Scott,

Depending on the primary technologies you are using at your institution, you could look at implementing the open source solution SobekCM. ( http://sobek.ufl.edu ) We are using this at our institution ( http://ufdc.ufl.edu/ufir ). We developed the system in-house over the last six-seven years and it works quite well as our IR.

It runs on Windows servers, with a Microsoft SQL DB and Solr backend. You can use the free Microsoft Express SQL if you like, assuming you don't reach 100,000 items in your IR. I am able to roll out a new instance within a day usually, but would obviously take slightly longer to learn the system and how to do that. The system is coded primarily in C# (with the obvious javascript thrown in as well) and we are actively pursuing two grants that will beef the IR functionality up even more; one to add a local authority system (for people, places, etc..) to assist with linked data and increased support for datasets.

It works with Shibboleth authentication, LDAP through Active Directory, or you can use a built in authentication system, which obviously doesn't give your faculty single signon.

We welcome active participation in the code development as well, so if you wanted to contribute back to the code base at GitHub we would love it.

Mark


Mark V Sullivan
Digital Development and Web Coordinator
Technology and Support Services
George A. Smathers Libraries
University of Florida
352-273-2907 (office)
352-682-9692 (mobile)
[log in to unmask]




-----Original Message-----
From: Code for Libraries [mailto:[log in to unmask]] On Behalf Of scott bacon
Sent: Tuesday, September 17, 2013 9:05 AM
To: [log in to unmask]
Subject: [CODE4LIB] Institutional Repository Costs

It may be a fool's errand to ask how much it would cost to implement an open source institutional repository, but here goes!


Let's first focus on open source and say that there won't be vendor costs for ingesting or downloading materials, that we already have our own purchased servers dedicated to the IR, our own digitization program (scanners and staff), and that we have already tallied cloud-based storage and preservation costs.


I have heard that the costs of implementing an open source IR can be as cheap as how much employee time is dedicated to the project. So you have a programmer spend a year or so on implementation and hire a librarian to manage it after that, let's say. Beyond that, are there any hard-and-fast costs associated with getting an IR up and running?


I have also read somewhere that it costs just as much to implement an open source IR as it does a proprietary one, but we'd certainly like to reap the benefits of having ultimate control over our own system if at all possible.


By the way, if it helps, my institution is classified *Master's S*, with an undergraduate enrollment of just under 10,000.


Numbers will vary wildly of course, but if anyone could give an idea of the cost of any component of a project like this, open source or proprietary, it would be most helpful.


Thanks,

Scott Bacon