September 16, 2008

Getting there

Filed under: Repositories — Neil Godfrey @ 9:01 pm

Golly gosh. No sooner do I go and start a new job at Murdoch University in Perth than another job with even greener grass beckons. Murdoch needed some help with getting their repository off the ground and I was free to move and liked the idea of helping out, so I’m really quite thrilled to find we’ve been able to make some real progress in my all too short stay here.

When I arrived the repository was simply broken. Perth is “the most isolated city in the world” and I wondered if that was partly to blame for their predicament despite all the bridging of distance technology is supposed bring us. I suppose most people who come to find work here are looking for the much bigger bucks offered by the mining companies. University libraries won’t compete with their offers.

Anyway, first job was to fix the technical side so we at least had a repository that worked, if not perfectly. By worked I mean that something that allowed a simple pdf file or jpeg image to be uploaded, with descriptive and rights etc metadata attached, and with a preservation and a Dublin Core datastream tossed in as well. And that that package was searchable and displayable.

Some of the broken bits that prevented even that level of progress turned out to be nothing more than not knowing where to find a particular window that was deviously hiding behind a larger one. Damn I hate computer software that shows no mercy for the neophytes!

Other issues took a bit more persistent and methodical diagnosis. Outside tech help had to be called in to discover an incorrect url namespace in a configuration. Thanks USQ Repository Services tech team! Another issue turned out to be related to a sneaky comma gatecrashing an alternate id value field. And the remaining issues are simple crosswalk or indexing matters.

All is now good technically. It works. And records look like they are supposed to look in the repository, or are in the final finishing touches of being so.

But there was one matter I had been postponing while focusing on getting the thing to actually work. I could never quite understand a certain discussion I heard regularly that related to publisher journal article links. It turned out that we were trying too hard to start with too much, and some directions travelled were not necessary at all. Not surprising, given that there had been so many handicaps — acute shortages in the IT department, limited skills required for crucial understanding¬† of other vital systems and procedures — faced by the library that held up the repository for so long.

I figured that trying to get everything in place before really starting the repository was only going to continue to hold up everything. Let’s just start crawling before getting ready to take our first walk. Don’t try to arrange to have 400 academics all come on board from the first weeks. Let’s just get 4 to start with. Even one will do. Suddenly we were working with real data. No longer trying to set up the whole she-bang before the get-go. And with real data, and with real academics who clearly were responsive to the idea of the repository, we could begin to see other fine-tuning that needed to be done. And we were responding to real needs as experienced by the real personalities and institutional climate of the institution.

Now we are crawling fast. Soon we will be able to walk — go into live production. And we will have a nice core of library and academic personnel enthusiastically working together to draw on and to build the core from which it will grow.

What we’ve had to discard is a lot of effort that was trying, I think, to set up too much before we started. And especially when I believe that much of that effort was really bypassing the immediate goals of the repository — open access to the university’s research. That was only delaying the start. Trying to bite off something too big, like a python struggling to gullet a cow. All that effort in trying to crush the beast to a manageable size before it can ever get anything near its throat. It will get there. But not very fast.

I’m excited about all the small things I’ve been able to get done so far, and to help set in train. I won’t be here long enough to see the repository fully bloom, but it’s great to think I’ll be leaving with things set on a positive course here.

Another issue related to the repository development has been the library’s relations with R&D here, or lack of such a relationship. Developing a libray research assistance strategy is soooo much easier with a repository thrown into the mix. Librarians can stop hitting their heads against the wall of trying to get academics to do things in ways that the librarians can see are better and offer some real goodies. Greater exposure that repositories have been demonstrated to bring. (And being in positions to be conduits for even further information goodies relating to national research databases and object reuse and exchange, etc.) A research repository has the potential, at least with the right strategies for building institutional support, to get libraries and R&D people and authors to collaborate for the benefit of all.

And one other thought I can’t escape from my experience here so far. All the progress in troubleshooting issues relating to the implementation of the proprietary repository have come from (a) consultation with the community of other users of the system; (b) tech support from USQ-Repository Services; and (c) my own testing and investigations. In the process of compiling a report on repository options for the library and university it became increasingly clear that those in this industry who say in relation to open source software repository solutions, “You get what you pay for. I.e. nothing”, are unaware of the realities of opensource repository development and supports. But that’s another topic entirely.


%d bloggers like this: