[data-protocols] New Data Package Format

Tue Jun 19 16:31:21 BST 2012

On 15 June 2012 18:23, Eric Busboom <eric at clarinova.com> wrote:
>
> On Jun 15, 2012, at 9:54 AM, Eric Busboom wrote:
>
>> Hi All,
>>
>> Rufus asked me to move this discussion over to this list from the CKAN dev list -- apologies for the repost -- so, I'll repost the original message and then
>
> Here is my reply to Rufus'  response to the original version of this message that was posted to a different list. ( Got that? )
>
> --
> On Jun 14, 2012, at 11:22 AM, Rufus Pollock wrote:
>
>> * CKAN datasets - datasets (metadata + possibly data) in a CKAN instance
>> * "Data packages" as per http://www.dataprotocols.org/en/latest/data-packages.html
>>
>>  I think we should focus discussion on "data packages".
>
> I agree; the package format comes first, then figure out where and how it should be supported and integrated.

Generally agree but think this must go hand in hand with some
implementation. My point re CKAN is that it is quite wide-ranging in
the types of data it must manage. I think we will gain by being more
restrictive in our first pass.

>> This is a good thing as it shows there is real commonality
>> here. This is also suggests that we could possibly combine the two!
>
> Absolutely. At the very least, the structures can be be harmonized at some level.

+1

>> PS: re your source spec see https://github.com/okfn/dpm/blob/master/doc/new-plan-2011-nov.rst
>
> Thanks;  I hadn't realized you also had a source format. It is really encouraging to see that two independent solutions to a problem have a similar structure -- means we're both on the right track.
>
> I'm reviewing the dpm code and design while I work on our version. I'll get back to the list when we have some functional code and can start exploring what the two formats should share.

Is it worth trying to draft an initial spec collaboratively in an
etherpad as a first pass (and try two independent implementations) --
it would seem to try this than harmonize afterwards? Also if you look
at the current data packages spec do you have specific change requests
you'd want now?

Rufus