[QGIS-Developer] Discussing default snapping and node editing settings

[Régis Haubourg]

Hi,
+1 !
However, I see lots of customers switching to the "Snap all layer"
settings, "because it is simpler". And then they face very slow snapping
index creation issues.
In the QWAT project, we have been thinking of exposing the indexing
strategies in the UI of the advanced snapping.

see here:
https://qgis.org/api/classQgsSnappingUtils.html#a2ea7577667b94657e14f63ba7dd0bf07
----------------------------
IndexAlwaysFull

For all layers build index of full extent. Uses more memory, but queries
are faster.
IndexNeverFull

For all layers only create temporary indexes of small extent. Low memory
usage, slower queries.
IndexHybrid

For "big" layers using IndexNeverFull, for the rest IndexAlwaysFull.
Compromise between speed and memory usage.
IndexExtent

For all layer build index of extent given in map settings.
--------------



And for WFS and any other cached provider, we may probably add some
exceptions to the global strategy to avoid fetching the remote datasource
by default. Using the local cache is better. However, when doing WFS
editing, this start to look like a can of worms.


Le mer. 12 sept. 2018 à 16:58, Andreas Neumann <a.neumann at carto.net> a
écrit :

> Hi,
>
> I noticed that by default QGIS is set to edit ALL layers and also snaps to
> ALL layer by default. I do think that this is a bad default setting, esp.
> with larger projects.
>
> We had issues because users had WFS layers in their project (read only, as
> reference) and QGIS tries to index these WFS layers and to get their
> vertices in order to snap to them. Then QGIS hangs and freezes and there
> are network time outs. Apparently one has to kill QGIS using the task
> manager.
>
> I think the better default is to edit and snap only in the active layer by
> default.
>
> Opinions?
>
> Andreas
> _______________________________________________
> QGIS-Developer mailing list
> QGIS-Developer at lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-developer/attachments/20180912/e2ca2b7a/attachment-0001.html>