So, the Hungarian government has launched a policy to use half of the IT budget on open source software. Effective in a few weeks, which in the alternate universe of public administration is equivalent to yesterday.

The more innocent corners of the open source community have been making much of this and expecting great mileage; it almost feels like a scaled down version of when Munich announced their open source intentions.

It's an extremely drastic move (and grossly miscalculated move were it even genuine), but what is really behind it?

In 2008, the Hungarian Competition Office took the Hungarian Public Procurement Authority to court over an EU tender to the amount of €100 million for the supply of "Microsoft or equivalent solutions". Hungarian open source providers of course complained to the European Commission. It's an open and shut case as the Hungarian public sector must issue an EU tender for a contract of that size, and they are explicitly prohibited from biasing the tender to a specific supplier.
(source: Rick Harris http://www.flickr.com/photos/rickharris/)
I wouldn't imagine for very long that this tender is a sign that doe-eyed Hungarian civil servants have finally seen the open source light. If anything, it looks more like a back-room deal wit the EC to avoid trouble for the new EU member. Corruption was a recurring theme in the progression of Hungary towards EU accession, and corruption continues to merit close monitoring in Hungary today. A policy instituting open source-friendly procurement terms is the equivalent of a bad boob job and some bright lipstick.

The deputy state secretary for IT and e-Government also said that the government would re-open the Software Competence Center, a body intended to help public institutions adopt open source. This center was closed back in 2005. It's probably a more effective public relations tool right now though.

We were asked to innovate, so we present: the broken clock that is NEVER right!

There is no overlooking that this policy has as intended effect something similar to affirmative action. The more distributed and internally competitive nature of open source is one reason that open source ISVs are not of Microsoft or Novell's scale; the historical market inertia of proprietary software (which was first-to-market in enterprises) is another. Both of these factors disadvantage open source in the enterprise public sector, and it is apparently this which motivates this policy of IT affirmative action.

Where this analogy diverges however is not in the intended effect but in the causes of the inequality which motivated the policy. Affirmative action in its social context bases itself on the assumption of parity in capability across ethnicity.(by Sappymoosetree: http://www.flickr.com/photos/bahkubean/) I think it would be wildly incorrect to think there is any form of parity between open source and proprietary software, however. Much of open source is radically better quality-wise than proprietary alternatives, and there are also instances of the opposite. Consider this: how many examples can you think of where there is an open source and a proprietary package which fill the same function to the same degree of competency? It doesn't happen.

A policy which by definition equates open source systems with proprietary ones disadvantages the greater body of open source technology by inviting the misconception that open source systems are drop-in replacements for the familiar proprietary systems. They aren't; to varying degrees, the systems that do replace specific proprietary products will usually require a very extensive migration. Secondly, the body of open source software is not a carbon copy of the body of proprietary software. You can't assume that Proprietary System X has an alternative in open source, and in fact the more innovative open systems will be disadvantaged since there are no existing proprietary systems for them to replace. Larger bureaucracies are used to thinking in narrow and discrete shrink-wrapped packages, and when they need to spend 50% of their budget on something else they will look for whatever appears most similar to their proprietary IT vocabulary. This shuts out a lot of the best systems opn source has to offer.

Now where'd I leave my crystal ball...

I foresee a rush of unnecessary and counter-productive purchase of open source systems in the Hungarian public sector. The civil servants in procurement are used to thinking in terms of discrete system functions (e.g. CMS, office suite, enterprise service bus) at best and specific package names at worst (SharePoint, Oracle, SAP). What has really happened here is that the government has halved their IT purchasing budget; they do not know the options available, the options available are not one-to-one mapped against what systems they do know, and the Hungarian open source suppliers are not large and sophisticated enough to effectively address the entirety of the awareness gap in the Hungarian state sector.

The re-opening of the Software Competence Center is insufficient and tardy; Hungarian agencies will have to implement the new (by Whinsical Cris: http://www.flickr.com/photos/treacletart/)policy within weeks of the announcement (which was made on the second of April 2009), and lord only knows when the Software Competence Center could be up and running again, never mind having a broad impact on public IT procurement decision making.

This has two consequences. The procurement offices won't really know what to do with the "other" half of their budget, and they will either mis-invest or not use it. Secondly, the procurement functions at the state level will be unhappy, and when it's the people sitting on the money that are unhappy that policies change back the quickest.

Further adding to the likelihood that this policy will be overturned very quickly is the fact that both Microsoft and Novell were mentioned by name in the minister's press release. This 50% policy is at least partially intended as a bargaining tool against the Hungarian states largest vendors, and it is far from the first time open source has been used as a threat rather than an option.

If the Hungarian public IT market were my God Sim

If the vendors are smart they will realise that this is a limited window of opportunity to sink their open hooks into the Hungarian state and act accordingly, and optimally in concert. The small to medium-size ISVs would each pick a very limited band of systems and present their shop and the products as their specialisation. Vendor A would be the case management systems and workflows integrator with Alfresco and JBoss, Vendor B would focus on desktop management and migrations. Vendor C could offer intranet solutions, and so on.

As argued before, there is a good chance the policy will be overturned sooner rather than later so this specialisation would help establishing footholds in the state from where the ISVs could work from inside to alter procurement expectations of software vendors, such as opening minds to source code availability and free markets for support and customisation.

(by d**: http://www.flickr.com/photos/d-/)This form of specialisation makes sense under any circumstances for small to medium size ISVs; the vendors limited size means that in practise their skill portfolio is limited. ISVs in this size band tend to make the classic mistake of trying to be everything to everyone, and this form of specialisation could in fact help mature them so they can grow more sustainably. I can think of a few examples of this kind of over-reach by small open source shops even here in Denmark.

Dear Hungary,

Thanks, but the rest of us will try to find more sensible ways to get open source in the public sector. Feel free to ask advice before acting next time if you're genuine.