[QGIS-Developer] The need for intuitive complex feature edit support in QGIS 3+

Hugo Mercier hugo.mercier at oslandia.com
Tue Jan 9 07:53:49 PST 2018


Hi Rui,

Co-author of the mentioned "GML application schema toolbox" plugin here :)

The people who first launched this plugin have similar goals I think (I
added Sylvain from BRGM in cc).
Even if this is still small (we are at a second iteration where the
first was a proof of concept), some interest is now raised among
possible users in Europe.

It was chosen in this first step to focus on a transformation from an
object model (GML) to a relational database model in order to fit QGIS'
data model. This is a hard exercise (I've done a first shot and Even
Rouault a bigger one in GDAL) but already gives interesting results.

The QGIS plugin is a graphical interface over this GDAL driver. It also
creates QGIS layers and relations and forms between them from metadata.

Moreover, the GDAL driver is already able to be used in "writing mode",
meaning that you can not only read complex features content, but modify
an existing one. This is a first step toward "producing" INSPIRE data as
you said.

This gives something usable, even if apart from simple toy examples, the
number of layers, fields and relations become very hard to follow. There
should still be a way to represent data and relations in a more
"hierarchical way" sometime I think.

If you have experience with telecom software that have more advanced
features and some funding to enhance QGIS in this field, I am sure we
can come with enhancements that make this kind of projects more and more
useable.

On 09/01/2018 16:13, Régis Haubourg wrote:
> Rui, I would formulate that differently.
> I have been working with INSPIRE datasets tooand a bit involved in some
> hydrography model workgroups.
> 
> IMO, gml is nice because it allows to fully respect the UML design and
> offer theoretical interoperability. Seems become darker when trying to
> use it for real interoperability and everyday's use.
>  It is not meant to be a physical implementation for editing. It is
> extremely hard to work with, with because of cascaded schema
> dependencies, XML heaviness, lack of index,  very wide possibilities of
> spatial object modeling (so much ways to describe a simple box is crazy).
> So to me, the GML is a exchange format, and should be implemented in a
> transactional DB for editing , or simply for fast reading. That's why
> BRGM approach relying on a spatial DB for real use sounds a goopd
> strategy to me.
> A bit like UML model of database can be used directly and needs to
> implemented in a physical model in a relational database for real work.
> 
> Regards,
> régis
> 
> 
> 2018-01-09 15:39 GMT+01:00 Rui Cavaco <rpcavaco at gmail.com
> <mailto:rpcavaco at gmail.com>>:
> 
>     Thanks for your response Régis.
> 
>     Yes, I read the documentation for that project. I think that's a
>     good, but small, starting point. Anyway the docs mention that "the
>     aim is to develop tools to manipulate Complex Features streams in a
>     GIS desktop application.". So they might be seeking for something
>     bigger.
> 
>     It seems that there are now some solutions, including that extension
>     you mentioned,   to consume complex data. But we see little, or
>     none, efforts on  how to produce such data. I think current user
>     interfaces are not adapted to the complex feature universe.
> 
>     Thanks for your suggestion. I should directly contact the GMLAS
>     project team.
> 
> 
>     Best regards
> 
>     Rui Cavaco
> 
>     Régis Haubourg <regis.haubourg at gmail.com
>     <mailto:regis.haubourg at gmail.com>> escreveu no dia terça, 9/01/2018
>     às 13:46:
> 
>         Hi Rui,
>         did you have a look at this project ?
>         https://github.com/BRGM/gml_application_schema_toolbox
>         <https://github.com/BRGM/gml_application_schema_toolbox>
> 
>         I think you purchase the very same goal, a common approach seems
>         a good idea !
>         Regards,
>         Régis
> 
>         2018-01-09 1:17 GMT+01:00 Rui Cavaco <rpcavaco at gmail.com
>         <mailto:rpcavaco at gmail.com>>:
> 
>             Hello list.
> 
>             My name is Rui Cavaco, a supporter for OSGeo Portugal, and I
>             see the need for some future major changes in desktop GIS
>             user interfaces in order to facilitate complex features
>             editing and querying.
> 
>             GML and INSPIRE are about complex features but so are fiber
>             optic networks. Complex features could be also very
>             productive in simpler cases like the management of city road
>             signs and indications and other municipality themes.
>             In order to properly support complex features I think we
>             need to go further than the simple and old three-part GUI
>             comprising TOC, map and attribute table. For example,
>             attribute and form views must have "drill down"
>             capabilities. As for the TOC, subdivding layers, as the
>             GMLAS extension does, is not enough. Something like
>             tree-view windows showing object hierarchies and complex
>             objects' internal contents must exist. This is the exact
>             same as schematics / synoptic views provided by specialized
>             "closed source" GIS tools provided for telecom and other
>             utilities management. Also I think the TOC should be very
>             interactive and adaptive, in order to make possible to
>             expose the intrincacies of sublayers without cluttering the
>             whole layer tree with details uneeded for the current user
>             context.
> 
>             Me and others discussing this subject in OSGeo-PT chat, we
>             are convinced that without a largely available and intuitive
>             editing support for complex features INSPIRE will soon be
>             (some say already is) dead, despite all the the EU legal
>             obligations.
> 
>             I suppose this is not a job for a single developer or a
>             small team. I imagine this might require some profound
>             changes in QGIS. I don't think that all these GUI changes
>             mentioned could be "compacted" in just an extension.
> 
>             Funding for this effort could be raised from
>             INSPIRE-interested EU organizations and member state
>             government agencies. Also telecom companies and other
>             utilities managers can be interested. Dedicated "closed
>             source" GIS solutions for utilities are so absurdly
>             expensive that this can open an opportunity window for Open
>             Source based solutions.
> 
>             I would like to join efforts with others sharing this vision
>             in order to help make it happen in future releases of QGIS.
> 
>             Rui Cavaco
> 
>             _______________________________________________
>             QGIS-Developer mailing list
>             QGIS-Developer at lists.osgeo.org
>             <mailto:QGIS-Developer at lists.osgeo.org>
>             List info:
>             https://lists.osgeo.org/mailman/listinfo/qgis-developer
>             <https://lists.osgeo.org/mailman/listinfo/qgis-developer>
>             Unsubscribe:
>             https://lists.osgeo.org/mailman/listinfo/qgis-developer
>             <https://lists.osgeo.org/mailman/listinfo/qgis-developer>
> 
> 
> 
> 
> 
> _______________________________________________
> 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
> 


More information about the QGIS-Developer mailing list