[ckan-discuss] A quick hello

[Rufus Pollock]

On 3 March 2010 19:56, Friedrich Lindenberg <friedrich at pudo.org> wrote:
> Hi all,

Weclome!

> I just wanted to introduce myself to the CKAN community. Together with
> Daniel and a few others, I'll start working on an open data directory
> for Germany in the coming weeks. Having explored CKAN for a few days,
> I'm incredibly happy to find that such a nice package does exist for
> metadata management.
>
> Having worked extensively with Pylons and SQLAlchemy in the past [1], I
> hope that the (mostly i18n-related) modifications we're going to make in
> that process can provide some value back to the CKAN project.

Great to hear you're familiar with Pylons and SQLAlchemy :)

> In the longer term, I would also like to add to the ongoing development
> effort - maybe someone could propose a few tickets for a CKAN newbie to
> get started with? Here's a few areas I would like to contribute to:
>
> * Multi-Server datapkg

By this I assume you mean datapkg using more than one different ckan
at once or different ckan at different times? Both are really good
ideas but the latter is already possible via the config :)

> * 0.5 SQLA migration

This is for vdm (versioned domain model) and may be a little dull as a
first task ;)

> * Solr search as replacement for DbIndex

This work is already "in-progress" but we'll probably need help at the
next stage.

> * Anything related to versioning

This is something worth discussing live (either via irc or skype or ...)

To make some further suggestions based on the areas you are interested in:

## 1. Work on datapkg.

There's lots that could be done but tickets are rather limited at the
moment [1]. Things I can immediately suggest are:
  * Automate creation of a "distribution" from a file on disk.
     * This would involve integrate package resources into the
metadata format (currently missing)
  * Integrate uploading into the registration process (for example by
storing the file on archive.org).

As a start I'd suggest taking a look at the codebase, running the
tests etc. Then I suggest we could talk directly about any questions
you have and to do specifically.

[1] http://knowledgeforge.net/ckan/trac/query?component=datapkg

## 2. Browse packages by rating

<http://knowledgeforge.net/ckan/trac/ticket/183> -- this is a nice
simple ticket that would get you familiar with the ckan codebase.

## 3. Integrate the CKAN registry with actual storage of the
content/data. Tickets along these lines are:

<http://knowledgeforge.net/ckan/trac/ticket/107>
<http://knowledgeforge.net/ckan/trac/ticket/186>

Another item along these lines we've thought about is a caching
service whereby you automatically go and cache (some or all) of the
datasets referenced on CKAN on a data storage service. This would both
be a service to people registering data (their data is backed up) and
ensure that if data disappears a copy still exists.

Let me know what you think and let's arrange to chat at some point
either on skype or irc, perhaps along with some of the other CKAN
contributors.

Rufus
-- 
Open Knowledge Foundation
Promoting Open Knowledge in a Digital Age
http://www.okfn.org/ - http://blog.okfn.org/