[QGIS-Developer] Can I backport bugfix with UI updates and fixing QGIS Server regression

René-Luc Dhont rldhont at gmail.com
Mon May 20 08:27:53 PDT 2019


Hi Devs,

I have removed the commit cherry-picked from Nyall's pull-request that 
update the layout designer dialog.
So my Pull-request only contains changes for QGIS Server to fix the 
regression about printing selection.

I have introduced the capability to print selection in QGIS Server for 
version 3.4 LTR with the pull-request [Server] Reactivate the capability 
to print selection with Server 3.4 https://github.com/qgis/QGIS/pull/8320
This capability is available for image export and was tested and this 
pull-request fix inconsistent use of layout render context flags and add 
test to validate it.

I really hope this will be fix it version 3.4.9.

Regards,
René-Luc

Le 14/05/2019 à 10:00, René-Luc Dhont a écrit :
> Hi all,
>
> In the PR to fix the QGIS Server regression I only need to backport 1 
> commit on the 2 provided by Nyall to fix the issue: Layout export - 
> raster divided into tiles, edges evident in pdf/svg 
> https://issues.qgis.org/issues/19500 in the Pull request 
> [needs-docs][layouts] Add checkbox to disable raster tiling for 
> PDF/SVG exports https://github.com/qgis/QGIS/pull/9016
>
> The one I need is a fix: Fix inconsistent use of layout render context 
> flags 
> https://github.com/qgis/QGIS/pull/9692/commits/0954deaf97638519e9f4b3e28855e03a8a34f5c1
> This one concoenrs only render context flags for PDF and SVG output. 
> In the case of image output, the layout render context flags are well 
> set.
>
> So do you think, the Pull Request could be accepted, if I remove the 
> UI modifed part and only conserve the modification needed for server ?
>
> Regards,
> René-Luc
>
> Le 13/05/2019 à 08:04, Paolo Cavallini a écrit :
>> Hi all,
>>
>> On 12/05/19 09:41, Régis Haubourg wrote:
>>
>>> I'd be pleased to test things here but use cases are so numerous that I
>>> fail to see how to do this efficiently.
>>> I hope we'll one day be able to run those smoke tests as discussed 
>>> at La
>>> Coruna.
>> I was about suggesting doing community-wide testing (having a test
>> package would greatly help). Is it so difficult to identify and test
>> possible regression areas?
>> Cheers.
>



More information about the QGIS-Developer mailing list