On an open registry of academic developers

Posted on January 27, 2011


As a community, we are pretty terrible at self-promotion. In fact, many of those coding and scripting as an academic staff member struggle to identify themselves as developers in the first place.

We are developers, whether we are writing code required for a funded project, adjusting webpage templates or even doing analysis of library usage data with excel. We all have a set of skills, and I have yet to meet someone who truthfully knows it all, although many claim to whilst deriding some skills and describing these as being ‘beneath them’.

What is simple and obvious to some, is impregnable to others without help and often, the reverse is true. What is difficult for the first group is often mundane to the second. It was the main principle of Dev8D, that developers can teach other developers and gain training themselves in the process.

What has become apparent is that discovery of who can help you is a difficult problem. Not as difficult as making people realise the skills they have are not as trivial or easyto acquire as they might think. Finding someone with the skills you need who also has some understanding of the academic world is another difficult, and currently, a underestimated property.

So what can we do? Do we really on word of mouth to find people who do good work or have the experience we are looking for, whether we want to ask them a question, to gewt them to speak or write on a topic or even pay them to fix a problem for us?

An inhouse developer with niche and desirable skills is a very valuable and crucial person and yet on project bids and reports they are often left unnamed.

The JISC fund many projects which have at their core at least one developer. If the project is worthwhile, it is inevitable that the develpers have gained some useful and potentially unique experience. And yet, the data held by JISC in their PIMS database typically names just the project’s PI.

And then there are all the internal projects that take place – from the can-you-fix-this? problems to the we-just-bought-this-software-now-make-it-run types!

Fundementally, I am striving for a community that is based around the academic context and not around a single product, foundation or project. Who can I ask about the realities of pushing data into $VENDOR product? or people who have experience in working with latex and webpages? In my personal network, I have people to call on but its not an easy query otherwise.

I propose that starting with a registry of projects – bootstrapped with data from pims.jisc.ac.uk and EPSRC and so on – where people are encouraged to put up both major projects and also internal projects. Experience has shown me that no matter how minor the hack or script, someone will be interested and enlightened by it.

Then allow developers to claim projects, to tag it with links to blog posts, code or even just the project outputs. The web will quickly sort the poor descriptions from the good. The important things to link up are the people and the subjects.

What about LinkedIn and others? Im dont find it useful or part of my life oline. Most importantly, it’s a walled garden and I can’t remembera time when a link to one of these sites appeared prominently in any search result.

And StackOverflow? Better, much more findable and you can certainly build up good, reliable kudos on the site. I still think there needs to be more to it to help us find people within the right context.

I do not wish to reinvent any wheels so would like feedback on this. I do think that starting with an open, combined data set of projects, taggable and attributable, is a great way to start!

Posted in: Uncategorized