Is the Repository Developer a dying breed, and should we care?
Cast your mind back, perhaps seven or eight years. It was the heyday of repository development. Projects such as DSpace and EPrints were taking off, and institutions around the world were watching the area closely and with excitement to see where this glorious new world would take us.
But back in those days repositories were similar to the early motor car – you needed a lot of money, several years, and your own mechanic/driver (developer) to make it work. Luckily, back then these resources were often available, money was perhaps a little easier to come by, and there were many funding opportunities from the likes of JISC to help out too.
As the repository developer worked with the repository software for a few years, they became intimately related with the software – they knew how it worked, how it was structured, what it could and couldn’t do, how to structure data within the repository, and often became key players in the development of the open source platforms by taking on roles such as DSpace Committership. Life was good, and I was lucky to be part of this, riding on the waves of e-theses, JISC projects (Repository Bridge, ROAD, RSP, Deposit Plait, SWORD) and the start of local institutional open access advocacy movements.
However… life moves on. The early repository developers have taken different career paths, and now find themselves in different situations.
- Some have left the domain when the project funded projects slowed down, repositories could be implemented without a dedicated developer, and new areas of interest arose.
- Some have progressed in their careers, and chosen to take a non-technical route up the tree.
- Some have taken a commercial route, choosing to take their skills into the commercial sector and providing development services back to repository-using institutions.
- Others have specialised as repository developers, but often find their emphasis has to be on compliance or marketing issues such as statistics, research assessment, or branding, rather than continuing to develop and apply core repository functionality.
It is rare to find a role these days where a developer can specialise in repositories, spending the majority of their time in that area.
I believe there is still a need for repository developers, as they bring many benefits such as:
- An understanding of the technology that helps them to know when repository technology can and should be applied, and when it should not. Often repositories do not appear to be suitable choices for some of our system requirements as we’re used to confining them to electronic theses and journal papers, but they have great potential in new areas.
- They know the underlying technology and data structures used by repositories, and how these can be mapped onto new domains. This can save institutions time and money, as they can re-use their existing repository infrastructure and expertise, rather than in investing in others.
- Equally and opposite, they know the weaknesses of repositories, and where current or future functionality will not be suitable.
- They provide technical credibility. Often repositories are run by libraries, but in environments where there are IT departments who may hold varying views on the technical development competence of the library.
- They make the running of the current repository/ies more smooth, and can help manipulate the data they contain (import / export / update / delete) in ways that are not supported by the native interfaces.
- Repositories are starting to become integration targets of enterprise systems such as CRIS systems. Having a repository develop around can make these integrations easier.
I think we’re seeing a downward spiral in the availability of repository developers. From time to time you see job adverts seeking experienced repository developers, and unfortunately they seem to be becoming a rare breed – a breed which I think we should protect, recognise, foster, and grow. I’m lucky to have worked in several large institutions lately where there have been small, effective, embedded, and valued repository development teams. However these types of teams are starting to become fewer and harder to find.
Opportunities for repository developers to network, learn, and share, are decreasing. There are exciting events such as Open Repositories with their Developers Challenge, and the Repository Fringe, but these are only annual events, and do not provide the opportunity for repository developers to show their skills and the potential of repository technologies to anyone outside of the repository community.
How will this affect the repository community? I think that there will be increasing problems in the repository world if repository developers become a very rare breed:
- The large open source platforms that we have come to rely upon (installations of platforms such as DSpace, EPrints, and Fedora number in the thousands) will find it harder to continue to develop and keep pace with current requirements. The large amount of development effort that has gone into these systems over the past decade could be wasted, and we’ll fail to see some of the benefits that only come to fruition after this length of maturity.
- There will be fewer exemplars of good practice of repository use to inspire and drive forward the innovative use of repositories.
- Those who administer repositories will lose their local allies who are able to provide the tools and integrations to make repositories a local success.
- The potential for repositories to be involved in new hot topics such as Research Data Management, the resurgence of interest in open access publishing, or the need for better digital preservation may be missed.
- It will be even harder to recruit experienced and passionate repository developers, and without well-established teams of these, new developers thrust into the arena will find it harder to grow their skills and knowledge.
What should we, and can we do about it? It think that we need to value the role of the repository developer, and continue to recognise that despite there being less requirement for a repository developer in order to run a repository, the absence of development skills may inhibit a good repository service from becoming a great repository service. As we value multi-talented systems librarians, we should value the repository developer as a multi-skilled employee that allows us to correctly apply and integrate repository technologies.
Looking around at commercial companies that offer repository development services (for example Cottage Labs and atmire) we see the sort of innovative thinking that has so much potential in this area, and when I talk to staff involved with these companies there seems no shortage of people wanting their skills. And this is good, and shows that there is a demand. But equally I feel we need to keep growing these skills within institutions, and not let the local Repository Developer become a dying breed.
Repositories are in their teenage years, we nurtured them through birth, messy childhoods, promising early years, and now we’re starting to get a glimpse of how they can become powerful embedded tools. But without the continued availability of skilled parents to shepherd their development, they may never reach their full adulthood potential.
[This blog post was written on the way to a DevCSI event for managers of developers, where we shall be looking at how we can show the positive impact of having local development teams within universities, and from my perspective and passion, their particular value to libraries.]