Of Open Source and the European Commission

At the end of last year I reported on the worrying signs of vacillation from the UK government over its support for truly open standards. At least it's relatively straightforward to keep tabs on what's happening in Blighty; Europe is another...


At the end of last year I reported on the worrying signs of vacillation from the UK government over its support for truly open standards. At least it's relatively straightforward to keep tabs on what's happening in Blighty; Europe is another matter – I find the labyrinthine bureaucracy and its digital shadow pretty hard to navigate. So I was pleased to come across the following page, entitled "Strategy for internal use of OSS at the EC".

For a start, it has this useful summary of the Commission's gradual adoption of open source solutions:

In December 2000, the European Commission (EC) defined a strategy concerning the internal use of OSS. It created at the time, for instance, the context for the recognition and use of the Apache Web Server as a recommended solution on UNIX systems.

In July 2003 a revised version of the OSS Strategy was presented to the Comité Technique Informatique (CTI), a committee gathering all Information Resource Managers, the persons responsible for IT in their Directorates General. The strategy was endorsed and a set of clear objectives was defined and approved by the CTI. This approach led to important achievements, the most prominent being the recommendation of Linux as Server OS, the use of Apache to power the europa.eu server and the use of open source software for the Commissioners' blogs and public forums on europa.eu.

Having met the initial objectives, the Commission further refined and revised its Strategy in line with its evolving organisational requirements while at the same time taking into consideration the worldwide developments in the field of OSS. The newly revised strategy was published in 2007 covering a timeline of four years (period 2007-2010). Within this timeframe a number of major achievements were accomplished in the Commission. Notably, the completion and formal approval of the European Union Public Licence (EUPL) is considered a milestone in the OSS domain. It is now used widely within public organisations as well as by the private sector.

Not quite sure about that last bit – is the EUPL that widely used? I don't think I've ever come across it used in anger anywhere, but maybe I've led a sheltered life.

There's also a good run-down of what exactly the European Commission is using currently:

The European Commission runs IT solutions on more than 350 Linux servers.

DIGIT's Data Centre manages more than 800 OSS web servers.

The Flexible Platform for Internet Services, project available under EUPL on OSOR.eu, offering a recommended set of Web 2.0 tools for social collaboration, is entirely powered by OSS tools; it provides, among others, 40 blogs for Commissioners, EC Representations and other EUROPA sites and hosts more than 400 wikis.

All new web applications at the European Commission are protected by an OSS-based solution for authentication, currently serving more than 300 existing web applications, more than 60 000 users and performing more than 1 000 0000 authentications on a yearly basis with more than 17 000 different users every day.

Several corporate solutions are entirely based upon OSS. Examples are in the area of content management, surveys, e-invoicing and e-ordering, etc. Within the Commission's IT network, an OSS-based developer collaboration platform hosts more than 770 projects accessed by over 3000 developers.

More than 60% of the information systems developed at the Commission are based on Java. All (100%) of these Java development projects include OSS tooling (e.g. to support the build process, for testing and quality assurance, for provision of core "runtime" functions such as frameworks for model-view controller, inversion of control, etc.). A corporate reference application eases the knowledge transfer to all development teams at the Commission.

An OSS browser is included in the desktop reference configuration available for all PCs at the Commission.

That's all good stuff, to be sure. But I'm less reassured by the accompanying "Strategy for the use of OSS at the EC for the period 2011-2013". The main elements are the following:

The Commission will continue to adopt formally (through the Product Management procedure) the use of OSS technologies and products where a clear benefit can be expected.

The Commission will consider OSS solutions alongside proprietary ones in IT procurements. Contracts will be awarded on a "value for money" basis.

For all future IT developments, the Commission shall promote the use of products that support recognised, well-documented standards. Interoperability is a critical issue for the Commission, and usage of well-established standards is a key factor to achieve it.

For the development of new information systems, where deployment is foreseen by parties outside of the EC infrastructure, OSS will be the preferred choice and in any case used whenever possible.

The first two are the usual wishy-washy commitments that could mean anything (how clear does a benefit have to be to become a "clear benefit"?), but at least open source gets a mention. And to be fair, the last point seems quite promising, with a clear preference being expressed for open source: that's good. But it will come as no surprise to long-suffering readers of this blog that it's the third point that causes my blood pressure to rise.

That's because it talks about "recognised, well-documented standards", not open ones. Which means that proprietary ones are fine if they meet those two criteria. This is part of a larger failure by the European Commission to stand up for truly open standards that might create a level playing field – manifest most painfully in the second, watered-down version of the European Interoperability Framework.

There's more of the same in the two accompanying documents, the "Conclusions of the discussion on Open Source Software and on the alignment of Open Source Software strategies" [.pdf] and "Conclusions on document exchange formats following the discussion on office automation platforms" [.pdf].

The first of these makes the following claims:

OSS products are not yet currently used widely in the Institutions in other areas, in particular:

in the collaboration and content management domain, where the OSS market is promising but very fragmented, although promising inroads have been made in areas such as the direct communication with citizens;

in the end-user desktop environment, where (with the exception of the web browsers) OSS solutions do not yet offer, as of today, all the functionality required to make them viable alternatives for organisations of the complexity of the Institutions and/or would require significant migration investments.

This seems curious. Content management is one of open source's strongest areas: it's not so much fragmented as rich and vibrant – there are lots of good alternatives. Equally, with the possible exception of calendaring, which remains a weakness in open source, I challenge anyone to find key functionality on the desktop that open source solutions lack. Free desktops can really do anything these days; the problem is simply that people think that the change will be too much trouble (in fact it won't, but human nature always imagines the worst.)

But the second document is even worse, since there we read that for "Interinstitutional exchanges":

As far as revisable document formats are concerned, XML-based international standards are the preferred approach. Given the fact that:

- on the one hand, OOXML is much more widely used than ODF at the moment, and this situation is not likely to evolve in the foreseeable future; and

- on the other hand, all the Institutions plan to migrate to office automation platforms which will produce XML files natively while providing 100% native support for legacy formats such as Office 2003, the preferred document exchange among the Institutions is OOXML.

This is an absurd, self-fulfilling prophecy, that says since OOXML is more widely used now, we'll keep using it because that's not likely to change - not least because we're not changing. It means that the European Commission remains locked in to Microsoft's products, and has almost zero leverage when it comes to negotiating prices.

That's because there is almost nothing that the European Commission could move to if it sticks with OOXML: Wikipedia lists just two one non-Microsoft applications that is capable of both reading and writing OOXML formats (one of which is OpenOffice/LibreOffice); this compares with dozens of applications that can handle ODF, and most of those support it natively.

This blind clinging to the past whatever the cost shows that for all the European Commission's history and fine words about open source adoption, when it comes to the crunch on the desktop, it can't be bothered. If it were, it would mandate ODF throughout the organisation – saving itself and us large sums of money, liberating itself from total dependency on a US company and giving a massive boost to the European software scene.

Follow me @glynmoody on Twitter or identi.ca, and on Google+

"Recommended For You"

Vista accused of breaching anti-trust laws Freiburg Goes Frei