State Library of Ohio ILS Options meeting minutes
Open Source Meeting
State Library of Ohio Boardroom
April 24, 2007
10:30 a.m. – 12:30 p.m.
Welcome and reminder of why we are meeting
Jo Budler, State Librarian, welcomed everyone and thanked them for their participation. The meeting was called in response to a request from a librarian who wanted the State Library to pull a group of libraries together who were affected by the SirsiDynix announcement regarding Horizon 8.0 to talk about other options. The invitation was expanded to include any library interested in the possibility of open source integrated library systems (ILS).
Jo indicated that the libraries in Ohio are really the drivers of the plan and input is needed now and in the future from the library community, especially given the impact on statewide resource sharing.
Open Source: an overview and presentation of pros and cons
Jo then introduced Stephen Hedges, Executive Director of OPLIN, to give a background on open source and open source ILS.
Open source is software that the source code is made available to the public in order to manipulate or change the code to make it do what you want. It is a different concept in how you treat property. Rather than following the property model of ownership and exclusion, open source follows the property model of stewardship and distribution.
The earliest days of open source were at IBM during the 1950’s when they distributed operating systems in source format. In the 1970’s, the notion of software as property to be bought and sold was foreign. However, in 1976, Bill Gates wrote the “Open Letter to Hobbyists” indicating that software, like his, should be bought and sold, not shared. There was a backlash and in 1985, the Free Software Foundation was created in the first organized attempt to define and support free software distribution. However, the term “free” was misleading. The software still has a cost in that you have to install it, configure it, and make code changes to suit your needs. Open source became the preferred terminology in the late 1990’s. Open source is not anti-business. Companies can sell the service of installing, configuring, and customizing, just not the product (code) itself. Red Hat is the best example, selling the service for supporting Linux, an open source operating system. Stephen suggested The Success of Open Source by Steven Weber for those interested in additional reading on the open source movement.
Open source is ultimately a production process. Code is created to fulfill a purpose and is expanded by adding modules and bug-tracking systems. When the community is large enough, development happens very quickly. If a community is not large enough, it becomes stagnant.
The process is self-selective. There needs to be an interest. There are “playground” rules so that the code does not go off in two different directions. People are in charge of certain aspects of the production process. Development is controlled by expertise and money. If a customer wants a certain feature developed, they will put money into it. Mozilla Firefox is an open source web browser. You download it and it installs itself and you can use it immediately. This happened because there was a sufficient customer base.
Librarians share a similar concept of property with open source, that of stewardship and distribution. Recently the two have joined forces. There is only a small community of librarians that can code. While the need is there, the expertise is not.
Currently, there are two major open source software ILS currently in production: Evergreen and Koha.
Evergreen
Evergreen is the software for Georgia PINES (Public Information Network for Electronic Services). There is one card and one automation system for 265 libraries in 127 counties. PINES created Evergreen. It has been in production for 2 years and debuted on September 5, 2006. The website is www.open-ils.org. Modules created are circulation, cataloging, statistics, and it supports self-checkout. They are working on the acquisitions and serials modules. They wanted OpenNCIP but have SIP2. The documentation for the process is very poor, which is typical because it is the last thing developers are thinking about. Equinox is the company that supports the Evergreen software, www.esilibrary.com. Version 1.0.3.1 was released on March 15, 2007. The University of Windsor recently joined Georgia PINES in development of Evergreen.
What is different about Evergreen is that it was put in place by Georgia PINES from the top down, instead of from a grassroots effort. The source code is open but they don’t really follow the “playground” rules for the open source production process.
Koha
Koha is a more mature open source project, http://www.koha.org/. Version 1 was released in 2000 and version 3 will be released in 5-6 months. Koha follows the traditional open source rules and the documentation page is pretty good, about 2-3 steps behind. Over 300 libraries around the world in dozens of languages are using Koha. Koha recently released Zoom, which handles major searches efficiently (for example search for “It” for Stephen King’s It or “O” for “O: the Oprah Magazine”). Koha has fully developed serials, acquisitions, cataloging, and statistics modules. Koha can handle Z39.50. Koha does not have NCIP developed. They are still waiting for a customer to put money into development of NCIP. LibLime is the vendor that supports the Koha software, http://liblime.com.
Currently, two libraries in Ohio are using Koha: Nelsonville Public Library and Stow-Munroe Falls Public Library. Barberton Public Library will go live in June.
Thoughts/suggestions regarding options
The SirsiDynix announcement regarding Horizon is indicative of the strange ILS market for business in Ohio. There are a fixed number of libraries in Ohio and no growth in the market. The vendors wait for migration to make their money. Historically, libraries need the vendors even though the business model is unusual.
Questions & Comments
Why did Georgia PINES create Evergreen rather than developing Koha? They didn’t go with Koha because it didn’t have experience with a large consortium.
How many developers are working on each project?
Evergreen had 6-8 original developers. Not sure how many active developers they currently have. Koha has 25-30 active developers.
How do you maintain the grassroots effort but with structure?
PINES already had one card for all residents and is mostly made up of small and medium libraries. The user-base must agree what it should look like. Expertise and money drives the development.
Sirsi states that the company is taking the best features of Horizon and putting these into Unicorn to create Rome. Rome will have a much bigger client, bookmobile functionality, database activity, and consortium customization.
Why did Stow-Munroe Falls choose an open source ILS?
The director of the Stow-Monroe Falls Pubic Library answered that they spent 2 years looking at new systems. The search committee evaluated 6 vendors and looked mainly at the financials and the stability of the companies. The committee narrowed the choices down to 2 vendors. In the meantime, the library hired a new systems person and just before deciding to go with one of the vendors, decided to go with open source ILS. This was mostly because vendors charge several hundred thousand dollars every 5-7 years for upgrades. The amount of savings realized and the success of PINES showed that the decision was not a gamble.
What where some of the challenges you faced?
One challenge was getting support and they got LibLime. The data migration was top-notch and the service was excellent. They felt good about the code. The biggest challenge was the concern of training staff, however, they learned very quickly. And if changes did need to be made, it didn’t cost an arm and a leg because they were affordable.
The amount of resources put into Koha, (time, money, & people) allowed them to tailor to their needs and wasn’t as expensive as migrating to Innovative (III).
Open source ILS are working great and it will pressure vendors to open up the code for their proprietary software. The open ILS are good systems but do not have the current capabilities of the current proprietary software.
What is the time commitment to bring an open source ILS to commercial level of use/functionality? 3-4 years?
It took 6 months – 1 year to develop Koha Zoom. Sometimes development takes a few months. Larger developments usually take 6 months – 1 year. The development of PINES took 2 years.
When you use a vendor, you don’t know exactly where your money is going. Using open source ILS gives you more control over where your money goes – into developing what customizations/features you want.
Schools in Ohio are largely automated with Sirsi. We should look the same for students using public libraries. Biggest concern is being responsive and timely. We must be nimble.
What additional features of Koha do libraries want?
Put the product on the screen and they will tell you. A list of requirements for users should be determined.
Library doesn’t want catalog to look like schools – wants their search to look like Google and Amazon. Some libraries in the audience stated that they really liked Koha Zoom.
Could a 3rd party vendor integrate ILS with other systems?
Should be easy using open source since it is open. Koha integrates with some.
What about small libraries with no IT staff?
An annual contract fee is based on the number of records. A majority of LibLime’s customers worldwide do not have IT staff. They offer top to bottom service, not just the software. LibLime charges for services on the software: data migration, annual support, customization and development, etc. LibLime employs 10-15 full-time employees and does hire contract workers as needed.
It is important to have a support/regional person on staff with access to servers at all times. Koha needs a fully functional acquisition module and consortia support.
The NEO-RLS Board just authorized NEO-RLS tech staff to support member libraries exploring open source ILS.
So, what one library or group of libraries pays to develop, then another can just take it?
There are always ways to cooperatively fund an open source ILS project. The more money that is put together, the more development and customization there is. Georgia would like to work with Ohio. Other state libraries have expressed interest in collaborating with Ohio.
LibLime is the only company providing support for Koha. What if they fold?
Because the code is open and not owned by LibLime, the code is still available and can be taken over and modified by someone else.
The State Library is not interested in imposing something from the top down, like how OhioLINK was created. There would be local control allowing libraries to make the decisions about what is the best fit for the individual library or group of libraries.
Is the State Library willing to give money to help libraries do data migration to an open source ILS?
There is a competitive LSTA grant program to which libraries can apply. It requires a local match of 25%. Libraries cannot migrate to stand-alone systems and should follow standards like NCIP. The State Library Board has made a commitment to funding library automation.
How compatible is Koha to Ohio Libraries Share: MORE?
Nelsonville Public Library is on Koha and members of OLS MORE. They have had no problems. NCIP is still needed to be less labor intensive for Ohio Libraries Share: MORE. Koha is working on NCIP compatibility.
Should we replace Ohio Libraries Share: MORE with an open source ILS?
The State Library is not wedded to Fretwell/Downing as a company. The State Library has to go before the Controlling Board on a yearly basis to get funding approved. However, the State Library is committed to having a statewide resource sharing mechanism in place. So yes, the State Library should and will consider an open source solution.
Should we re-ask if we are using the correct model for statewide resource sharing? Who are the most important libraries to share with? What are the best combinations of libraries to work together to fill the needs of our patrons and how do we make that work? Is there an open source solution?
Those are political questions, not technology questions. Patrons want what they want when they want it. Delivery is more of an issue. With the increase in usage, the cost of delivery will increase. Statewide funding issues will be discussed at the May 21st Library Partnership Summit.
Identification of possible next steps
Organize libraries to get together and evaluate features of LibLime and determine what is there and what is missing.
Koha libraries work together and apply for an LSTA grant to develop NCIP.
Form a research committee of trusted constituents from around the state to:
- Determine what the needs of public libraries are
- Collect and distill comments
- Develop a list of requirements
- Interview vendors and support vendors
Libraries that have and like commercial systems should be asked what commercial systems offer that the open source solution would need.
Bring in a consultant to gather data about the functionality and analyze the needs of the libraries. Could the State Library fund the consultant?
Collaborate with libraries on the same system to go collectively with one voice and say “we’re going to open source if you won’t meet our collective needs”.
Use SEO’s RFP as a base document to get started.
Keep the two-way discussions going. Preferred methods of communication: listserv, blog
All libraries interested in attending a meeting in May about moving to Koha, contact Debbie O’Connor.
Jo closed the meeting thanking everyone for their interest and participation. She was surprised at the amount of interest but thinks that it is indicative of the changes that need to be made regarding ILS and statewide resource sharing. She said that minutes would be distributed to attendees and that additional communication would follow.
Contact Information
For questions, please contact:
Jo Budler by email or by phone at, 614–644–6843.
or,
David Namiotka by email or by phone at, 614–728–4988.