Essentially we found the bottom line for many data owners we talked to was that they wanted attribution if data was used in other contexts. While some data owners might be happy with CC0 or PDDL, we decided at this stage it was easier to apply one license across the piece.

I’m not sure that CC-BY is the correct one – I need to discuss with the rest of the team, but I know (apart from CC0) CC licenses aren’t really suitable for data – we may need to look at something like the Open Government License instead (http://www.nationalarchives.gov.uk/doc/open-government-licence/) – but I’ll need to look at this further and talk to others.

In terms of the attribution trail – my personal view is that we probably shouldn’t get too worked up about this (for the moment at least). While I know some feel that we should look at how we track provenance for individual pieces of data my view is that we shouldn’t worry about this level of detail and should think about attributing overall sources (e.g. “This data was compiled from locah.ac.uk and data.open.ac.uk”) rather than tracking where each piece of data came from. I’m aware there are going to be issues if you only track at a high level, but I find it hard to get to worried about this. However, it is early days and I may well change my mind!

I see from http://data.open.ac.uk/ that you’re using CC-BY for your Linked Data. I wondered how you came to that decision, and if you foresee any attribution trail problems? I’m looking at CC0 and ODC Public Domain Dedication and Licence (PDDL) right now, and ‘no going back’ aspect is making me think twice.

Cheers, Adrian