We are approaching the 30 June, and the open source community is again being asked to perform our annual miracle of delivering twelve months worth of software services in two. Yes, after living on crumbs from July to March, we're swamped with calls for help in April, and expected to deliver between May and June.
This would be all well and good if we could stock pile product, but our product is open source software, which we give away for free!
Yes, we are in the business of selling free software. And governments love building their systems on our free software. They've even written policies and guidelines on how to use it. You see, using open source reduces vendor lock in, which reduces financial and technical risk. It facilitates international collaboration, rapid development and rapid innovation, which makes it an enabler for the government's innovation agenda. And it is based on openness and transparency, a core tenement of open government initiatives.
But if we give away our software for free, what do governments want to pay us for? It is our services, our time: installing, maintaining, extending and supporting the software, and training people in its use. It is a specialised skill which takes years to develop. It is not practical to quickly ramp up and down software teams for a two month peak load. As explained by Brook's Law on software engineering, you can't use nine women to create a baby in one month. Likewise, throwing fresh developers at a delayed software project typically makes the project even later.
So as government investigate open government opportunities, I urge understanding and tackling some of the hard, root causes hindering open source adoption, such as flattening spending spikes out across the year.
No comments:
Post a Comment