The VIDaaS Project has now concluded its requirements-gathering interviews and conducted a national survey of researchers and IT staff engaged in supporting research. The time has therefore come to take our findings and translate them into technical requirements which our development team can get their teeth into. After going through all of the survey responses and interview notes, we managed to come up with a grand total of 81 separate requirements – enough to keep our programmers busy until well into 2013. Unfortunately, the project concludes on the 30th March 2012, so we’ve had to rationalise and prioritise the requirements to ensure that even if we cannot cover everything our users might want, we can at least get the really essential aspects of the Database as a Service software implemented.
Firstly, we took out any requirements which could be better dealt with via documentation and training rather than by the software itself, or which related more to the level of service we could offer than anything that could be measurably implemented by the technical team. Next, we merged a handful of requirements that specified different functionality but which could actually be provided via a single technical mechanism. Finally, we used the ‘MoSCoW’ system of prioritization to determine whether each requirement was a ‘must have’, ‘should have’, ‘could have’, or a ‘won’t have’ for the project. After comparing the user requirements (or in some cases aspirations), considering the functionality already in place, and ditching the ‘won’t haves’, we have managed to reduce the list to a more manageable 34 development tasks, 10 of which are regarded as truly essential to the success of the project.
Now it’s up to the DaaS development team to estimate how much team each piece of functionality will take to implement and test, and then we can reschedule the relevant work package in the Project Plan and get started on the new features, which we’ll be saying more about in due course.