[ckan-discuss] CKAN activity

[David Read]

Here's the latest from the CKAN world:

CKAN 1.5 finally got released on Wednesday and the dev team is very
pleased! Lots of new features released: see the announcement
http://ckan.org/2011/11/09/ckan-1-5-release/ It is great to also
release an Ubuntu package, which we've already used to deploy eight
CKAN servers.

Mozfest hit London last weekend, with three of the team there as a
"Human API" for data.gov.uk. We helped hackers take advantage of the
UK government data for various hacks related to Mozilla's theme
"Media, Freedom and the Web". Aided by superb espresso and working in
refreshingly modern facilities, it was great to work with data
wranglers and data-driven journalists from around the world.

The allowable format for CKAN's tag (taxonomy for datasets) is being
widened to allow spaces, capital letters and foreign characters. This
is needed to accept tags in metadata imported from elsewhere -
essential for projects where CKAN is used to aggregate datasets, such
as publicdata.eu, IATI and data.gov.uk.

OKF's OpenSpending team provide an easy way to examine UK government
spend http://data.gov.uk/openspending, and it turns out it's getting
lots of use amongst the upper echelons of the government itself! So
now we're improving the intergration of the app and giving the service
a better server.

We're having a big push on harvesting of INSPIRE datasets for the UKLP
project at the moment. Some improvements are rather technical - Adria
just added something called 'harvesting multiple resource locators for
each dataset.' but others are really visible to end users: we are
currently doing some 'real-life' testing on the bounding box based
geo-spatial search powered by PostGIS. It shouldn't be long before
this goes public using UK data, and soon afterwards we'll have an
international equivalent on thedatahub.org.

Activity Streams is a new feature being kicked off in CKAN by new
developer, Sean. It will be a bit like what you get on your Github
dashboard, to help publishers and data consumers keep track of
datasets.

The QA Extension, which checks links, file sizes and formats, is being
refactored by John to run off the Celery queue framework. This will
help us cope with the sheer volume of links to process and allow us to
slot in more background processing tasks, for which there are plenty
of ideas!

The Groups Refactor work is now started, making datasets more clearly
grouped by publisher or subject, and allows easier set-up of the
permissions.

DataExplorer is being merged with Max Ogden's Recline, so browsing
table data in Webstore with Google Refine-like controls - a really
easy way to do data clean up and analysis. See Rufus' demo here:
http://okfn.github.com/recline/

CKAN has now embraced Git, since more people know it these days,
compared to Mercurial. The new repository is here:
http://github.com/okfn/ckan