[Qgis-developer] Vertex editor
Denis Rouzaud
denis.rouzaud at gmail.com
Sun Sep 13 23:35:26 PDT 2015
On 09/14/2015 08:16 AM, Andreas Neumann wrote:
> Hi,
>
> Marco agreed to participate in our paid bug fixing program that should
> start after feature freeze (which - according to our website - is on
> September 26. If PSC votes for him - he can then work several days
> dedicated to bug fixing.
I am not sure that QGIS bug fixing program is the right way to go.
It wouldn't be fair that an unfinished paid work is pushed to master and
later on the group sponsors the end of the development.
I suggest to wait for Marco's answer regarding the schedule for fixing
the node tool.
I think that waiting again 2 weeks is a bit long as this tool is crucial
in QGIS. If he can't make it, then the solution of paid bug fixing by
another dev can be discussed, but that's a bit odd to me.
>
> Marco is traveling the next two weeks (FOSS4G South Korea) - but I
> assume - after he is back, he can continue working on the
> digitizing/geometry issues.
>
> Andreas
>
> On 14.09.2015 07:19, Denis Rouzaud wrote:
>>
>> On 09/13/2015 11:32 PM, Nyall Dawson wrote:
>>> On 14 September 2015 at 07:07, Saber Razmjooei
>>> <saber.razmjooei at lutraconsulting.co.uk> wrote:
>>>> The new changes assume:
>>>> - User always wants to move a vertex
>>>> - Only one single vertex is selected
>>>>
>>>> But in general, the node tool is also heavily used for deleting,
>>>> adding and moving vertices. With the new behaviour it is a bit
>>>> confusing for users how the other tasks can be done, when it gets
>>>> automatically locked on a single node.
>>> There's a long blocker list relating to regressions in the map tools,
>>> eg:
>>>
>>> - deleting a node: the rubber band is not removed
>>> - deleting a node: when finished, the next node should be pre-selected
>>> (this pre-selection should by synchro with the table view selection
>>> model).
>>> - right-click should cancel the edition of the node (but keeps it
>>> pre-selected)
>>> - for point layers, there is no visual feedback of the point being
>>> moved (no rubberband)
>>> - editing in table view: changes should be applied to the rubberband
>>> on keypress (not on editing finished), and when editing is finished it
>>> should ends the digitizing (clear rubberband)
>>> ( all from http://hub.qgis.org/issues/13276)
>>> - split tool errors: http://hub.qgis.org/issues/13273 and
>>> http://hub.qgis.org/issues/13347
>>> - inability to add parts to null geometries - refs
>>> http://hub.qgis.org/issues/12885
>>>
>>> I tried to do some digitising in master last week and honestly it was
>>> unusable for geometry editing. Lots of regressions and unexpected
>>> behaviour compared to the smooooth editing machine that 2.8 is! I
>>> understand master is a work-in-progress, but can we get some
>>> confirmation of which issues will be addressed prior to 2.14?
>>> Marco/Denis? If there's no funding to address these then I think they
>>> should be prioritised for the 2.14 bug fixing period.
>> All these issues shall be fixed prior to the release.
>> Marco, do you have any schedule regarding this?
>>
>>
>>> Nyall
>>> _______________________________________________
>>> Qgis-developer mailing list
>>> Qgis-developer at lists.osgeo.org
>>> http://lists.osgeo.org/mailman/listinfo/qgis-developer
>>>
>> _______________________________________________
>> Qgis-developer mailing list
>> Qgis-developer at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/qgis-developer
>
> _______________________________________________
> Qgis-developer mailing list
> Qgis-developer at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/qgis-developer
More information about the Qgis-developer
mailing list