SSC as the catalyst
Laurence Millar, the Chief Information Officer for the NZ Government recently blogged about some thoughts around government data and Web 2.0. It is exciting to see someone in such a position of influence raise these issues, and hopefully lead the charge to promote change in government agency practices.
One angle I would like to extend upon, is Tim O’Reilly’s quote.
What if you don’t think of what you produce as the “final product” but rather as a step in an information pipeline, what do you do differently to add value for downstream consumers?
I think that this is a fundamental point that Government needs to get to grips with. When you look at government websites these days, most of them are designed around the philosophy of being just the terminal point of an information production chain – the result being information products expressed entirely in the form that the government agency ‘expects’ citizens to digest them. Most Government websites therefore only produce web pages, and pdf documents – they contain relatively little in raw data in a format that is more accessible for citizens. The exceptions of course are those agencies that have extensive mandates for publishing vast datasets such as Statistics New Zealand and Land Information New Zealand.
Laurence concludes his post with absolutely the correct next action required
…open up our content, expose our data so that it is easier to consume, rather than applying resources to redesign information dissemination. By creating objects that others can assemble we are likely to be significantly more successful at ensuring New Zealanders have access to the government information when, where and how they prefer.
A classic example is the recently released Atlas of Socioeconomic Deprivation in New Zealand (NZDep2006) published by the Ministry of Health. Whilst it contains a wealth of data, all of the downloadable forms from the website have been constricted by their publication in a read-only pdf format. The multitude of maps are all pdfs. The tables and reference information are in pdfs. This makes it near impossible to extract and utilise the data “when, where and how they prefer“. All is not lost however, I did contact MOH and they did provide a CD with a couple of hundreds MBs of data, including shapefiles for use within Geographical Information Systems. However, the licensing of the data is still somewhat unclear – Crown Copyright is not ideal terms, for example, to enable the republishing of the data by uploading it to a geospatial data hub such as Koordinates. What I mean to say is that it doesn’t allow automatic republishing without having to clarify conditions of use with the agency, in this case Health. If it were released under Creative Commons, then this would greatly speed the republication and disitrbution of the data. So, whilst the NZDep2006 data has been released, it is really not yet ready in prime time for wider Web 2.0 use.
This has also been replicated with Health’s Atlas of New Zealand’s District Health Boards. Using pdf’s to provide spatial data is admirable, but in this day of Google Maps or Earth – surely Government should be considering publishing data as KMZ’s or even live network services that can be loaded dynamically into a far richer and intuitive client than Adobe Acrobat.
Coming back to my point, SSC is certainly making all the right noises about where we should be going. However, right now, it appears to be left up to individuals like myself to actually go to agencies and say – “Hey, we’d actually like access to this data in a more reasonable format”.
What I would like to see is a mechanism whereby individuals such as myself, can instead approach SSC with a request to an agency, and the SSC will actually engage said agency to ensure agencies make said data available in a consistent manner across government.
I’m sure agencies would stand up and listen to someone making a request on behald of the Government CIO instead of a lone citizen or three. It would also mean that a consistent playbook could be promoted (Government Geospatial Information Web Access Guideline) that includes formats, hosting and most importantly licensing agreements by encouraging widespread adoption of the Creative Commons v3 NZ license.
For Government agencies to really start opening up their data, this needs to be driven from within Government, and only the SSC has the voice to be able to catalyse this process. Sure, individuals such as myself are engaging with success, and in some cases we’ll be able to obtain access – such as my recent win with Transit’s 2008 road survey trackpoint data. But for an individual to engage multiple agencies is very time consuming, and it is a slow process – especially when we are often cold calling and have to restate our case for publicly accessible data every time. And honestly, it is not something we as volunteers should have to be doing, this is really work that paid government personnel should be doing.
SSC needs to short-circuit this process by stepping up and creating an inter-agency mechanism to accept requests from citizens, and use the position of the SSC to engage, promote and ensure release of the data – whether spatial or not.
Hi Gavin – I’ve left a comment, responding to yours, on In Development (http://blog.e.govt.nz/index.php/2008/07/24/what-does-web-20-mean-for-government/#comment-327).
Kind regards
Richard Best
Richard Best
5 Aug 08 at 17:50
For my records, I’ve copied my reply to Richard here as well as on In Development.
Hi Richard,
Thanks for the reply.
Firstly, you give great examples of what SSC is doing, and doing well! It is a pity however that many government agencies however are not able to keep up with the leading edge, and indeed, are only able to incorporate some new functionality into their websites when they see a major refresh every 3-5 years. I think in this regard, the SSC needs to be a bit more proactive in getting government agencies to modernise their web presence.
I’m interested in point 3 – the Guidelines on the Treatment of Intellectual Property Rights in ICT Contracts. At a quick skim, it would appear that the interest is solely on the commercialisation of intellectual property rights, and suggests little in terms of incorporating open source software as a potential solution (noting that option 1 would appear to allow the Customer Agency to own IP and release it under open source licensing if it so chooses).
An example that is close to my heart is Sahana, an open source web-based disaster management system that start in Sri Lanka following the tsunami. I am now on the Board and I am interested in finding Government funding for development of Sahana – given that Government agencies are likely to be the main users. I’m a firm believer that if a system can be developed for emergency management, and it can be funded by government, then we can actually provide this as a public good back to not only other users in New Zealand such as local government, but also contribute to a worldwide body of work. If I was to be cheeky, I’d suggest that this non-commercial IP option should also be presented as an option in the guidelines. Not everything needs to be commercialised – sometimes tax payer funds should support public good projects such as Sahana that is trying to provide powerful tools to manage disasters. And yes, wealthier countries with experience such as New Zealand should contribute this knowledge so that other countries, such as our neighbours in the South Pacific are able to benefit from our funding and knowledge. They can’t afford commercial solutions for disaster management. I digress
Re: Crown Copyright, Creative Commons et al – perhaps inconsistency between individuals licenses was not what I was referring to. Crown Copyright is quite a close neighbour to Creative Commons in terms of freedom. The point I was trying to make is that LINZ Crown Copyright licensing is different from Statistics New Zealand Crown Copyright is different from Ministry of Health Crown Copyright is different from New Zealand Fire Service Crown Copyright. Crown Copyright’s seem to differ in subtle and minor ways as defined by the agency that owns the information. This make it difficult to say take and republish this information, say by creating a data-warehouse, because Government isn’t doing this. the difficulty comes because as you said, there is a different bundle of rights with each piece of data that Government releases, subtle differences, but time consuming to work with, particularly in aggregation.
And yes, even Creative Commons is not perfect. There is a gap I have identified with geospatial data. With authoritative Crown data, there are some datasets that you don’t want people editing, so you could license it with CC No Derivatives. However, this goes a step to far, as this now disallows the conversion of the data into different formats, such as that required to convert shapefiles into use for handheld GPS receivers. The gap here is that No Derivatives needs to be clarified further, as some derivatives should be allowed (format-shifting) whilst other types of derivatives should not be allowed (editing/manipulation of underlying data) for certain authoritative Crown datasets. For example, one may be election boundaries – something where the underlying data should not be allowed, but format-shifting should certainly be allowed so that you can convert it to the application of your choice. But I don’t believe that the current No Derivatives license allows format shifting.
Having tens or hundreds of subtly different variants of Crown Copyright can create barriers to uptake of that Crown resource. This is less of an issue with end products such as reports and guidelines, but can be a major stumbling block for geospatial or statistics data.
As an example, I will email you the license attached to the New Zealand Fire Service ‘NZ Localities’ geospatial data that I recently licensed for our business from NZFS (for free). Whilst the terms and conditions outlined in the license are understandable, they do present barriers to more widespread use and adoption of the data. And naturally, as it is a custom license, it requires reading and understanding. I’m relatively comfortable reading licenses now, but not everyone is, and that is something that Creative Commons do well – they provide human-readable ‘translations’ of licenses.
I guess where I’m heading to here is that SSC may want to consider producing guidelines as to license options for the type of product they are releasing. We should not be encouraging agencies to write custom licenses and proliferating licenses with minor differences. From a public perspective, we need a limited number of clearly understood licenses with clearly communicated rights.
I am aware that LINZ and Statistics were happy licensing their topographical data under Creative Commons which made it easy to post their datasets to Koordinates. It is non-trivial to do this with custom licenses.
It is also good to see the work being done in the UK, however, as I noted when I looked at some of the data made available for the mashup competition, most of the data requires the uses of web API’s and detailed click-through licensing. Once again, these create further barriers to entry in the use of publicly-funded data.
Cheers Gavin
rediguana
7 Aug 08 at 23:05
Hi Gavin
Many thanks for raising these points. I’ll try to address each of them briefly in turn.
Modernisation of agencies’ web presence: hopefully the revised Web Standards will go some way towards achieving this, together with other guidelines or a revised Policy Framework for Government Held Information (in the works). I appreciate, however, that there is a divergence in licensing regimes across government and in the functionality and data formats offered. Without wishing to sound like a Pantene ad, I think change will happen, but probably not overnight (I’m probably aging myself a bit in making that remark).
IP Guidelines: Crown retention of new IP under an ICT contract is certainly still possible when retention is desirable in the interests of making that IP available to the public under an open source licence. As you note, the Guidelines expressly contemplate that. It’s also a point that was made when the Guidelines were released. The SSC press release quotes these words from Laurence Millar, GCIO: “Government agencies will also retain the right, where appropriate, to allow free use of the intellectual property on open source terms” (http://www.e.govt.nz/resources/news/2008/20080131.html). The Guidelines don’t explicitly refer to public good projects; perhaps they should. That’s something we can consider when the Guidelines are next updated. I’ve referred your comment to the person in charge of the Guidelines. One might also note that the Guidelines do not apply to “content that could be covered by the Policy Framework for Government Held Information”.
Differential licensing: I take your point about inconsistency between different copyright licences across government. It is something I have noticed when reviewing different governmental websites’ copyright statements. There is also sometimes a lack of clarity as to whether RSS feeds may be reused or not. While the copyright-related standards in the Web Standards are currently under review, as is the Policy Framework referred to above, I’m not sure we’ll ever achieve absolute consistency, as there may be reasons for some information and data sets being more liberally licensed than others. I do agree, however, that there is scope for better alignment and for a number of generally accepted licences for government use (whether Creative Commons, Creative Commons-modified or otherwise) so as to significantly reduce the differences and make life easier for those who wish to reuse and add value to government information.
Crown copyright and Creative Commons: Crown copyright is only close to Creative Commons to the extent that agencies, either pursuant to Government guidelines or mandates (such as the Web Standards) or on their own initiative, choose to grant a licence to material/information on terms approximating Creative Commons licences. In other words, it is the licensing of Crown copyright material on terms approximating Creative Commons terms that renders them close neighbours. Without such licensing, they are different beasts. Without licensing, reuse/copying etc of Crown copyright material would constitute a breach of copyright. Another way of looking at this is to say that it’s the approach to reuse taken in the Policy Framework and in the Web Standards that gets close to Creative Commons, not Crown copyright itself.
Gaps in Creative Commons licences: I agree that a given Creative Commons licence may not always be perfect/suitable. Without binding the Commission in any way (which I’m not entitled to do) I suggest we will need to consider the “No derivatives” point you make as part of the Policy Framework review.
Thanks again for your thoughts on this topic. Very helpful.
Richard
Richard Best
8 Aug 08 at 12:57