LESSONS LEARNED: An Academic
Library's Conversion from DOS and CD-ROM to Windows 95 and Online Access
Francie C. Davis, M.S.L.S.
Electronic Resources Librarian
Joyce Renfroe Gotsch,
Information Instruction Librarian
Lori Kim Troboy, Ph.D.
of CIS and Management
of the Project
Dowling College is a 43-year-old, liberal
arts institution located on the south shore of Long Island. It serves 6,000
nontraditional, primarily commuting undergraduate and graduate students
and specializes in education, business, aviation and transportation.
A major project was initiated in October
1995 to move the Library from dependence on CD-ROM versions of databases
to Web-based access where possible. Limited space and finances suggested
that an appropriate strategy for our Library would be a heavy reliance
on electronic resources. Because we had a relatively small collection with
limited depth, we sought access to more full-text information. This conversion
was intended to provide much faster access to more up-to-date information
and to improve service to Library patrons. The project did not come to
fruition until January 1998 and required fortitude, patience, and humor.
This paper will examine the sociological, technological and managerial
AND DESCRIPTION OF PROJECT
Prior to this project, the Library
housed 13 public access 386-based microcomputers connected to a LAN networking
22 CD-ROM databases and the online library catalog. Since the project's
implementation, the LAN has been upgraded to Pentium-based computers workstations
with access to an expanded networked tower of CD-ROMs and with Web access
to the Library's homepage and online databases.
Most of the databases were in CD-ROM
format and needed to be updated frequently. Because our outsourced Computer
Services Department performed this update irregularly, we deemed it necessary
to seek an alternate solution that would better serve our customers and
provide more up-to-date information.
We upgraded our CD-ROM tower for more
speed and space allowing greater campus-wide access to materials that did
not need frequent updates. This upgrade provided additional new space that
permitted us to convert periodical subscriptions to electronic CD-ROM forms,
eliminating the time consuming management of paper subscriptions. We also
converted from access to CD-ROM programs at one stand-alone station to
campus-wide network access.
We saw a golden opportunity when major
vendors announced that they would no longer support DOS. This was a compelling
argument to update our resources and convince administration that we must
convert from DOS to Windows versions of our databases and migrate to Windows
95 / Office 97 in the Library.
The Library's organizational chart
is unique. There is no library director. Seven full-time, faculty librarians
share professional library management decisions. An additional 10 part-time
librarians assist in staffing the reference desk, teaching Information
Instruction, conducting research and suggesting specific service improvements,
but are not part of the Library's management team. All facility management
and clerical staff management fall under the auspices of a non-librarian
Information Services Director who holds an MBA. The group of librarians
and the Information Services Director report jointly and equally to the
Outsourcing is a significant feature
in the background of the project. The entire Computer Services Department
is outsourced to one company, which manages both hardware and software
systems for the College. A completely separate company hosts and manages
the Web site for the College. Part of the challenge in this project involved
communications between the College and these two companies.
During the course of this conversion
project, key personnel changes occurred. The Associate Provost/College
Information Services Administrator, the immediate boss for librarians,
resigned to take a position elsewhere. The outsourced Computer Services
Director left and was replaced. The outsourced Computer Services Department
project manager quit. The college-wide web master resigned, as did a long
string of students hired to update the web page.
Lastly, the library budget was hit
by major budget cuts. An initial budget cut of 15% of the previous year's
allotment was followed by an additional 40% cut. These budget cuts had
a significant impact on the speed of implementation of the project and
increased the necessity of moving to electronic sources.
This section lists the lessons we learned
during the course of the project. Taking a sociotechnical perspective,
the lessons fall into three categories: social, managerial, and technical.
The social lessons revolve around using
clear and frequent communication to overcome resistance to change and to
invite everyone to participate enthusiastically.
It is vital to enlist support for the
project from every level. You will need it. It is important to understand
that we all have different mental models. You will encounter that when
you are trying to enlist the support of everyone involved. Some will be
enthusiastic. Some will be obstructive. You will need to help them appreciate
your vision. Be patient. Remember that while you may have been mulling
this conversion over for some time and have been eating and breathing it
for months, they have not.
Get your peers involved at the ground
level. They need to buy into the project if you want it to fly. You
will need their involvement in many ways. The project will impact on them
and the way that they interact with your customers. If you provide an Information
Instruction program, it will have to be completely revised because of your
project. If you do reference work or research for your customers, your
project will impact on your staff's efficiency because all of the interfaces
will have been changed. Remember that if there are budget cuts that are
necessary, it is entirely possible that it is your peers' budgets that
will be impacted. So be patient and be sure to keep them informed.
Be sure to keep your administrators
informed about your project. They are much more apt to support it if
they have been consulted and kept apprised of the benefits of the conversion.
You need to prove that your plan will benefit the library and your organization.
It will need to be cost effective, as well. You will need their financial
support as well as their influence with the powers that be.
Be careful whose advice you take. Know
your resources, both people and reference. One of our goals has been to
provide electronic library resources campus-wide. When we were having difficulty
with providing access to the multiple formats we have to deal with, one
person said to just say, "It's only available in the library." To do that
would have been breaking a trust and going 5 steps backward. That was not
an option. That was not good advice for our situation.
The managerial lessons include using
good communication skills but also suggest that good planning and thorough
training is essential to the success of a project like this.
The project needs to be justified
to the administration. They need to understand that it will benefit
the entire community served. In planning the project, we noted that with
funds being tight, we would have to cancel the subscriptions to journals
that were available from our new full-text source. We would also have to
cancel redundant databases, in other words, databases that indexed the
same journals as our full-text source. We budgeted for new hardware to
run our system and upgrading our CD-ROM tower. What we did not count on
was a budget cut of 15% of, not the proposed budget but of the smaller
budget for the previous year. This smaller budget was later cut by an additional
60%. These budget cuts made the decision to cancel redundant subscriptions
very easy. If the project was to be at all viable, this was a good place
to start. Next we had to attack our individual budgets. This was where
cooperation among colleagues proved extremely helpful. Everyone wanted
the project to succeed. Everyone cut her budget. We determined to lease
the computers and to cut the multi-user licenses for databases where we
could. Once we proved to the Provost that everyone was behind the project,
we got his approval.
Let your Computer Services
Department know what you are investigating. You will need their advice
on hardware specifications and their technical expertise when it comes
to making the shift. Despite what we think, migrating from DOS to Windows
and CD-ROM to Web access is not a simple process. It can be a time-consuming,
tedious endeavor and it is not you whose time you are volunteering for
the project. It is the Computer Services Department's. So keep them informed.
If you plan access from your
web page, it would be wise to get control of your web page as soon as you
can. Whoever is working on it should understand your goals and understand
that your web page is a dynamic unit. You need to be in sync. When there
is an update to be made, you need it to be done in a timely fashion.
Create a plan with clear
lines of authority and areas of responsibility. Define milestones and
remember to adjust the plan as you go. This is vital. There may be key
personnel changes that you have not anticipated and they can seriously
impact your project. After we had worked towards our "conversion" for two
years, the Chief Information Officer, an enthusiastic supporter of the
plan, resigned. We then had to report directly to the Provost. This created
the need to convince the Provost that we had a viable plan and that it
was fiscally responsible. That meant the creation of a new proposal and,
ultimately, a Technical Plan for Information Services. In addition to that
key personnel change, we also had the director of the outsourced Computer
Services Department quit. The Computer Services Department project manager
for our project quit right after having signed off on a half re-built but
not yet working CD-ROM tower. The web master quit. And there were a total
of 8 different students working on our web page, each of whom needed time-consuming
orientation to the idiosyncrasies of the page.
Preview all software and
ask the vendors about authentication, display, licensing, and costs. Authentication
refers to how you will limit access to the database. Will it be by password
or by IP address? If it is by password, how will you control access to
that password? How will you distribute it? Can you provide dial-in-access?
What will the display be? Will it be just the items to which you subscribe
or will it be everything that they offer with that particular database?
If it is not limited to your subscription, what happens when a customer
tries to access that part? Will you have a disgruntled customer or will
you get a surprise bill? Will the display be what you have been used to?
Can you control the default settings? Or will it take a two month battle
with the vendor to have it changed to your specifications? What is the
cost of the number of user licenses that you will need? In most cases converting
from CD-ROM to Web access was no more expensive. There were, however, instances
that the number of user licenses available was reduced. For instance, Wilson
CD-ROMs provided for unlimited licensing. Web access to the same databases
is more costly and unlimited licensing was not available. SilverPlatter
costs were the same whether online or CD-ROM.
Training is a vital part
of the project. It will contribute to its ultimate success. This must
be planned from almost the beginning. You must remember to train the trainers.
We have already provided many in-service sessions for the librarians but
plan to do more. They may be very comfortable with the old resources but
the new ones have very different interfaces despite the fact that the content
is the same. They need help in understanding the best way to approach the
databases now. Remember that everyone has a different learning style. Some
are visual learners. Some are auditory and some learn by doing. Make allowances
for each style. And be patient.
Handouts and manuals will
have to be prepared. Your Information Instruction program will have
to be revised. Faculty need to be approached and educated. We volunteered
to teach one of the faculty colloquia. We have been scheduled to present
at the next Administrative Council. This spring we will offer campus-wide
workshops. We have learned that it is more successful to offer sessions
by category rather than to expect that the groups will mix. Therefore,
we will offer some sessions for faculty, some for staff, etc. We speak
about our new program at every opportunity both to inform and drum up support.
Plan for the transition.
It will probably
be your most difficult time during the project. Even though you have been
warning people that the new system is coming, they never believe you until
it is actually here. Keep in mind that the CD-ROMs will be gone from use
while your technical person has to map them. Until the whole system is
up and running, you will still have some of the old and some of the new.
Prepare your staff. Let them know what is happening. Give them a chart
of what programs will be eliminated; what will now be available on the
network; what will now be online. Let them know what will now have full-text
information. We kept referring to the project as the "full-text project"
which created the misconception that everything would be full-text.
While you are in transition, you may keep the old CD-ROMs and use them
on the network even though you have started using the new Web access. Just
remember to let everyone know that the CD-ROM disks are out of date. That
may be enough of an impetus to get your staff to stop using them. As soon
as you have completed the project, get rid of the old databases.
We have found that planning and communication
are absolutely essential to preventing and solving technological problems
Thoroughly investigate your options.
The technological lessons have been interesting. When we decided to upgrade
the software, it was a logical step to go from DOS and Windows 3.1 to Windows
95 and Office 97. This conversion was complicated by a non-standard installation
of Windows 3.1. Upgrading the software required a hardware upgrade. We
determined to go with Dell Pentium 2 computers with 2.1 gig hard drives.
They would replace 386 machines in the library. Ultimately, the librarians
got the new machines and their old Pentiums were placed in the LAN because
they did not have enough hard drive space for Office 97. That was a glitch
in advice. Leasing the hardware made sense because it meant that the equipment
would be kept current.
The decision to rebuild the CD-ROM
tower, however, was probably penny-wise and pound-foolish. We chose to
expand the number and speed of the drives of the tower and had been advised
that it would be as good as new. The project director signed off on the
tower before the installation was complete. Dead fans resulted in a dead
power supply. The boards were wrong. The planes were wrong. The
tower overheated and kept crashing so the air conditioning needed to be
redone The new version of CD-Commander caused the tower to crash.
The network card was wrong and caused the tower to crash every 10 minutes.
All of this delayed the implementation by 3 to 4 weeks and caused the loss
of hours and hours of labor.
Standardize your environment and allow
plenty of time for the project. A non-standard environment meant that
the project took much longer to accomplish than expected. We had decided
to go from DOS to Windows and from CD-ROM to Web access where possible.
We would also go from stand-alone access to network access where possible.
Because of the budget cuts, the user licenses for the networked CD-ROMs
would be limited to single user only. The project manager for Computer
Services had to create an image for our non-standard environment. There
was shared Windows on two servers; local Windows versions 3.1 and 3.11,
and local Windows 95A and 95B. He was not happy. Mapping the drives was
a tedious, time consuming task. What was easy in Windows 95 was complicated
in Windows 3.1. Windows 3.1 loads different versions of the dynamic link
library (DLL) files with ease. There are numerous versions and they are
not necessarily compatible. For Windows 95 this became a major headache.
Each program had to be loaded piece by piece. Once installed, the order
in which the various programs were opened needed to be checked. Each and
every one of these mysteries had to be solved and each and every one of
these mysteries took a great deal of time. It was important to be there
for support and information. It was important to explain to those who did
not understand just what was taking so long. It was not an easy process.
Once the drives were mapped, the process
was automated so that it would appear seamless to the user. S/he would
click on and the program would simply open. WinBatch Compiler was
used. Microsoft client for NetWare Networks worked well. Novell
NetWare did not. WinFrame is on our wish list for the future
but well out of our current budget range.
Next the display had to be standardized.
We chose to have icons for each database and arranged them in alphabetical
order for ease of access. We used System Policies to locked down
the system so that it would be tamper resistant. "Network Neighborhood"
was removed. Access to DOS was eliminated. The display can not be changed.
Icons can no longer be moved or deleted. And finally a ghost image of the
computer was created so that it could be easily and quickly reloaded in
case of a problem. We are currently testing a program to do remote maintenance
of the system. Automation of the project has yet to be completed.
Plan for security. Security issues
need to be dealt with in other ways. We have chosen to use IP authentication
for allowing access to the database subscriptions over the web. It would
be too cumbersome to deal with passwords and distribute them to 6000 students,
etc. We do not currently have dial-in access.
One problem we have had to deal with
is the change in URLs when vendors have had to make corrections. If at
all possible, urge vendors to allow you to keep the same URL. If you have
distributed handouts or manuals to your new system, a change in URL could
In addition to handouts for our customers,
we have provided bookmarks to all of our Web databases on our browser.
That way, if they do not want to or cannot get to our web page, they can
still go to the databases. This has already proved handy when we were teaching
an Information Instruction class and we could not get our web page to open.
Plan for access. We had hoped to
have access to all of our databases from our web page. With our budget,
we were advised that would not be possible. We, therefore, determined that
we would have to instruct our customers about alt+tab to switch
to Windows from the web page.
The Web page is a work in progress.
It has much of the usual information about hours and policies. There are
subject area links that are useful to areas of the curriculum. Then there
are links to our databases. Another division handles corrections to our
pages. We have had to deal with personnel changes here and the necessity
to justify our goals repeatedly. In addition, the time involved in checking
and rechecking pages for accuracy and currency is extraordinary. Add to
that the time involved in explaining what one wants to each new employee,
and you have much time that could be more productively spent. It also is
not being updated in a timely and appropriate fashion. Corrections delivered
in June have just been loaded in January. It is important to have control
of your own web page if at all possible. We have requested that we be allowed
to update and upload our own pages.
In reviewing the lessons learned three
key factors emerge. They are communicating, planning, and training.
Share the vision, informally and formally
with all levels
Ask questions; solicit advice;
involve peers with previewing
Give plenty of advance warning of impending
Communicate anticipated stages
Consider how change might impact individuals
Include them in project and planning meetings
Share some of your frustrations
Develop a thorough project proposal and
Provide anticipated change timetables
Create a resources transition chart
Create network bookmarks for Web resources
Learn web editing
If attention is given to each of these
factors, your conversion project has a far better chance of success.
Train the trainers
Update all teaching materials
Conduct separate training sessions for
peers, faculty, staff, administrators