

[Benjamin Ducke]

Hi Paolo,

On 08/04/2019 10:30, Paolo Cavallini wrote:
> Hi Cory,
> 
> On 08/04/19 00:57, Cory Albrecht wrote:
>> I was wondering when Bug #21460
>>  was going to get some attention?
>> It's about  QGIS not forming proper PostGIS SQL updates correctly,
>> specifically for null values.
> 
> may I suggest you an interesting read?
> https://nyalldawson.net/2016/08/how-to-effectively-get-things-changed-in-qgis/

There is a (common) problem with that text
and your reply that is increasingly giving
my headaches when thinking how to advertise
QGIS among colleagues:

Quote (from the URL above):

"unlike the commercial GIS offerings"

This suggests that QGIS is a non-commercial
project.

But a few lines further down we read:

"either organisations who offer commercial QGIS
support"

"if you hire a core QGIS developer to make your
changes"

These are clearly commercial offerings/activities,
and people are being advised to make use of them.

To be clear: I don't think there is anything wrong
with that. But I think that it is misleading to
present QGIS as "non-commercial". It is open
source and non-proprietary, in the sense that no
single individual or institution owns the rights
to it. But commerce is clearly part of QGIS now.

As I see it, there are two "honest" options:

(1) Stop calling QGIS non-commercial and keep
advising people to hire developers for fixing
bugs and implementing features.

(2) Keep calling QGIS a non-commercial project
and stop advising people to hire developers
to fix bugs or develop features for them.

Best,

Ben


> Cheers.
> 
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

[QGIS-Developer] Plugin [1587] Instituto Canario de Estadística (ISTAC) approval notification.

[noreply]

Plugin Instituto Canario de Estadística (ISTAC) approval by pcav.
The plugin version "[1587] Instituto Canario de Estadística (ISTAC) 0.3" is now 
approved
Link: http://plugins.qgis.org/plugins/istacqgis/
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] No LTR on SUSE repos

[Alessandro Pasotti]

On Mon, Apr 8, 2019 at 10:25 AM Paolo Cavallini 
wrote:

> Hi Yann,
>
> On 08/04/19 10:01, Yann POUFFARIX wrote:
> > Back in time.
> >
> > No answer from maintener and of course nothing listed here
> >
> > https://software.opensuse.org/package/qgis-ltr
> >
> > what to do x)
>
> I think you are left with the choice of either doing the package
> yourself, or hire someone to do it. Of course you can always compile
> qgis yourself, it shouldn't be too hard if the dependencies are properly
> packaged.
> Cheers.
>
>
Just curious: why should we invest money for Apple packaging and not for
SuSE packaging?

I think we should be inclusive and supportive for all users on all
platforms.



-- 
> Paolo Cavallini - www.faunalia.eu
> QGIS.ORG Chair:
> http://planet.qgis.org/planet/user/28/tag/qgis%20board/
> ___
> QGIS-Developer mailing list
> QGIS-Developer@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer



-- 
Alessandro Pasotti
w3:   www.itopen.it
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] Bug #21460?

[Paolo Cavallini]

Hi Cory,

On 08/04/19 00:57, Cory Albrecht wrote:
> I was wondering when Bug #21460
>  was going to get some attention?
> It's about  QGIS not forming proper PostGIS SQL updates correctly,
> specifically for null values.

may I suggest you an interesting read?
https://nyalldawson.net/2016/08/how-to-effectively-get-things-changed-in-qgis/
Cheers.
-- 
Paolo Cavallini - www.faunalia.eu
QGIS.ORG Chair:
http://planet.qgis.org/planet/user/28/tag/qgis%20board/
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer

Re: [QGIS-Developer] No LTR on SUSE repos

[Yann POUFFARIX]

Back in time.

No answer from maintener and of course nothing listed here

https://software.opensuse.org/package/qgis-ltr

what to do x)
___
QGIS-Developer mailing list
QGIS-Developer@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer