I have posted here before about work at ADFI that may be of interest to ANDS.

After a recent visit to Canberra I have another couple of ideas presented here in a very quick blog post.

1 An implementation fund for CAIRSS

Currently we run CAIRSS as a support service, but as some of our discussions with stakeholders move along I foresee a time when people will start to say stuff like “we need a normalizing OAI-PMH proxy” (people other than me that is). We could have a mini project fund administered by the CAIRSS steering committee or a subcommittee which could dispense funds on the basis of short proposals. $1,000,000 over two years would be fine.

2 A developer engagement project

I have been trying to put my finger on what is lacking in the spread of projects we have now; ANDS, ARCS, NeAT and the outcomes of previous things like our own ICE-RS project and the ARCHER and DART projects. There's lots of good stuff going on but it feels to me like the community is somewhat fragmented by discipline, or by repository people vs grid computing people or along other dimensions.

So how about a program to do something like we did on RUBRIC but for software developers instead of repository managers - and for the new role of eResearch Analysts around the country. RUBRIC was Regional Universities Building Research Infrastructure Collaboratively. This would be Something-or-other Building Research Infrastructure Collaboratively. The acronym we can leave to Dr Treloar.

I am basing this suggestion on my experience of what JISC is doing to build professional networks amongst their developer community by running competitions alongside conferences, hosting loosely structured meetings like the develop happiness days and so on. So why don't we hire a few people; one to organize events and communications and one or more experienced open source developers with good communication skills (can we clone David Flanders or entice him out here?) who can promote:

1. Community. Getting developers in touch with each other and with the implementers, the eResearch analysts. For example we could really use the help of some semantic web people on our work on The Fascinator Desktop, do I know who to ask? (Well, actually yes I do, but that's just an example).

2. Harmony. Getting teams working in similar areas to work together and share rather than re-inventing.

3. Openness. Encourage and assist projects to work with their code under an open trunk from day zero so that others can join in at any stage and offer assistance, rather than working in a closed environment and then (maybe) releasing the code after the project has finished and it is too late for others to offer useful assistance. I won't name names, but some publicly funded software that should be open is not and we have had to rewrite it.

4. Friendly competition. I was forced by David Flanders to enter the developer happiness days competition and made two new connections (hi Mia, hi Ian), and I met Anna Gerber from down the road at UQ partly because I was a judge on the Developer Challenge at OR09 and now I will take my semantic web questions to her.

   Via Les Carr comes the idea of giving developers hard disks full of assorted data and documents and saying 'what can you do with that lot'. We could do that at eResearch Australasia 2009, mix up data from the humanities and sciences to encourage discipline specialists to work with each other and throw in some documents to get the repository people involved.

I have not worked out a budget but to have decent events, travel and guests from overseas this might need $1,000,000 or more a year.