









































194

Implementing an Open 
Source Catalog in a Consortial 
Environment
Evan E. Boyd, United Lutheran Seminary

ABSTR AC T The Eastern Cluster of Lutheran Seminaries (Lutheran 
Theological Southern Seminary and United Lutheran Seminary) 
have had a joint library catalog between their three libraries for over 
18 years. This has allowed for inter-campus lending, some shared 
purchases, and joint collection development. After using a shared 
instance of Voyager for 17 years, the Cluster knew it was time to 
upgrade. During the 2018–19 school year, the Cluster researched 
alternative software and ultimately selected the open source tool 
Koha, hosted by Equinox Open Library Initiative. This paper explains 
the reasons for change, the selection process, key decisions in the 
install process, lessons learned, and an overall assessment of the 
transition to Koha.

During the 2018–19 school year, United Lutheran Seminary and 
Lutheran Theological Southern Seminary of Lenoir-Rhyne Univer-
sity entered into a process to replace their shared integrated library 
system with a modern, currently-supported system that better met 
the two institutions’ needs. This process began with an assessment 
of current needs, discussions with vendors, and final selection of the 
best system to meet the Cluster’s needs: Koha with Equinox Open 
Library Initiative. This essay will walk the reader through the assess-
ment & selection process as well as some information about the 
installation and assessment of the new software’s ability to meet the 
Cluster’s needs.

BACKGROUND OF THE EASTERN CLUSTER

United Lutheran Seminary Library is a multi-campus library 
composed of the Krauth Memorial Branch Library, A. R. Wentz Branch 
Library, and the archives of Region 8 of the Evangelical Lutheran 
Church in America. ULS Library partners with Lutheran Theologi-



Listen and Learn Sessions  195

cal Southern Seminary of Lenoir-Rhyne University for a consortial 
catalog as well as intercampus delivery of books and articles. ULS 
and LTSS consider their collections a joint collection, or “one library 
under three roofs.” This partnership has been the longest-lasting 
program of the Eastern Cluster of Lutheran Seminaries (the Cluster). 
For more information on this history, I highly recommend reading 
David Wartluft (2002) as well as Karl Krueger (2007).

THE PROBLEM

When I arrived at ULS Library to manage the new institution’s two 
campus libraries, I immediately recognized that the Eastern Cluster 
Libraries were in need of either a new library catalog or more trained 
staff on the catalog. Key issues with the current system included:

• Voyager had essentially reached end of life. No major features 
were coming out and Ex Libris was trying to get everyone onto 
their newer product, Alma.

• Across three libraries, only one staff member was still fully 
trained on the software.1

• The software was inadequate to the needs of a 21st-century 
library: URLs in the OPAC were not permalinks; automatic 
logout from the staff client after 10 minutes of no use; a software 
client was required to operate staff functions.

• A central feature of the Cluster is inter-campus lending. 
However, patrons could not perform unmediated requests; a 
staff member had to manually enter their paper (or emailed) 
requests.

• Notices had to be produced manually and Voyager had no built-
in mailer daemon for sending those messages; separate mail 
server had to be provided by ULS’s IT staff.

• The OPAC was not mobile friendly, making on-the-fly lookups 
while in the book stacks frustrating.

The Cluster directors came to agreement that a change was needed. 
Initially, the Cluster approached Ex Libris to train all library staff on 
Voyager, hoping that more training would reveal ways the Cluster 
staff could better utilize the software. Ex Libris rebuffed the request, 
pointing to the manual rather than offering custom training. Ulti-
mately, they recommended replacing Voyager with Ex Libris Alma 



196  AT L A 2020 PROCEEDINGS

with Primo or Summon; they had said the costs would be similar to 
current costs.

NEW SYSTEM PRIORITIES

At this point, the directors knew they needed to look at what other 
options were available. The directors developed a priority list of 
needed features of the new system:

• Designed for consortial and multi-branch use. 
In the beginning, Gettysburg, Philadelphia, and Southern 
shared circulation policies. Over the years, multiple library 
directors’ varied interest in the “One Library/Three Roofs” 
model led to distinct policies being developed at each campus. 
While Gettysburg and Philadelphia now have matching policies, 
Southern still varies. Being able to allow for varied circulation 
rules continues to be important.

• Designed for inter-campus lending and “floating collec-
tions.” 
Under the “One Library/Three Roofs” model, patrons in Phila-
delphia are able to request materials from Gettysburg or South-
ern to be delivered to Philadelphia or delivered to their home. 
Not only that, but books sent to another campus are expected 
to stay there until requested back by their home library. This 
system has been successful over time but, because of the way 
Voyager was originally set up, this process was completely 
manual. Paper requests were accepted at the library front desk 
and circulation staff would message the lending library to have 
them mail the book on. In an ideal world, staff wanted patrons 
to be able to make unmediated requests, avoiding reliance on 
their own staff to submit the request.

• Must have a course reserves module.
• OPAC 

A web interface must be modern, easily-updatable, and custom-
izable by someone with a basic knowledge of HTML standards. 
One of the problems the Cluster had with Voyager was that 
libraries wanted to make e-books easily discoverable. However, 
Voyager’s OPAC wasn’t designed to have its interface easily 
updated. To make the slightest public interface change required 
changing multiple data tables on the server, rather than being 



Listen and Learn Sessions  197

easily manipulated with CSS or other common HTML standards 
anyone could master.

• Hosted by a reputable vendor with active development. 
Voyager was in use by the Cluster for almost eighteen years and 
LTSS first went live with Voyager earlier than that. The Cluster 
probably would have kept on with it longer, had it not been glar-
ingly obvious that Ex Libris was preparing to end support with it. 
This was hammered home for the Cluster when it had requested 
to purchase videoconference training sessions and was rebuffed 
by Ex Libris’s support team with links to the software manual. 
Just like with Voyager, the new software would be expected to 
be in use for a long time with regular updates. The new software 
had to be stable, with a large user base, and actively being devel-
oped.

• Easy-to-learn, web-based software. 
Voyager is not an easy piece of software to utilize, whether you 
are just a student worker checking out books or a full-time cata-
loguer. Because of this, the directors wanted software that could 
be learned fairly easily and was web-based, as the staff had 
several bad experiences with the Voyager desktop software. In 
addition, being web-based meant that staff could perform duties 
away from the office on any computer (a feature that has been 
very useful during the 2020 COVID-19 pandemic).

• Reduces overall costs.

REACHING OUT TO VENDORS & QUOTES

Once the team had this list ready, I made initial contacts with Ex 
Libris, Innovative, OCLC, and Equinox Open Library Initiative. 

Innovative informed us they were only selling Sierra and no 
longer offered their Virtua software, which, based on marketing, 
seemed a better fit for the Cluster. Some staff members’ personal 
experiences with Sierra were filled with the same issues as the expe-
rience with Voyager. Ex Libris and OCLC both gave presentations; 
the directors saw the many great advantages of their products (Alma 
and WMS) and were initially hopeful that quotes would be similar to 
current costs rather than an increase. Sadly, both vendors arrived at 
similar price points, wherein annual costs would have increased by 



198  AT L A 2020 PROCEEDINGS

fifty-six percent and the initial set up fee would be quite high—as 
much as the annual hosting fee. 

In my previous work experience, I had used a fork of Koha at 
Chicago Theological Seminary. Considering open source software 
was important to me, as I had a good experience with Koha, but I 
also wanted to see what else was out there. At this time, the FOLIO 
project did not have a complete project to demo so they were not 
considered. Both Koha and Evergreen and their various software-
as-a-service vendors were researched.

Equinox Open Library Initiative is both an open source software 
developer and hosted software vendor. They are known for having 
been created by the initial developer of Evergreen (Equinox Initia-
tive 2020), an open source integrated library system that was devel-
oped for use by library consortia from day one (Weber 2006). In the 
last few years, Equinox had branched out to hosting Koha, as well. 
The Cluster reached out to them, shared its needs, and they assessed 
that Koha best met those needs, as Evergreen did not yet have a 
course reserve module.

After a demo, quotes for annual hosting and installation were 
received. The Cluster members were shocked that Equinox’s annual 
hosting fee would be 10% of current costs with Ex Libris Voyager. 
In fact, the installation and data conversion costs would be roughly 
what was paid for Ex Libris Voyager each year. This meant the Clus-
ter could move forward without any significant impact on the Clus-
ter’s budget!
After final approval from the Cluster’s board, contracts were signed 
with Equinox in March 2019. In April, they were granted access 
to the data tables in Voyager. In May, the initial project manage-
ment meeting was held to go over major concerns and to take the 
staffs through the process of the various pieces of data needing to 
be reviewed. The new Eastern Cluster Catalog Online (ECCO) went 
live on August 22, 2019. Working with Equinox, the Cluster was able 
to get from the signing date to fully running in slightly less than six 
months. This was perfect, as the Ex Libris Voyager contract ended in 
October, which was just enough time for us to work through Koha 
and fix any problems.



Listen and Learn Sessions  199

GENERAL ISSUES WITH KOHA

Throughout this process, the Cluster knew there were some messy 
things in Voyager that needed to be resolved. In addition, some 
aspects of Koha were not fully understood until after either the test-
ing server was in place or the final go-live happened. Some of the 
major problems include:

• Too many item types. 
In the past, whenever a new shelving location was created, a 
corresponding item type was created. The Cluster simplified 
this, so that it could get by with as few item types as possible. 
This simplification makes updating circulation policies some-
what easier. While the Cluster did greatly reduce the number of 
item types, from 100 down to 23, it is possible a few more could 
have been eliminated.

• Permanent location – current location – shelving location. 
Koha displays current location information in an odd way. For 
instance, on a search result page, Koha only indicates if some-
thing is available and what branch it is at. This information 
lacks the most important detail: the shelving location, as each 
branch has a number of closed stacks. Patrons can get frus-
trated by this.

• Loss of the MFHD. 
Voyager uses the MARC Format for Holdings Data, or MFHD, as 
well as an item record for physical items. MFHDs can be used 
to create local holdings information, in particular periodical 
holdings information as well as series information. In addition, 
bibliographic records can be added to Voyager and are findable 
without item records but with MFHDs. Koha has a two-level 
structure: there is the bibliographic record and there is the item 
record. While bibliographic records can be searched without 
item records, one cannot narrow by owning library without an 
item being attached. 
     There are other related concerns we’re still working on, such 
as 5,000 records with no items attached, some of which are old 
acquisitions/order records for the old acquisitions data as well 
as bibliographic records for items that were withdrawn. 



200  AT L A 2020 PROCEEDINGS

• Item and branch info affects display and search. 
In order to narrow search results by format or location, Koha 
utilizes item data. Since there are no MFHDs to also provide 
that data, every bibliographic records needs an item attached, 
regardless if there is a physical item. In addition, ULS purchases 
e-books for its entire student population and not just for a single 
campus. These factors all affect the number of “branches” that 
exist in the catalog. In order to make electronic resources stick 
out, three additional library “branches” have been added.

 ˏ The Cluster eResources Branch was created for any open 
access e-resources added to the catalog.

 ˏ The ULS eResources Branch was created for any paid content 
ULS added to the catalog.

 ˏ Southern eResources Branch was created for South-
ern resources after it was noticed that ULS and Cluster 
e-resources are more visible than Southern’s, as the basic 
branch location designation in a search results page doesn’t 
cue the user in to the fact that something is an e-book.

LONGER-TERM ISSUES STILL BEING RESOLVED

While the Cluster has resolved most of the major issues since install-
ing Koha, there are some lingering issues: 

• Periodicals information. 
MFHD information for periodicals and other serial resources 
were not necessarily transferred correctly into Koha. In addi-
tion, the Cluster staff didn’t quite understand what relevant 
fields would be most easily seen in Koha, so some of the info 
that did transfer is hidden or shows up multiple times in the 
same record. Cody Swisher, public services librarian in Gettys-
burg, is slowly working his way through periodicals to update 
holdings information. Luckily, Koha has robust, easy-to-use 
reporting tools that have made identifying records in need of 
work simple. Those reports can be used to create downloads of 
problematic records that can be edited in MarcEdit and reup-
loaded, overlaying the problematic record.

• E-books owned by more than one library. 
Koha has a very easy-to-use batch upload tool. Unfortunately, 



Listen and Learn Sessions  201

that tool does not allow you to run record updates—that is, 
compare the current record and the incoming version of the 
record and only update the current record with new infor-
mation. This is mostly not an issue; however, it does mean 
that, currently, there are duplicate bibliographic records in 
the system, as ULS and Southern use different authentication 
systems. Eventually, staff will test and develop a process to 
download records needed, update in MarcEdit, and then reup-
load with both institutions’ 856 fields.

QUESTIONS FROM CONFERENCE ATTENDEES

Q: How did you handle duplicate entries while migrating your data?
A: The Cluster sub-contracted with Backstage Library Works through 
Equinox to have all records updated with current subject headings 
and added RDA elements. Record duplication was also checked for 
but there was very little.

Q: Would using Koha for your consortium work as well if you were not 
so closely integrated in circulation and cooperation?
A: I believe it would. Each branch can maintain its own circulation 
policies and any features around inter-campus sharing can be turned 
off. There was an alternative set-up where each branch would have 
a completely separate install, only sharing an OPAC, which we did 
not examine.

Q: How do you manage inventory between [campuses]? You mentioned 
books from one campus are held until requested back to the home 
library.
A: Koha has both “home location” and “current location” for item 
records. Home location does not change. It is easy to run a report 
looking for items that have a different home than current location.

Q: Have you integrated Koha into any other campus systems?
A: We have not.

Q: What is your biggest savings by moving to Koha?
A: The software and its hosting fees are a tenth of our previous costs. 
We also save significantly on staff time because of the easy-to-use 
automation.



202  AT L A 2020 PROCEEDINGS

Q: How do you manage electronic resources? Do you manage e-books 
in Koha or use something else?
A: Currently, all e-books in Koha are unlimited simultaneous users, 
so no e-check-out feature has been researched by the staff. E-books 
in Koha work best when item records are created. One challenge is 
that Koha’s bulk import tool does not allow you to add new elements 
to existing records, only replace. E-books held by both Southern and 
ULS are currently separate bibliographic records, but, in the future, 
we will likely use Koha’s export tool to identify duplicate records, edit 
in MarcEdit by adding both institutions’ URLs, and then re-import 
as replacement records.

ASSESSMENT

The Eastern Cluster Library has reduced costs, made staff training 
simpler, and is able to automate many formerly manual processes. 
Most problems the Cluster encountered during install and in the first 
year were due to not fully understanding how Koha structures and 
presents data, rather than a failure of Koha’s features. With some 
adaptation, the Cluster staff has learned the new integrated library 
system and users are generally happy with the new service. Over-
all, Koha has been a success for the Cluster, meeting all the Cluster’s 
key needs.

WORKS CITED

Krueger, Karl. 2007. “One Library under Three Roofs: The Exciting Sequel.” 
Atla Summary of Proceedings 61: 127–41.

Wartluft, David J. 2002. “One Library under Three Roofs: A Case Study.” 
Atla Summary of Proceedings 56: 207–17. 



Listen and Learn Sessions  203

ENDNOTES

1 Early on, many staff were well-trained on what was then 
Endeavor Voyager. After the first few years, the Cluster 
stopped sending staff to biennial trainings. Since the early 
days, many people had retired or resigned in the last 15 
years, taking with them a lot of the in-depth knowledge 
needed to keep the catalog running.


