Open Source in Government - Policy Framework

 

Notes from e-govt barcamp – Discussion on the policy framework around the use of Free and Open-Source Software (FOSS).

 

 

 

Policy Framework was split into three levels:

 

• Agency
• National
• International

 

Much of the discussion centred around agency level frameworks.

 

The chair (Don Christie) introduced the following concepts:

 

Low level agency centric policy

• Purchasing policy
• Legal
• Evaluation process

 

National Framework

• Legislative
• Protection of IP
• Standards
• Avoidance of capture, “paying an overseas tax”

 

International Framework

• Standards and interroperabilty
• NZ Role in the world
• ”Free” Trade
• Sovereignty

 

Awareness of broader benefits / Issues

• Import substitution
• Ability for bright NZers to trade their brilliance and foster the local software industry
• Integration point with global economy
• Threats

 

Agency Policy Framework - Discussion

 

A good conversation around SSC current policy recommendations. The main points noted were:

 

• SSC recommends FOSS should be considered on the same grounds as commercial software
• There was a discussion around the appropriateness of government contracts – in particular whether the global jurisdictional definitions of IP mitigate against smaller suppliers and perversely increase the attractiveness of overseas organisation suing
• SSC position on DRM is - don't use it
• Government agencies don't generally submit on legislative issues at Select Committee phase (e.g. the patent bill and Copyright bill)
• SSC influences and persuade, it cannot mandate
• MoJ has established some policy principles for FOSS adoption which can readily be fed into SSC advice or individual agency policy development

 

Other points made:

• There has been no economic impact analysis carried out on use of FOSS
• The Auditor General's office has visibly adopted FOSS solutions which could be regarded as a strong signal to other agencies in terms of the appropriateness of FOSS
• It was agreed that most IT support comes from local companies whether the software used is commercial or not. These companies generally also carry most of the contractual liability for ensuring the software works as they proposed
• Discussion around how commercial licencing models (e.g. per CPU, user, web hit) compromise the ability of enterprise architects to build the best technical solution for the job

 

Ministry of Justice representative Barry Polley spoke about a draft policy document the MoJ have been working on. Main points:

 

• MoJ will prefer to use FOSS. If no suitable FOSS solution exists it will use commercial software
• They spent some time identifying perceived weaknesses in FOSS and addressed all the objections that were raised
• One objection, that FOSS was flaky, does apply, but only in the context that all software can be flaky and part of the evaluation process includes a readiness for use criteria
• Legal risks are overstated. MoJ lawyers were unable to find a single instance, world wide, of an organisation adopting FOSS and being sued

 

 

 

The above process has already been used and results in a faster adoption cycle and a lower cost of purchase.

 

Suppliers will still carry the liability of making sure FOSS works in the way they have proposed (similar to a point above and the current situation with commercial software). The difference being that if necessary code fixes don't have to come from a single commercial supplier with little interest in NZ specific problems.

 

There was a discussion about Crown Ownership of tax payer developed systems. Crown Copyright invests ownership in the “crown” but this could also be compatible with releasing code under an Open Source licence such as the GPL. The latter option could see considerable return on investment in terms of functionality added by third parties which could then be used by NZ agencies.

 

Related to the above point it was noted that Australian Government Agencies often release systems under FOSS licences. This is known as “White Branding”. An example of this is the content management system, MySource Matrix currently used by Radio New Zealand.

 

Agency level policy discussion was by far the most detailed.

 

(As an aside I would note that a number of NZ Government agencies are making considerable and public use of FOSS. As well as the three mention there are the Ministry of Education (particularly TEC funded collaborative projects), MSD, National Library, Archives, Treasury, DoC, Ministry of Women's Affairs and Ministry for the Environment.)

 

National Policy Framework

 

The initial reaction from some government representatives was that they could not concern themselves with FOSS “philosophy” or issues that impact on its long term survival.

 

Others pointed out that this perception may need to change, in the same way that the MoJ is beginning to change its approach to the purchasing process.

 

The point being that the environment of the last 20 years under which FOSS has flourished is constantly changing. IP rules are changing, issues around DRM have an impact in that they allow commercial operations to exclude FOSS solution and therefore restrict consumer choice.

 

In an other example there are now opportunities for businesses to claim ownership of work through business and software patents that never existed when many FOSS licences and code were written.

 

This can have a detrimental impact on the economic benefits of using FOSS and also on the incentives to develop FOSS which would be a pity as it would occur just at the point in time that FOSS is having maximum impact.

 

There were a number of suggestions around this topic:

 

• Cabinate should include specific software and technology impact statements when the set out a “regulatory impact statements” on proposed legislation
• Government Agencies should give advice that considers the above points when advising Ministers on policy issues
• The MED should be far more aware of the impact on both the local commercial and FOSS industries when giving economic advice. There was a suggestion that MED is hard to influence and that there were multiple sub-agencies within MED that need to be more fully briefed.

 

There was also talk about the upcoming G2009 round of negotiations with Microsoft. However, this mainly centred around justification for the previous deals than any strategy for the next round.

 

 

International Policy Framework

 

There was no time to move the discussion onto this topic.

 

Conclusion

 

As the chair my optimism regarding Government Agencies' policies on software in general seems well founded. I am very encouraged by some of the cross agency communication and also by the deep thinking taking place.

 

This e-govt barcamp was one indicator of this, as was the GOVIS conference earlier in the year and some of the many initiatives I have come across over the last two to three years in particular.

 

I would like to see further work and research on the broader national and international context of what encourages software development and how best NZ developers can participate on a global scale.

 

So far I think NZ software companies have had to make do with rather bland platitudes and catch phrases (“The Knowledge Economy”). I think Government Agencies can have a role and influence in this.