All posts in the topic UK govt data licencing model (Short link)
Summary
- There are 7 posts — by 4 authors — in this topic.
- Latest post made by Mike Pearson at 2009 Jun 18 11:28 NZST
Hi
Th OPSI has released a new licensing model for government data.
"The new licence terms are compatible with other standard
licences such a Creative Commons and GNU GPL;"
I think it is very relevant that the UK, which is touted by many as a
leader in this and other initiatives, is dotting the i's as they go
along.
I would like to see a similar licensing regime in NZ. All the work has
been done elsewhere, all we need to do is adopt it.
On 2009-06-18, at 0915, Don wrote:
> Th OPSI has released a new licensing model for government data.
>
> "The new licence terms are compatible with other standard
> licences such a Creative Commons and GNU GPL;"
>
> I think it is very relevant that the UK, which is touted by many as a
> leader in this and other initiatives, is dotting the i's as they go
> along.
>
> I would like to see a similar licensing regime in NZ. All the work has
> been done elsewhere, all we need to do is adopt it.
Of course the ultimate is actually to have them just adopt licences
such as the Creative Commons, rather than just claiming compatibility.
I think the last thing we want to encourage is a proliferation of
licences, when, for most of the data we're trying to open, CC is
mostly* adequate. Otherwise we'll end up with 50+ licenses, as has
happened with FOSS licences.
One grey area I've seen, and I'd been interested to hear from others,
is the following situation. The organisation releasing data would like
the following:
1. No derivatives are to be created (e.g. dataset needs to remain
authoritative).
2. Format-shifting is allowed (e.g. conversion to device-specific
format).
When I've looked in the past, it would appear impossible to use CC to
licence data in this way. E.g. CC-ND appears incompatible with format
shifting, but protects the original authoritative data. Everything
that is not CC-ND does not protect against the creation of alternative
works and threatens the authoritative nature of the data.
It seems to me that CC actually needs the option to allow format
shifting, but not modifying the meaning of the underlying data.
Consider the importance of this for things such as electoral
boundaries where only the Electoral Commission is able to modify them,
but format shifting whilst retaining the meaning/accuracy of the
boundaries is OK (e.g. to convert the boundaries to a format that a
handheld GPS can utilise).
Cheers Gav
Further to my last email and just to clarify why the subject is
important the following article about Tim BL finishes with this point:
"Sir Tim pointed to a mashup of bicycle accidents using
statistics from the Department of Transport. Except someone –
not me – then pointed out in the comments that the mashup must
breach Ordnance Survey's rules on "derived data" (data plotted
on an OS map re-plotted on a non-OS map)"
http://www.guardian.co.uk/technology/2009/jun/17/charles-arthur-free-data
On 18/06/2009, at 9:31 AM, Gavin Treadgold wrote:
> It seems to me that CC actually needs the option to allow format
> shifting, but not modifying the meaning of the underlying data.
> Consider the importance of this for things such as electoral
> boundaries where only the Electoral Commission is able to modify them,
> but format shifting whilst retaining the meaning/accuracy of the
> boundaries is OK (e.g. to convert the boundaries to a format that a
> handheld GPS can utilise).
It seems to me that codifying this in a license is pointless, because
anyone stupid enough to alter election boundaries will do it
regardless of what the license says.
On 2009-06-18, at 0935, Don wrote:
> Further to my last email and just to clarify why the subject is
> important the following article about Tim BL finishes with this point:
>
> "Sir Tim pointed to a mashup of bicycle accidents using
> statistics from the Department of Transport. Except someone –
> not me – then pointed out in the comments that the mashup must
> breach Ordnance Survey's rules on "derived data" (data plotted
> on an OS map re-plotted on a non-OS map)"
I recall quite a few issues like this being raised a few months back
when UK.gov had their mashup competition, including click-through web
licences and the like that created barriers to entry in adopting and
utilising the data.
Cheers Gav
On 2009-06-18, at 0937, Nathan Torkington wrote:
> It seems to me that codifying this in a license is pointless, because
> anyone stupid enough to alter election boundaries will do it
> regardless of what the license says.
Agree - but I still think there is a benefit to encouraging release
under widely used and understood licences in favour of custom ones,
and would prefer to seem the CC expanded rather than see new licences
spring up due to the inadequacies of existing ones.
And surely the same argument could be made for FOSS licences - if the
source is available, people will do with it whatever they choose
regardless of the licence ;)
Cheers Gav
There are advantages of the CC licensing framework, in that: - it is internationally recognised; - has sufficient critical mass to grow; - has mechanisms to change - has started to be embodied in US popular software (e.g. Flickr, Google searches). Most NZers already use such US tools to manipulate the data. I think the issues will be around liability / ongoing obligation. The NZTA Infoconnect agreement demonstrates this. The agency wants to make sure information it supplies is used responsibly, and does not cause loss/death. -- Mike Pearson mobile: +64 (21) 631 731 http://www.linkedin.com/in/mikepearsonnz