ILS Symposium

Marshall Breeding - The ILS: Past, Present & Future

Presentation by Marshall Breeding at the Symposium on the Future of Integrated Library Systems on September 13, 2007 in Champaign, IL, sponsored by Lincoln Trail Libraries System.

Presentations Available

Links to the ILS Symposium presenter files are now available.  To download the files, visit the ILS Symposium web site.  Presenter slides are linked from the Agenda page. 

Carl Grant Gives Us the Scoop on Open Source

Carl Grant, President of CARE Affiliates, gave us a wonderful presentation: Gone Open Yet? on open source software.

Grant gave us an introduction to open source software. He explained that open source software has its code widely available under a special license that allows users to change, improve and customize the software. It can then be redistributed to the community and shared to make the software better for everyone. It seems that most people who use computers use some sort of open source software, whether they realize it or not. Word Press, Open Office, FireFox, Apache, MySQL, Thunderbird, and Jabber are some of the most utilized open source software. As for libraries, open source is just beginning to rise in popularity. It seems like just in the past few years, we see open source as a viable option in libraries. The two most recognized open source names in the library field are probably Evergreen and Koha. Both are used by over 300 libraries. But what is it about these open source ILS? It seems that with all of the vendor consolidations and the commodity/infrastructure technology that people are naturally looking for different alternatives. Open source software are easier to procure, they are more reliable products, and there is no vendor lock in. They seem to have better functionality and are able to support a wide range of applications that are now available.

Since Grant came to the open source from the vendor world, he really has an understanding of both sides. Many vendors and skeptics will give libraries a list of disadvantages of open source. Grant tries to de-myth some of those.

Is open source really free?
-No, though the initial cost of a vendor ILS isn’t there, there are other costs associated with open source. Grant believes that in the long run open source will cost about the same as a commercial option, however, the open source ILS will be more of what you as the libraries want, and more of what your user expectations are.

What will it cost to add functionality?
-The great thing about open source is that adding functionality is easy. There may be a cost in staff time, but since all the code is open you can decided which function to add and what you can afford to add. In a vendor situation, you must just do without or purchase the package they have compiled and pay what they want you to pay.

Open source companies lack maturity and have no product road map.
-Grant says that this is absolutely not true, that open source companies have a much more defined road map than many of the vendors. The developers of open source love libraries and could be making more money elsewhere, but are devoted to making great software for libraries. They are not out for profit.

Is there really support?
-Open source is based on community. Most open source software companies have a large community website where developers share ideas. These sites have FAQs, wikis, blogs, and multiple facets for support. Since the code is open, there is often new features and code source on these websites too

Is open source viable over the long term?
-Open source software is no more or less viable than the commercial options. Other presentations have talked about vendor abandonment. Grant believes that because of the community aspect of open source, that there is less of a chance of fading away.

Grant concludes that open source has a very real, applicable place in libraries. It might not be the solution to every single library, but open source will really make the whole ILS market a better place.

Karen Schneider Talks to the Trustees

Karen Schneider, Research and Development Consultant, College Center for Library Automation (Florida) presented ILS Issues for Trustees on Saturday.

Schneider says that it is necessary to have librarians in the selection of the ILS process, but there are positives and negatives in that. Librarians understand the back-end operations better than anyone. Season librarians have years of working with users and know functional aspects needed in the ILS. The newer librarians are often savvy about user trends and behaviors outside library land. Schneider, however, has some concerns, because librarians can have staff centered views, which focus on staff functions. Sometimes there is a belief that users can and should be taught complex functions, but this shows a lack of familiarity with user behavior. Some librarians may not be familiar with concepts such as: standards, interoperability, or web services.

Having trustees being a part of the selection process can bring perspectives from other industries. Trustees don’t have to be library software experts; they just need to understand that libraries are different than other industries. It is important for trustees to realize that some vendors are predatory.  It is important to have a long range plan for technology, but if the plan is longer than 2 or 3 years (in great detail) then it’s a fairy tale. It is important to build in as much flexibility as possible. Trustees have to be thinking forward and don’t know what technologies will be available in the next 6 months. The must assume that there will be massive change in technology. It is important to serve today’s users as well as possible. It can be assumed that users’ expectations will rise faster than libraries can keep up. Existing technology prices will drop, but new stuff will come in to fill the gap. No matter how much you spend on technology, you’ve under spent by 50%. The library does not have enough technology staff and those staff members are paid far below market price, so it’s very important to treat them right.

Schneider suggests that trustees try to read some of the following resources in order to keep up with library technology trends:
    Library Journal
    Information Week
    American Libraries Direct
    Chronicle for Higher Ed.
    Gartner
    Pew Internet and American Life
    OCLC Reports

Some of the above sources will give trustees information about the realities of ILS market. Currently, merger and buyouts dominate the ILS market. There has been no new commercial product in the last decade. Nobody likes their commercial ILS vendor. There have been several cases when vendors have promised upgrades and then left libraries at the alter.

Schneider talks about what she has learned about users’ research. Most people make typos some of the time, and typos constitute about 10% of failed searches. Most users are searching with 2, 3, or 4 words without using Boolean. No one reads help screens anymore, and libraries are the only people who are using the advance search features. For users, the search is a hesitant, random process of discovery, and most people start their searches elsewhere (like Google.) Most of the legacy software now have turned our OPACs into “walled gardens.” There is no interconnecting them. The thing to remember is that users want to connect to the software.

A good ILS should be robust and reliable. It needs to supports core staff function in a manner that doesn’t lead to expensive work-arounds. It should provide a user experience that satisfies customer expectations. It’s very important to have modules that can be mixed and matched with external products needed to keep up with technology change. All that the user really wants is: ease of use, personalization, doablility, and connection. Last gen  OPACs are very poor at known item searches. They are very week at discovery research process. They don’t engage the user. When it comes to adding digital content, it ends up feeling like an artificial limb. These OPACs are very out of sync with user expectations. They are policy driven and have software limitations.

Some new enhancements in better OPACs that are happening now are ranking, spell check, search as a box, and improved digital content. Other options that people are adding to make the OPAC better are: faceting, editions, working lists, review/rate features, persistent links, unified search and best bets. This gives the OPAC a more personal feel and gives the user a more emotional attachment.

Schneider talks a little bit about open source options. She says one of the problems is that there is a lot of fluff talk to make you feel good and attached to the software, which isn’t necessarily what you want at first. Open source comes with a lot of fear, uncertainty and debt. There are many options out there and it’s hard to decide which one is the best for your library. Schneider suggests reading the Gartner reports. The option of commercial software often has many hidden costs. With commercial ILS software you can’t see, touch, discuss or change the software. Many of the problems get hidden behind the institution because of the model. The vendors tend to skimp the most when it comes to service and development and those are two important things with ILS. There is always the chance the vendor will go out of business or merge with another company, and this vendor abandonment can come at a very high cost. Open source software adds a lot of value. Open source causes better vendor competition and a healthier service orientation. It disrupts the ILS market in a way that benefits the consumer. There are actual fixes in open source software. Third party vendors can support open software, and it is a very wise model for libraries.

What the Studies Tell Us About the Future

On Friday night after dinner, George Needham, Vice President, Member Services, OCLC and Jasmine Gaia, OCLC gave a presentation about the OCLC studies and what they tell us about the future. The presentations were given virtually using iLinc and will be available in the near future on the ILS Symposium website.

Developments in the OPAC World: A Panel Discussion

The panel presentation moderated by Karen Schnieder, focused on developments in the OPAC World, The presenters were Cindi Holt, Information Services Manager, Phoenix Public Library, Kate Sheehan, Coordinator for Automation, Danbury (CT) Public Library, and Chip Nilges, OCLC. Each presenter discussed information access and cataloging platforms: Endeca, Library Thing for Libraries, and WorldCat Local, respectively.

Holt detailed a case study of using Endeca, a commercial search platform that provides guided navigation and allows customers to refine search. Large retailers such as Barnes and Noble, Home Depot and Nike as well as smaller companies interested in internal communication have successfully used Endeca. The staff of the Phoenix Library, aware of changing user expectations tested Endeca and Aquabrowser in an attempt to refine their goals. The staff sought an approachable website and catalog, an increase of speed and effectiveness when searching, a means of exposing the richness of MARC in a user-friendly way, to expose the depth of the collection and incorporate and extend the platform. Endeca works by harvesting MARC records from the bibliographic database and creates multiple indexes from the data nightly. Searches are performed against a dedicated Endeca Server and full records are retrieved from the database in real time. In order to implement Endeca Phoenix Public Library needed to define and refine dimensions and implement user-friendly guided navigation. Approaches to the challenges began with mapping Dewey Classification number to BISAC heading. The PPL staff had to devise a method to do this so cataloguers switched gears and broke the rules. A preview of the new system was set up on the old site with both environments available to staff. The customer response was positive. However, there are some librarian concerns such as where the LC subject headings go, why the display differs, real time of items, authority browsers, and the inaccuracy of data and practices. Holt offered some long-term solutions like having all new items have BISAC headings, excluding BISAC for audio-visual materials, and adding BISAC to fictions. Additionally, Holt laid out the next steps for the project, which includes continuing to enhance searching and browsing, developing web 2.0 strategies, and incorporating 3rd party data.

Similar to Holt, Kate Sheehan explored the possibilities of a system not intended for libraries or library system. Library Thing is a social catalog for booklovers. Regular people catalog their own books using tags they create. In addition, the site offers reviews, suggestions and widgets. There is a lot of data on the site. The Danbury Public Library implemented Library Thing for libraries. Library Thing for libraries is not the same as Library Thing. It takes the information from Library Thing and integrates it into the library’s catalog. Sheehan gives reasons to use Library Thing for libraries: First of all, it’s fun! It is a way to add innovative features, without putting out a lot of money. Most libraries don’t have a programming team, or access to the ILS code, and that isn’t necessary in this case. It is incredibly easy to integrate. Library Thing sends the javascript, copy, paste, and there you go. There were no major problems in Danbury’s implementation, and the one small kink was fixed by LT in an hour. The staff loves it, the patrons love it, it’s a great addition to the ILS.
***Kate Sheehan has a cool “techie librarian blog” check it out!

Chip Nilges gave us an overview of WorldCat Local (WCL). WorldCat Local is trying to deliver library searches in a manner that is easy for users to understand. It is important to encourage users to participate in network and community libraries, and reach out globally. User expectations are changing and WCL is trying to answer those expectations. WCL aggregate services around the world, discovers resources, delivers needed items, and allows for sharing and collaborating. It has configurable feathers, library branding, relevance ranking, interoperability with ILS, and includes Metadata in the local search. Each library has a custom search box, and a custom URL so that they can have true customization of the service. Once a search is conducted, it gives local holdings first, followed by consortia and then global. It shows you not only which library owns it, but also if it is available. Since it uses metadata, a single search can access multiple resources: the library catalogs, WorldCat databases, digitalized materials, and full text. The pilot libraries right now are: University of Washington, Peninsula Library System, The Ohio State University, University of California, and the state of Illinois. Release is planned for late 2007. Upcoming features are: institution search, identities interactions, and social functionality.

Karen Schneider Tells Us Why the OPAC Sucks

Karen Schneider is a Research and Development Consultant for the College Center for Library Automation in Florida. Her presentation was originally titled, The OPAC Sucks but has been changed to: Where We Stand, Late.

Schneider, in her presentation tells us why the OPAC sucks. In the past, OPACs have been enhanced with more library centered features, and not user-friendly features. This is what librarians asked for and this is what the vendors gave us. To really develop a good OPAC there are things we need to be aware about searching for materials. For one thing, most people make typos some of the time. A misspelling will not yield any search results. Today’s searches are two, three, or four words, and we aren’t talking about Boolean operators. Users’ searches are hesitant and are random discovery processes. People today are starting their searched elsewhere, like Google and Amazon. No one is using the help screens or the advance searches. Over all, users want to like your software, but libraries just aren’t providing that right now.

Last-gen OPACs are very poor at known item searches. They are weak when it comes to discovery searching and are pretty bad at user engagement. It is very hard to add digital content smoothly into these OPACs. These have become walled gardens, or silos, that you can’t branch out of or get into. This is very out of sync with the users’ expectation, and lacks emotional connection. It is necessary to provide better OPACs with features that meet the users’ expectations. Some new enhancements in better OPACs that are happening now are ranking, spell check, search as a box, and improved digital content. Other options that people are adding to make the OPAC better are: faceting, editions, working lists, review/rate features, persistent links, unified search and best bets.

Schneider gives four approaches to bettering libraries OPACs: Libraries can ditch the vendor’s OPAC and add a separate product onto their ILS. They can migrate to a whole new ILS with next gen features. Old OPAC can be replaced with a unified finding aid. Or they can just wait around for the vendors to catch up. The new concepts for these next gen OPACs are decoupled modules, user center designs, making the ILS a middle ware, union catalogs, leveraging other people’s data and data models, and using the open source.

Schneider talks about the values of open source software. Open source software actually fosters better vendor competition, and provides a healthier service. Because of the open source communities, actual fixes happen, and happen at a relatively fast rate. It disrupts the ILS market in a way that benefits the consumer

There are some problems we need to fix and some changes we need to make. There is a ranking problem, and with the surrogate record environment is hard to get ranking working well, but you should try. The surrogate record: a database record stands in for the object. In the digital world, there is no record, just the actual digital resource.

Should the ILS be a suite of tools for maintaining the collection with a different OPAC on top? There are issues with decoupling. Eventually all the applications will need to hook up. Decoupling increases maintenance at the library level. All this can leave to a scramble to develop interoperability standards that meet realistic needs.

Schneider concludes with some hard problems to solve before everyone can move to next gen OPACs. The current systems have aging, legacy data structures. The core taxonomy of the OPACs are complex and extensive. We still have record based data and need to think about record vs. content. What really needs to happen is libraries need to get in the mind of the user experience and develop from there.

PINES and Evergreen

Elizabeth Garcia, PINES Program Director, Georgia Public Library Service
Mike Rylander, Vice President, Research and Design, Equinox
Open Source – The Good, the Bad, and the Wonderful

Garcia started the presentation out with a brief introduction of PINES. PINES is the Georgia library consortium. It has 47 public library systems and 270 facilities. There is one card for basically the entire state. The patrons can check out books anywhere, put holds anywhere, and return the books anywhere. Patrons can request books from other libraries at no cost, however, new books aren’t intra-PINES loaned for 6 months. This has been a state wide service since October 2004 and they have had 540,000 intra-PINES loans. PINES has consistent policies across the state, there is one, easy to use interface, it is centrally administers, all automation costs are paid for member libraries, all the training is done through PINES and there is a centralized help desk. The approximate cost per patron is 1 dollar, which is 1/10 of the cost if there were individual library systems.

PINES had an ILS, but was coming up on their contract end. Staff conducted a survey of the library automation marketplace. They were trying to decided if the software was driving the policies and procedures, or if it was the other way around. PINES didn’t have anywhere to go, they had outgrown their ILS. Staff got together and figured out what they needed and set up focus groups. They decided to develop their own ILS and they named in Evergreen.

Evergreen development began in 2004. They wanted to include search capabilities like the popular websites, Google like spell check and search suggestions, added content like book covers and reviews. They needed Evergreen to be able to grow as the amount of libraries and collection grew. Mike Rylander helped to explain the core technologies of Evergreen. Everything they used was open source:
    Database: Postgresql
    Programming Languages: Perl and C
    Web server: Apache
    Operating System: Linux
    Server Hardware: x86-64
    Messaging Core: Jabber
    Client Side Software: XUL
Since all these are open source, it is very cost effective, as well as being reliable and flexible.

Of course in all of this they could have used more staff, more time and more money. There were initial loading challenges and the first day didn’t go as smoothly as planned, but there was a lot of things that were done right. Garcia talked about how they involved their libraries in the process and that the libraries feel they have ownership. There were multiple dry runs and staff from all over was involved with the testing.

Where is Evergreen going from here? They want to add more self-service options – like online bill pay. They look forward to cooperative projects with university system libraries, as well as partnerships with libraries world-wide. New acquisitions and serials modules are in development. It is important to develop resources to make Evergreen a more community project as  well as make a foundation and users’ group.

Carl Grant Gives Us the Vendor Perspective

Carl Grant is the President of CARE Affiliates

Grant’s presentation today was the vendor perspective of integrated library systems. He talks about the different kinds of people in the profession, where the problem areas are, where vendors are headed, and solutions to some of the problems.

Librarians have a certain reputation with vendors, but it isn’t as bad as you think. Grant says that librarians have adverse feelings to risk and change, they tend to be conservative because they are spending public funds, they are highly educated yet not adventurous, and usually use an RFP as a buying mechanism. There tends to be different kinds of personalities in libraries. The innovators are those tech enthusiastics who want all the new stuff and are always checking things out, unfortunately they don’t have a lot of money. There are the early adopters who are the revolutionaries, they embrace the new technologies and talk about how to implement them. However, they want the software highly customized. These two groups only make up 20% of the market. Early majority people want things to work efficiently and want to buy from the market leaders. They want safe and reliable software. The late majority are conservative and price sensitive. They are the skeptics, but are demanding. They don’t accept the technology until it is a commodity. The laggards are those who challenge everything and vendors don’t even try to sell to them.

Grant explains what the vendors have to go through during the process. The purchase process tends to be a 2 year process. When libraries send an RFP it costs the vendor about 5000 dollars. Multi-day demon cost another 5000, all these costs are going into the actual cost of the ILS cost. Vendors get about 1/3 of the RFPs they receive and this adds to the cost as well. Many libraries bring in consultants and they are brought in so that the responsibility is placed on someone else.

In many cases, libraries get the lowest common denominator that doesn’t piss anyone off. If you want to be on the edge you have to be daring and lead and choose software research the systems out there. Since there are so many different browsers, and systems and different things it’s hard to get the ILS to keep up to date.

Grant gives us his opinion on where he sees things going in the future. The OPAC will be going away, because there are just too many interfaces. The acquisitions model will also go away, because it will be mainstreamed. Circulation and interlibrary loan will become self-service modules, and users will need to pay for the convenience factor. The cataloging module will morph. Personalization will be added. It will be necessary to standardize workflow and partner with non-library companies.

Planning for the Future with Rob McGee

Rob McGee, RMG Consultants
Planning for the Future

McGee started with a very entertaining sermon-style history about the dark days of library automation. It was hysterical. The concept of the ILS was actually developed at the University of Chicago in the early 60s, the project was to develop an integrated bibliographic data processing systems for a large research library. Today’s ILS have been around for about 30 strong years, but there are many problems with the major enterprise. As for the future, McGee sees the need to go beyond bibliographic data, add metadata for e-resources, integrating the management of resources, managing all financial and human resources. We will see best of breed combinations vs. integrated systems. The ILS as a service is gaining popularity.

When thinking about the ILS it is important to have a strategic plan. You need to know what to spend you should spend your money on, and know what to consider in planning and procuring next generation integrated library systems. It is important to have a ten-year plan and have goals as well as know the problems that you are trying to fix. A lot of times there are conditions that trigger this planning. It is these conditions push people to open source, but it give libraries opportunity to do some good. There are many things to consider. It takes 2 years of planning to actually implement the change. It is necessary to do a gap analysis. Do you need an upgrade or a replacement?

There are possibilities and constraints to both upgrade and replacements. Upgrades happen within a given operating systems or compatible OS environments. The upgrades have to happen within certain hardware. It is fine to add a new interface to an old ISL. If you can change a few things and make them better, then you might not have to train as if you would if you were in a new ILS. If you decided to go with a new ILS you need to have a long range, comprehensive plan. There are three important questions to ask yourself: What are some of the technologies important to libraries? Which ones should my library implement? And Should my library partner with other libraries? You should think about integrating new technologies from outside and inside the library industry.

McGee concluded by showing the need for a multimedia management system. There is so much in the library land that we can add and use ILS as a solution. Technologies will be very key in this as well as collaboration between libraries and outside the library.