Open source and ILSs Bob Molyneux Rowan Public Library Salisbury, North Carolina July 30, 2009

Open source dates: Koha – 2000 Nelsonville 2002 Evergreen – 2004 PINES 2006

What happened on March 13, 2007?

Here the market divides 1. Traditional ILS vendors Terms: “legacy” “proprietary” Two types: Founders still around Founders have sold out to VCs

1. Traditional Vendors (continued)‏ Characteristics: Compiled code (machine readable code) ‏ Intellectual property laws apply They own the code and you rent it. Advantages “turn key” comprehensive solution to many problems documentation and support

1. Traditional Vendors (continued)‏ Disadvantages Slow development cycle One size may not fit all Who owns your data? Restrictive licenses Forced migrations vendor lockin

2. Open source

Reading assignment: Eric S. Raymond: The Cathedral & the Bazaar

What does “open source” mean? “A set of principles and practices about how to write software the most important of which is that the source code is openly available....”[additionally] “...one should have the right to use it.” Wikipedia, “Open source”

What does “source code” mean? “human readable” What does “compiled code” mean? machine readable

Advantages of the open source method Easy customization for your own local situation Fast development - “release early, release often” Cost—it's free. can have about the same support as proprietary software

Disadvantages of open source It's free but it may not be cheap Support—if you can't, who supports it?

What happens if your ILS won't do something you need? Proprietary Wait until the next version Open source Do it yourself (or with others in the community) ‏ Either pay someone yourself or get others to contribute Cajole, persuade, or charm someone into doing it

State of the OSS US public library market about 1-2%,give or take no figures for the academic market but they are assuredly lower

But there is a new wind blowing: Evergreen Indiana Open Source ILS Initiative Sitka Michigan Library Consortium Koha MassCat INCOLSA WALDO

OSS opens the way for us to change libraries and their interaction with our users Evergreen offers one way

ILSs reflect their beginnings Evergreen started on PINES Large, resource-sharing consortium with a single catalog Universal borrower card

What have we learned from the PINES experience? Library users LIKE access to the large virtual library They don't care about our politics or the difficulties under the hood They will bypass libraries without access to consortial resources in favor of libraries with that access Welcome to the long-tail, Google world

Silos Separate, barely communicating collections of information Logic of IT is to break down silos and to integrate these collections We have these persistent silos for three reasons: Legacy vendor's lack of vision Our lack of vision and/or politics Think locally, act locally Until now, no software to run these large consortia

Traditional cost model Library ALibrary BLibrary C Module X $$$ Module Y $$ Module Z $

Open Source Cost Model Library ALibrary BLibrary C Module X $ XX Module Y X $ X Module Z $ XX

Evergreen design parameters Had to handle PINES 40+ PL systems, 250+ outlets 14+ million circs on a statewide resource-sharing network and scale up from there

Architecture Open Service Resource Format (OpenSRF) ‏ Service oriented architecture Modern, modular, scalable The only ILS software that can currently run large and distributed resource-sharing networks.

FullfILLment (tm) ‏ Attempt to get past the silos to one dynamic, real time search mechanism Evergreen backend, with opportunistic connectors to legacy vendor software

What other futures are possible in an open source world? Modular—not one size fits all More open source choices, perhaps more specialized Index Data VU Find Blacklight Mix and match Why not open source and proprietary? Code sharing

Eric S. Raymond: The Cathedral & the Bazaar Thanks... Bob Molyneux OPEN-ILS