Open Source doesn't matter

I recently had the pleasure of presenting at the British Computer Society, as part of a panel discussion on the challenges of web and document management in the public sector, and the ways in which open source software might be suitable for addressing them. Given that my presentation only contained pictures, I’ve summarised the key points from my presentation below in lieu of uploading it.

The key message in the presentation was that Open Source doesn’t matter. To over-simplify the point, since when has being able to see the source code of an application been a key factor in deciding what platform to choose?


Selection of a product is all about whether it’s the right fit for your organisation, and this involves many factors, including functional fit, total cost of ownership, support model, quality and availability of services. What weight you put on these will depend on your specific needs and type of organisation, but given the relative maturity of Open Source in the ECM space (particularly in WCM), it seems just as ill advised to discount Open Source entirely as it does to blindly insist on it over commercial software without some form of deeper analysis.

Public Sector organisations have a number of different drivers for WCM and EDRM (e.g. delivering better interactions with citizens, improving inter and intra department collaboration, keeping better track of information to comply with directives such as the Freedom of Information Act, all against the backdrop of needing to keep costs down and improve efficiency). However, in the UK at least, there is still a relatively low level of adoption of Open Source solutions for EDRM and WCM, compared to commercial products.

To explore why this might be a little further, we looked at what a stereotypical public sector IT or procurement manager might initially think when presented with Open Source as an option. The following opinions (exaggerated for effect) may run through their mind:

• It's risky : Did anyone ever get fired for picking a leading commercial product? Who’s going to assure my design if the whole product is community developed? Clearly Open Source carries a different type of risk to adopting a commercial solution, but that doesn’t mean that those who have understood the risks well haven’t been able to deliver projects efficiently.
• I can't get it supported: With some research, the ecosystem that surrounds each Open Source project you may be considering can be understood well enough to decide whether it works for you or not. One potential challenge here is if your organisation outsources its application support and maintenance to another organisation, and whether that organisation could readily support Open Source technologies for the same cost as COTS products.
• It doesn't fit my strategy: If your IT Strategy is to always consider products from a particular vendor first due to licensing and support arrangements you have in place, or to always go for .NET/Java/COBOL products, then this obviously significantly limits the options available to you. The most important thing is whether any of the products that align to your strategy meet your business requirements – if they don’t, then it may be that your strategy needs a rethink.
• It's insecure: Greater visibility of issues doesn’t mean less security – and the ‘many eyes’ argument has been rolled out a number of times in support of Open Source projects.  There are obviously aspects to consider here, such as how often are security patches issued, and how quickly could you deploy them once issued, but these apply equally to commercial and Open Source alike.

Assuming you’ve managed to address these concerns at a sufficiently early stage in the process, you may then encounter some challenges when trying to run a balanced procurement that includes both Open Source and commercial products:

• Open Source doesn't currently get considered equally: It can be harder to find complete, detailed information about Open Source projects to compare with the glossy sheets you receive from vendors, and, conversely, it’s also quite easy to discover reports about issues that might put you off (as vendors are quite good at hiding these). Consider getting integrators or subject matter experts involved in the process so that they can provide insight and balance into the process.
• Commercial vendors explain things in business terms: One of the main advantages commercial vendors have is their sales team, who can clearly explain how their product will address your specific business challenge. Open Source projects are getting better at this (e.g. through including client testimonials on their sites), but don't yet have the vertical solutions that vendors offer. This could potentially be a key consideration in an assessment. Again, using SME’s can help out here by providing a balanced view of how long it might take to make an equivalent of the vertical solution in an Open Source project.
• Too much / too little choice, even within products: There are not many open source EDRM products out there, particularly ones with certification. WCMS, on the other hand, has a very large number of independent projects, and even more ways of configuring them. Translating all this information will take time and effort, but it is always time well spent.
• Flexibility of frameworks: If you are following public sector procurement frameworks in the UK, are they sufficiently flexible to allow you to consider Open Source on an even keel? When constructing tenders for use within frameworks, the desire to look at Open Source as well as commercial products needs to be highlighted, so that potential suppliers know you’re considering this as an option.

We finished off by reiterating that Open Source doesn’t matter – and it certainly isn’t the key to making public sector projects more successful. Most issues on projects are related to the usual mix of people, management and politics, and these are rarely resolved by making the source code of your application visible.

A lively debate followed the presentations, with topics ranging from whether Open Source puts you in a better position with regards to IP to whether being able to adjust the core of the software yourself is a better situation than dealing with a sometimes reluctant vendor to address your problem. The answer to a lot of these questions was the classic Consultant answer (as Janus put it) ‘it depends’. The one thing everyone seemed to agree on was that Open Source generally likes Open Standards, but the two things are different (you can get commercial products that support Open Standards, and you can get Open Source products that don’t).

The point that interested me the most was in relation to how the selection of Open Source would affect the procurement process in public sector, and the question of whether a procurement process was even required given that Open Source doesn’t have a license fee. Some further research is required on this point, and I would imagine that this would very much depend on how your relevant procurement departments viewed the situation, and some specifics like whether you required support and maintenance or configuration services as part of the project (at which point a procurement might be required, but for those support/config services rather than the product).

The event also gave me the opportunity to make the acquaintance of a number of people in the industry whom I haven’t met before, including Jon Marks and Janus Boye, who have also posted their views of the day.

Many thanks to the BCS for giving me the opportunity to present, and I welcome your views on the topics raised.