[Qgis-developer] Improving the legend for map-units, data-defined symbology, diagrams and labels

[Régis Haubourg]

Andreas Neumann wrote
> 1. Support for legends for graphical properties defined with map units. 
> Currently these legends display in the print composer at very strange, 
> unpredictable and unusable sizes. E.g. a stroke-width or dash defined in 
> map units is either way too big or too small. Same for symbol sizes 
> defined in map units. We need a way to translate these map units in 
> usable paper units suitable for display in the legend. Let the map 
> author define a corresponding mm size (along with the map units) - 
> solely for the display of the legend.

+1 !


Andreas Neumann wrote
> 3. Legends for diagrams (pie charts, bar charts) - something similar to 
> http://esri.force.com/servlet/rtaImage?eid=087300000008Nki&feoid=Body&refid=0EM30000000Co0O 
> - showing both sizes and colors - nicely stacked into each other.

Hi Andreas, 
we are now launching a contract that will add to QGIS size varying legend in
composer and mapcanvas legend, pretty similar to the one you show. It will
not solve the diagram color part, but adds a size varying legend below color
legend when it exists (and it doesn't for diagram now). 

It will also add label data defined widgets to all symbology options,
instead of a new dialog. 

It will also add an assistant GUI to help a user choose proportionnality
rules (min size, max size fetched from data or expression, flannery, surface
size or linear function..)

I hope we will make it for 2.8, I got delayed a lot on that (we talked about
it a while ago with Martin)


Andreas Neumann wrote
> 2. Display of legends for graphical properties defined with data-defined 
> symbology. We need a way to specify common graphical properties in the 
> legend that are of interest in the legend. F.e. - if I define the colors 
> data-defined I would like to be able to specify the expected colors that 
> may appear in my data/project. This could be either done dynamically (by 
> scanning the data) or statically (decided by the map author). Perhaps 
> both methods. To me, the static method may be more predictable and more 
> performant.


Unfortunately, we will not have in this contract this data defined legend
collect tool, except when using size varying assistant. I would have liked
to, but I think it deserves a QEP. 
I  would suggest a tool to collect asynchroneously existing symbols rendered
on the map extent, in a new tab of layer properties, where user can finalize
the legend by hand. Agreed with you, I would love an automatic tool, but I
can see how we could do that. 

Cheers
Régis



--
View this message in context: http://osgeo-org.1560.x6.nabble.com/Improving-the-legend-for-map-units-data-defined-symbology-diagrams-and-labels-tp5172915p5172990.html
Sent from the Quantum GIS - Developer mailing list archive at Nabble.com.