[Aust-NZ] Geoscience Australia goes CC-BY [SEC=UNCLASSIFIED]

[Bruce Bannerman]

Thanks Gavin,

That clarifies a few points.


We have data that we manage as data custodians, and as custodians, we choose not to allow others to edit that data for issues relating to data quality etc (In our business that is a valid concern). Therefore we're releasing 'whole' datasets as is. 

We have a couple of choices as to licenses:

- go to a lawyer and draft a separate license agreement to allow access to the data.

- use a commonly accepted license, e.g. a CC license, that our laywers will accept.


(My personal preference is to use a CC license, however, we have yet to see if that will be the way we go.) 

The first option will have you going to your lawyer for a translation to see if it is OK to use. The second option (CC), should avoid that.


Wrt Spatial Metadata, I believe that all spatial professionals have a responsibility to *maintain* and release a (url to their) quality metadata record with their data sets.

Wrt 'provenance' (isn't that part of the Metadata ;-) ), as a user of a spatial dataset, I'd prefer to know where that data came from and what its metadata had to say, particularly:

- if I needed to rely on it for a particular reason; or

- if I needed to rely on an analysis derived from it.

This is especially important if I had to make a decision that had $$$$ or similar related to it.


Now when data gets included within another dataset 'provenance' becomes more important. ISO 19115 can handle this situation with record level metadata if required (with your help of course).


Wrt mashups etc, this is the current nature of the application of the technology. Why look to governments to fix this by releasing data 'on a free for all basis'?


Just a few thoughts...


Bruce

 

> -----Original Message-----
> From: Gavin Treadgold [mailto:gavin.treadgold at gmail.com] On 
> Behalf Of Gavin Treadgold
> Sent: Monday, 7 December 2009 12:44 PM
> To: Aust-NZ at lists.osgeo.org
> Cc: Brent Wood; Bruce Bannerman
> Subject: Re: [Aust-NZ] Geoscience Australia goes CC-BY 
> [SEC=UNCLASSIFIED]
> 
> 
> On 2009-12-07, at 14:01 , Bruce Bannerman wrote:
> > Sorry, I'm missing the point. What is it that you don't 
> like about CC?
> 
> I was going to jump in earlier, and now seems a good time.
> 
> I believe the point that Brent is trying to make - is that 
> when it comes to data governance, licensing and copyright is 
> just a small aspect of what is required when releasing 
> government data. As Brent has pointed out, metadata is also 
> as important, as is the provenance.
> 
> It is great to see Australasian government agencies adopting 
> CC-BY for copyright and licensing, but I (and I assume Brent) 
> also recognise that there are plenty of other aspects related 
> to the release of government data, and whilst we should be 
> congratulating NZ and Australian agencies for releasing data 
> under permissive licenses, issues such as provenance, 
> metadata, and indeed how attribution of data in complex 
> mashups should be handled. 
> 
> On the attribution angle - we've had plenty of discussions on 
> the NZ Open GIS list about how attribution should be handled 
> for data going into OSM - is a single reference in the OSM 
> wiki OK? Or should we be tagging every node and derived line 
> with a source tag that attributes the source? Both of these 
> would attribute the data correctly, but one approach is 
> likely to be far more robust at tracking both the provenance 
> of the data and correctly attributing the original source.
> 
> CC does not provide any guidance on best practice for how the 
> attribution should occur, it just says that it must be 
> attributed. That is fine for creative works, but is a lot 
> harder to manage when you have factual data such as roads, 
> that anyone, including someone like myself can easily edit 
> and correct errors in the originally released government 
> data. Especially if you have to manage the provence down to 
> the node and connecting line level that geospatial types 
> sometimes expect.
> 
> Cheers Gav