[ckan-discuss] Summary of the CKAN Geo Community meeting

[Adrià Mercader]

Hi all,

We had a great meeting yesterday regarding the CKAN geo features,
thanks very much to all who attended!

After some introductions, we did a quick overview of the current
geospatial model and the existing features (have a look at this wiki
page [1] if you want to know more).

We then discussed the future enhancements and features that could be
added to CKAN. Three main areas of work are more or less defined:

* Closer integration with the web interface
We would like to make easier for users to define the geographic extent
of a dataset, via drawing a box or point on a map or using a geocoding
service to extract location information from a place name (or a
solution that combines both). The general agreement is that it will
require a previous work of defining the UI and requirements right.
We also want to integrate the spatial search into the web search form.
In that case, the UI is not so complex (as is limited to the
capabilities offered by the spatial API i.e. bbox), but the challenge
will be integrate both the textual and spatial search, which right now
are two different services.
Maybe Sean Hammond will be able to give as a hand with one of this tasks.

* Provide geo-enabled feeds and API (GeoJSON, GeoRSS, KML)
The goal is to make really easy to map the geometries of the datasets
in mapping librarires like OpenLayers or Google Maps or desktop
applications like Google Earth or GIS programs, enabling user to build
geo mashups or use the metadata stored in CKAN in new and interesting
ways.

* CSW and INSPIRE Support
This is another area where we see a lot of potential for CKAN.
Enhancing the CSW support and other geo features and taking full
advantage of the existing catalogue features of CKAN, we believe it
can be an ideal tool for publishers wishing to publish their geo and
non-geo datasets. To achieve it, we need a better CSW interface, and
here is where the excellent pycsw library [2] can play a crucial role.
Tom Kralidis, one of its developers, gave as a detailed introduction
to the library, its capabilities and internals. Both pycsw and CKAN
underlying technology and data model are fairly similar, and we are
really optimistic in regards to the possibility of an integration
between both. Tom kindly offered his support in making this happen.
Next step could be to make the necessary UI modifications (e.g. custom
forms and validators) for turning CKAN in a web front-end for a fully
fledged CSW/INSPIRE server.

As you can see, there are lots of exciting things going on and we
would love to hear you thoughts on them (even better if you are
willing to contribute!). Feel free to ping this list or meet us on the
IRC channel (ckan at irc.freenode.net).

You can check the notes of the meeting in the public etherpad:

http://ckan.okfnpad.org/ckan-geo-meetup


Best regards,

Adrià


[1] http://wiki.ckan.org/Geospatial_Capabilities
[2] http://pycsw.org