Caveat – I work for the Government Digital Service but this post is entirely my own opinions.
Like a huge number of other organisations around the world, the British government has benefited greatly by using open source software. Previously government IT was defined by such things as: huge scale outsourcing, very long contracts, waterfall development, very expensive and time consuming updates live systems, focus on internal rather than user needs and heavy dependencies on the likes of Microsoft and Oracle. In the new ‘digital’ (definition still up for debate) era we now expect in-house agile development, frequent releasing, user research and the extensive use of free-to-use open source software.
The use of products such as various Linux implementations, Apache, Nginx, Varnish, Squid, Postgres, MomgoDB and Piwik – to name just a few – has saved government very significant costs in license fees.
The government Technology Code of Practice states that departments must “Improve transparency and accountability by giving equal consideration to free or open source software when you choose technology“.
The GDS Service Manual quotes the Government IT Strategy as saying “Where appropriate, government will procure open source solutions. When used in conjunction with compulsory open standards, open source presents significant opportunities for the design and delivery of interoperable solutions.”
Digital teams across the government estate talk about why contributing to open source is important. From the very beginning GDS has always talked about “coding in the open” and the Service Standard still has a criteria for making all new source code open.
This has been achieved, in spades, with the GOV.UK / Government as a Platform / etc repository and the GDS Operations repository being just two examples.
In addition to brand new code, government developers also contribute code updates to a variety of external open source projects.
The value we put on open source is reflected in the fact that we’re currently looking to employ a full time Open Source Lead.
All of this is fantastic – but I’m been pondering recently if it’s enough. Being part of the global open source community is great, but does it reflect how much government has saved, and continues to save, by using such products? GDS quotes figures of savings of well over one billion pounds since it inception. A non-trivial part of that must come from not paying for software licenses that we no-longer need.
I’ve been wondering whether government should reflect this, and our continued dependence on such products, by contributing financially to long running and highly respected open source organisations. I’m not suggesting that each department should work out which products they use and have to come up with a figure accordingly. I’m not even suggesting a rough estimate of usage across the board. Instead, I’m reflecting on a central Treasury funded pot, allocated through GDS, that represents the relationship government owes to producers of open source software.
Obviously this is an unnecessary cost to the taxpayer – we could continue to use the products for nothing – but an initial spend of, say, £100K is less than the cost of three developers for a year and we’re currently taking advantage of many times that number of people contributing to such projects.
I don’t think it’s a hard requirement, but I think it’s an interesting conversation about what debt we feel we own the open source community and what kind of relationship, beyond using and contributing code, we want to have with it.